US 20190179861A1 (19) United States (12) Patent Application Publication (10) Pub. No.: US 2019/0179861 A1 Goldenstein et a1. (43) Pub. Date: Jun. 13, 2019 (54) (71) (72) (21) (22) (63) (60) CONTAINING DISINFORMATION SPREAD USING CUSTOMIZABLE INTELLIGENCE CHANNELS Applicant: CREOpoint, Inc., Cliifside Park, NJ (US) Inventors: Jean-Claude Goldenstein, San Francisco, CA James E. Searing, New Hope, PA Edward J. Finn, Cliffside Park, NJ (US) Appl. No.2 16/268,329 Filed: Feb. 5, 2019 Related US. Application Data Continuation-in-part of application No. 15/642,890, ?led on Jul. 6, 2017, now Pat. No. 10,223,465, which is a continuation of application No. 14/772,598, ?led on Sep. 3, 2015, now abandoned, ?led as application No. on Mar. 11, 2014. Provisional application No. 61/776,587, ?led on Mar. 11, 2013. 12 Publication Classi?cation (51) Int. Cl. G06F 16/9535 (2006.01) G06F 16/248 (2006.01) G06F 16/2457 (2006.01) (52) U.S. Cl. CPC G06F 16/9535 (2019.01); G06F 16/24575 (2019.01), G06F 16/248 (2019.01) (57) ABSTRACT Techniques are provided for rating the veracity of content distributed via digital communications sources by creating an ontology and selecting keywords for a topic of the content, creating a customizable intelligence channel for the topic, and extracting from the customizable intelligence channel a ?rst list of potential experts on the topic sorted by at least relevance and in?uence. The list of experts may be supplemented by mining trusted media sources to extract a second list of potential experts or witnesses on the topic. The ?rst and second lists of potential experts are evaluated as a function of at least one of professionalism, reliability, prox- imity, experience, responsiveness, and lack of self-interest in the topic to identify a short list of experts. The content is provided to the short list of experts, who are polled about the veracity of the content to create a veracity score for delivery with the content. Chennai Prepmduction 14 Chame? Development 21.5 Aggregate, Filter, Qua?itv Test 1.3 Assemble 22 Update if 20 Provide Top?cmReiated information 12 Channeiprepmductim M- Chamef D-?ve?apment 15 \m Aggregate, F?iter, Quaiity Test 18 22 Assemb?e Update ?j Al 263 Fmvme Topic~Re?ated infarmatim HG, Patent Application Publication Jun. 13, 2019 Sheet 1 0f 18 US 2019/0179861 A1 Channel Prepmduction 24 {Determine A Topic?s} T0936 Data Mining And Researah Qetermine Search Criteria 26 Patent Application Publication Jun. 13, 2019 Sheet 2 0f 18 US 2019/0179861 A1 (Shame? D-eveiapment 39 Betermi?e A Scurcds) 32 1. \m impiemem Search Criteria 34 -Perfmm Test Search 33 35 znca abie A riate Ne Remove Source . . . - pp Quaiity? Yes 46. . Another YES SourceDetermine Structure And meat Fer Search Queries HG. 3 Patent Application Publication Jun. 13, 2019 Sheet 3 0f 18 US 2019/0179861 A1 Aggregate, Fiiter, Quaiity "Fest 44 impiement Search Query 45 Aggregate ?8 Giebai Fitter 3 Scarce Fiiter .. Dupiicate Fiiter .. N05523: Fitter - meaniiy Fi?-?er Pass 54 Cha?ng! Based Fiiter - Language Fi?ter it Saurce Hitter Noise Fiiter - Prmfa?ity Fi?ter 55 .. . Rem-0V9 Source/information Pass N03 \53 59 Cieanup And Enhanced Formatting 52 Fiitering? N0 52 5 Stare-Resuits Patent Application Publication Jun. 13, 2019 Sheet 4 0f 18 US 2019/0179861 A1 64 EChanneEj Channef Channei ECha?neE: Channei 63 Mare Channels Channet Chamne? Channei iChamnei: Channe? Channef Chennai Channei EChannei: Channei Channef Channei Channei Channe? Channei aw?65 Patent Application Publication Jun. 13, 2019 Sheet 5 0f 18 US 2019/0179861 A1 34 MTGPEC Curate 72 y?wmgre Channais Share Source-21 Saurcef? Saurced 6? 9 9 Somme?? 73 76 35 KM M36 Artiste 3. Base-Timmy} {Titia} 1cm Time Source Articie 2 Bescript?on {Titie} icon Time Source . I van-32 Arti?cie 3 Bess-ript?m (Tithe) icon Time I Sauna Articie ?1 Description {Titia} {can Time I Source 3 0 8 Patent Application Publication Jun. 13, 2019 Sheet 6 0f 18 US 2019/0179861 A1 Channeis Patent Application Publication Jun. 13, 2019 Sheet 7 0f 18 US 2019/0179861 A1 5:2: >E=mnmmo= mic?moans. Hm, we; mrga (In/d (. 3:2: S. we; om 3 dm 2:93:.5555? .3 23215:?? #3 A.) W0 Wm. . . .3. 30.. 47M . a . 4.?.35 . 27x. 3 ulm. 957.?aw(2.. .. a5 a . .Hwmy.? . . ?wt. . . MM .5 z. 2.03? ?2?Wu. aw?2.3 C. .. v. ?12up. we.vvr.fb? . ?hm 31w J/?.13.32 106 115 113 \w 3.13 Updates Pram-5505'- Data Appiication Layer 112 Database Layer 126 Data ha 353 122 Search Engine: 124 Web Layer 126 Admimstratwn User inter-fa ce Enterfa ce HG. Patent Application Publication Jun. 13, 2019 Sheet 10 0f 18 US 2019/0179861 A1 132 MG Database . . . . 133 Load Baiancer Metadata Extra-3cm 149 {Bate Fi?ters Request Manager .. 142 144 Processor Data Narmaiizer A 134 1.45 k? Query Builder [?esta Aggregates" 3.36 Sources Patent Application Publication Jun. 13, 2019 Sheet 11 0f 18 US 2019/0179861 A1 US 2019/0179861 A1 if? if? {3.5.1 t? 22' 2232 2222222222 2:222; 222222 2222222 2: 22: 22222 Li 22-2 . ?2222212222222 :22" 2222222., ?222 :2 222222 22222222222222; Jun. 13, 2019 Sheet 12 0f 18 Patent Application Publication 23%? @531 i . than.? ~11111- 1m.? In?; an" a. 2 222222222222} . up? .2 m. .. Apparatus 192 Remevame Storage ?532 134 Non-Removabie 188 VGiatiie Stmrage Patent Application Publication Pracegsing 1&6 Nan-Vaiatzie 193 input Devisem 19% Output Device(s) 293 135 Cannec?mm Jun. 13, 2019 Sheet 13 0f 18 13 US 2019/0179861 A1 Patent Application Publication Jun. 13, 2019 Sheet 14 0f 18 US 2019/0179861 A1 ?k 226 Processing Memory inputiQutpu?t User lnte?a?e 2% HQ. '14 US 2019/0179861 A1 Sheet 15 of 18 2019 9 Jun. 13 Patent Application Publication $2 $23{an'n'qa 13:1 535:3; i?gtga's?zgmwisd 6231:5122 #3 Wf?gi?i?? 51535wa 19% 3:112me 35$ Zain?i?i ?1 31311 3H Patent Application Publication Jun. 13, 2019 Sheet 16 0f 18 US 2019/0179861 A1 Rhea? Fact? ?11.11}? 6: Wing stats iniaus u?miizw ?news?? So we? ?C?REOpgdm (gramme mm: were pars-ides a haw-mi}: ?32173: hi?asizasis: gm wanna: gum-ii}! ?sh} em Miami-2'; $231235 Subj?ini ma?a: awn: areaxgmg; 21x39: 33o ibis is {223:9 m. 525hggiuh gamma-$3 mew papmxf VA -L wan-am {Kw-u 2 war. mm} or: warmly-5am in ibis case mi. F0 ?xmed fay se'ummi Mamas-wk}: iivks the. cia'sm {mam 52m; 315E 'E?V'Rgp?n?w Witms: 3 aha? 13m ?fths:- Suit-25 mum: FIGURE 16 Patent Application Publication Jun. 13, 2019 3?3 FIGURE Sheet 17 0f 18 US 2019/0179861 A1 Grazia. amiss- 53 mm imagissgzi I camexpms UniQue. new EREOpoin'tuser interface a?ows user to. 513mm wag/chestspumes ,?hnuki'he: 3,:in {mama-?3 .exg?g?a @733; Ei?fs?} I ma .1: 17 Patent Application Publication Jun. 13, 2019 Sheet it?! 2" Kim}: T3 32}? ?eet: is: {gig-53:93:96. {si?mz?iugg Eta? US 2019/0179861 A1 FIGURE 18 US 2019/0179861Al CONTAINING DISINFORMATION SPREAD USING CUSTOMIZABLE INTELLIGENCE CHANNELS CROSS-REFERENCE TO RELATED APPLICATIONS [0001] This application is a continuation-in-part of US. patent application Ser. No. 15/642,890, ?led Jul. 6, 2017, which is a continuation application of US. patent applica- tion Ser. No. 14/772,598 ?led Sep. 3, 2015, which is the National Stage of Intemational Application No. ?led Mar. 11, 2014, which claims the bene?t of and priority to US. Provisional Application No. 61/776,587 ?led Mar. 11, 2013, the entireties of which applications are incorporated herein by reference for any and all purposes. COPYRIGHT NOTICE AND PERMISSION [0002] A portion of the disclosure of this document may contain material that is subject to copyright protection. The copyright owner has no objection to the facsimile reproduc- tion by anyone of the patent document or the patent disclo- sure, as it appears in the Patent and Trademark O?ice patent ?les or records, but otherwise reserves all copyright rights whatsoever. TECHNICAL FIELD [0003] The technical ?eld generally relates to searching for and providing information, and more speci?cally relates to customizable provision of information and containing the spread of disinfomiation. BACKGROUND [0004] 100+ social networks offer billions of people a growing ability to consume and produce news. Unfortu- nately, intentionally and veri?ably false information (re- ferred herein as ?fake news?) misleads readers, and these falsehoods that spread through social media ?take root quickly and die hard.? A person skilled in the art would appreciate the different types of fake news including disin- formation, misinformation, fabricated hoaxes, satire, B.S. news, etc. Thought different in meaning, these are generally referred to herein as ?fake news? for ease of description. This purported ?news? impacts public con?- dence, public health and well-being, personal and organiza- tional reputations, the markets and democracy. [0005] Each day for example Facebook and Twitter pub- lish hundreds of millions of new story link updates and tweets, including too many lies to imagine. If a lie in a story is picked up without critical context by multiple clickbait sites, dozens of other blogs, cross-posted over thousands of social media accounts and read by hundreds of thousands, then it becomes fake news. It does not help that algorithms sometimes favor sensational content over substance. [0006] Fighting fake news is a race against time, similar to stopping a ?re in its tracks before it spreads as wild?re and causes irreversible damage. Social networks and society are increasingly challenged by the di?iculty of rapid detection and di?usion of disinformation, despite investments in machine leaming, information literacy, third-party ?agging, fact-checking programs, and news feed updates. [0007] Fact checking is an intellectually demanding pro- cess that often requires hours and sometimes days to com- Jun. 13, 2019 plete. In the case of disinformation impacting a company, the public relations ?rm Edelman reports ?it takes an average of 21 hours before companies are able to issue meaningful external communications to defend themselves.? Layers of organizational approvals lead to turnaround times too slow to work and the optimal prevention window is often missed. During that time, more damage is done in early reposts by many users, creating periods of confusion and uncertainty. [0008] In addition, some extremists and hostile foreign governments are applying social engineering concepts to manipulate social media platform algorithms to further their agenda, potentially endangering brand perceptions reputa- tions of long established companies, products, as well as the personal brand image of any person in the public eye. [0009] Disinformation is not new. In the 19905, Russian agents planted a story that the US had created the AIDS virus in a lab. However, ?Falsehood ?ies, and the truth comes limping after it? is more relevant now than ever before. Rumors spread much faster due to the eifect of social media operating as an echo chamber as more people became active communicators rather than passive receivers. The problem is compounded by the ?4 Vs of news?: growing Volume (from a few newspapers to billions of sources), Variety Tweet, Instagram post, WhatsApp message, deepfake, etc.), and Virality (images, videos, or links that spread rapidly through a population by being frequently shared with a number of individuals)idecreasing their Veracity. [0010] A 2018 Gallup/Knight Foundation survey about perceived accuracy and bias in the news media shows that Americans believe that 64% of the news and information they see on social media is inaccurate. Edelman?s 2019 Trust Barometer reported that 73% of the people surveyed are worried that fake news is being used as a weapon. [0011] Fake news proliferation is an adversarial attack on our information ecosystem and a new breed of cybersecurity threat. It is partly a ?inction of information overload and the failure to have su?icient ?ltering tools. To date, techniques to conveniently address information overload have been found to be impractical. Also, previous attempts at ?ltering at scale and other approaches have fallen short. The frus- trations are often centered on solutions that are ?too little, too late,? similar to trying to clean up the internet by building a ?rewall or attempting to purify the entire Ganges river when pervasive pollution, or vitriol, is in the air and the public?s cognitive infrastructure is under siege. [0012] Improved ?ltering techniques and alternative meth- ods are needed to contain the threats and help slow or halt the spread of potentially damaging fake news in a high-stakes information war. These elforts remain a critical and urgent need in the art. SUMMARY [0013] Various examples are now described to introduce a selection of concepts in a simpli?ed form that are further described below in the Detailed Description. The Summary is not intended to identify key or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter. [0014] A system is described herein for creating ?customi- zable real time intelligence channels? that enables time starved people to more conveniently get to the point about their topics, brands and personalities of interest. This is generally accomplished by ?ltering traditional and social media to: US [0015] Remove duplicates (?old news? identi?ed with Natural Language Processing are not displayed, which will help decrease social reinforcement and contagious- nessitherefore popping the ??lter bubble?); [0016] Identify traditional and social media sources with the highest relevance and in?uence; and [0017] Allow users to delete some news content or even ban bad or fake news sources. To contain the spread of fake news, in sample embodiments described herein, these unique capabilities are expanded by adding a hybrid human expert network augmented by AI. A practical method to reduce this potentially hamiful content enables a networked platform of relevant active experts (rather the current ?wisdom? of the crowds, one of the causes of the infocalypse). [0018] Recent MIT Sloan research indicates lay people lazily judge around 40% of legitimate news stories as false and 20% of fabricated news stories as true (Academy of Sciences?David Rand, January 2019 ?Fighting misinfor- mation on social media using crowdsourced judgments of news source quality?). The CREOpoint system enables the preparation and creation of a visible and shareable measur- ing stick, namely, the probability that news content a breaking news story, deepfake video, doctored photo, viral WhatsApp message, or forged document) is fake. A CREO- point Veracity Score (called is assigned to the evaluated news content. For example, a minimum CREO- Score means the panel of experts is completely convinced that the news content is a fake. The Veracity Score can be transposed to develop a ?Fake News Warning.? For example, a CREOpoint Veracity Score of 14 out of 100 would correspond to a high 86 (100-14) fake news indicator, and a red warning would be automatically enabled. [0019] Fighting fake news is a race against time with similarities to the more established and regulated ?elds of pandemics, wild?re and aircraft structure crack propagation. Sharing the CREOWarning early and intervening before the news content has spread past a tipping point (similar to ?re extinguishers quickly used by professional ?re?ghters to stop a damaging wild?re) helps: [0020] Combat fake news with the AI indicators of unreliability and the consensus of crowdsourced experts; [0021] Provide con?dence to society about the level of veracity of news content; [0022] Quickly share warnings to enable the public to make more informed decisions before disinformation has gone viral; [0023] Raise some doubt to encourage readers to pause, be skeptical, and use more critical thinking; and [0024] Decrease the consequences of fake news on elections, democracy, public con?dence, organizational and personal reputations, public health and well-being, and product or equity markets. [0025] Sample embodiments described herein are further directed to a computer-implemented method of rating the veracity of content distributed via digital communications sources. The methods include the steps of creating an ontology and selecting keywords for at least one topic of the content and creating a customizable intelligence channel for the at least one topic of the content. A ?rst list of potential experts on the at least one topic of the content is extracted from the customizable intelligence channel and sorted by at least relevance and in?uence. Trusted media sources are also Jun. 13, 2019 mined for the at least one topic of the content to extract a second list of potential experts on the at least one topic of the content. The ?rst and second lists of potential experts on the at least one topic of the content are provided to a database, and the potential experts are rated and ranked as a function of at least one of professionalism, reliability, proximity, experience, responsiveness, and lack of self-interest in the at least one topic of the content to identify a short list of experts. The content is provided to the short list of experts for evaluation, and the short list of experts are polled about the veracity of the content to create a veracity score. The veracity score is then delivered with the content. [0026] In sample embodiments, the method further includes creating a third list of potential experts and any local witnesses on the at least one topic of the content based on at least one of a relationship and a proximity of the potential experts to a breaking story on the at least one topic of the content and providing the third list of potential experts to the database to complement the polling. The polling of the short list of experts about the veracity of the content to create a veracity score may occur in near real-time. [0027] In other sample embodiments, delivering the verac- ity score with the content comprises at least one of issuing a pre-populated press release, initiating a social media and press campaign including the veracity score and at least one of a warning and denial if the content is not completely true, issuing a quote from an expert from the short list of experts, issuing a quote from a local witness to the at least one topic of the content, and issuing a quote from an influencer on the at least one topic of the content and related reassuring metrics including information about trustworthiness of sources of the content. A fake news warning also may be presented with the veracity score and content along with insights and metrics relating to the content. In the sample embodiments, the veracity score, content, fake news wam- ing, insights and metrics relating to the content are delivered via an interactive interface enabling a user to select the types of sources by level of trust or proximity to the news content or user. [0028] In other sample embodiments, providing the ?rst and second list of potential experts on the at least one topic of the content to a database includes predetermining a list of experts who could best crowdsource veracity signals for a given topic of content. Also, mining trusted media sources for the at least one topic of the content to extract a second list of potential experts on the at least one topic of the content may be performed upon the release of a new story. [0029] In still other sample embodiments, the methods include incentivizing experts to be active in near real-time when consulted by compensating experts based on how accurate their predictions are and creating a decentralized register including expert trust ratings. [0030] In yet other sample embodiments, the methods include creating at least one customizable intelligence chan- nel for at least one topic of the content relating to potential sources of fake news and the semantics of fake news content. [0031] In yet other sample embodiments, a decision matrix is created to evaluate a breaking news story to decide whether the news story is a candidate for determining a veracity rating based on at least one of the nature of the breaking news story, a source of the news story, and whether a relevant population of experts readily exists. US [0032] In yet other sample embodiments, the methods include modifying the veracity score to re?ect the behavior of additional experts and trusted sources in sharing and commenting upon a breaking new story. [0033] In yet other sample embodiments, the methods further include benchmarking the veracity scores to create a predictive fake news spread containment model and iterating to revise the model and overall performance of the model over time. [0034] Any one of the foregoing examples may be com- bined with any one or more of the other foregoing examples to create a new embodiment within the scope of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS [0035] Aspects of a customizable, real time, intelligence channel are described more fully herein with reference to the accompanying drawings, in which example embodiments are shown. In the following description, for purposes of explanation, numerous speci?c details are set forth in order to provide an understanding of the various embodiments. However, the instant disclosure may be embodied in many different forms and should not be construed as limited to the example embodiments set forth herein. Like numbers refer to like elements throughout. [0036] FIG. 1 is ?ow diagram of an example process for generating a customizable, real time, intelligence channel and providing information via the intelligence channel. [0037] FIG. 2 is a ?ow diagram of an example process for intelligence channel preproduction. [0038] FIG. 3 is a ?ow diagram of an example process for intelligence channel development. [0039] FIG. 4 is a ?ow diagram of an example process for aggregating, ?ltering, and quality testing. [0040] FIG. 5 depicts an example interface comprising a plurality of intelligence channels. [0041] FIG. 6 depicts an example illustration of an intel- ligence channel interface. [0042] FIG. 7 illustrates an example interface for provid- ing access to a plurality of intelligence channels. [0043] FIG. 8 illustrates an example interface that pro- vides topic related information pertaining to incubators. [0044] FIG. 9 is an example functional block diagram for developing an intelligence channel. [0045] FIG. 10 is an example block diagram of a system for generating an intelligence channel and providing infor- mation via an intelligence channel. [0046] FIG. 11 is an example block diagram of the updates processor. [0047] FIG. 12 is an example block diagram of the data processor. [0048] FIG. 13 is a block diagram of an example apparatus that may be utilized to implement and/or facilitate an intelligence channel. [0049] FIG. 14 is a block diagram of an example device that may be utilized to generate and/or implement an intel- ligence channel. [0050] FIG. 15 is a sample block diagram of a system for containing disinformation spread using customizable intel- ligence channels in a sample embodiment. [0051] FIG. 16 illustrates a sample presen- tation of an evaluated news feed. Jun. 13, 2019 [0052] FIG. 17 illustrates a sample interface with an associated toggle for allowing users to select their desired types of news sources. [0053] FIG. 18 illustrates a sample poll for experts in sample embodiments. DETAILED DESCRIPTION [0054] It should be understood at the outset that although an illustrative implementation of one or more embodiments are provided below, the disclosed systems and/or methods described with respect to FIGS. 1-18 may be implemented using any number of techniques, whether currently known or in existence. The disclosure should in no way be limited to the illustrative implementations, drawings, and tech- niques illustrated below, including the exemplary designs and implementations illustrated and described herein, but may be modi?ed within the scope of the appended claims along with their full scope of equivalents. [0055] In the following description, reference is made to the accompanying drawings that form a part hereof, and in which is shown by way of illustration speci?c embodiments which may be practiced. These embodiments are described in suf?cient detail to enable those skilled in the art to practice the systems and methods described herein, and it is to be understood that other embodiments may be utilized, and that structural, logical and electrical changes may be made without departing from the scope of the present disclosure. The following description of example embodi- ments is, therefore, not to be taken in a limited sense, and the scope of the present disclosure is de?ned by the appended claims. [0056] The functions or algorithms described herein may be implemented in software in one embodiment. The soft- ware may consist of computer executable instructions stored on computer readable media or computer readable storage device such as one or more non-transitory memories or other type of hardware-based storage devices, either local or networked. Further, such functions correspond to modules, which may be software, hardware, ?rmware or any combi- nation thereof Multiple ?inctions may be performed in one or more modules as desired, and the embodiments described are merely examples. The software may be executed on a digital signal processor, ASIC, microprocessor, or other type of processor operating on a computer system, such as a personal computer, server or other computer system, turning such computer system into a speci?cally programmed machine. Customizable, Real-Time Intelligence Channel [0057] A customizable, real time intelligence channel as described herein may enable a user to keep apprised of, and/or deliver to others, selected topics by identifying and delivering relevant information culled from, for example, greater than hundreds of thousands of relevant sources while ?ltering out irrelevant and/or redundant information. In an example embodiment, an intelligence channel may provide the ability to identify, isolate, and access the most reliable, relevant sources for a topic, and deliver a stream of targeted, focused information pertaining to the topic. A topic may comprise any appropriate topic, and is not limited to any genre, category, or the like. [0058] As described herein, an intelligence channel may be generated for a topic. The topic may be selected from US predetermined topics, a topic may be selected from topic already available online, a topic may be provided by a user, a topic may be provided by a user and de?ned and developed through an iterative process that de?nes search terms, key- words, includes and excludes, language recognition capa- bility and prede?ned ?lters, a topic may be a speci?c topic, a topic may be a speci?c topic for a speci?c user, or the like, or any appropriate combination thereof. Topics may be determined in any appropriate manner. For example, pre- dictive data mining based on a user?s email address, social graph, prior history of search or use of an intelligence channel, or the like, may be utilized to identify likely appropriate areas of interest and channels for the user. [0059] In an example embodiment, predictive data mining may be utilized by identifying a user?s home location, and accessing publicly available information to determine, for example, the income and interests of persons located in that area to predict interests of the user and recommend an intelligence channel. Predictive data mining also may rely, in whole or part, on prior use characteristics of the user or others related to the user by location, social media, or otherwise, to predict interests of the user and recommend an intelligence channel. Predictive data mining may be facili- tated by the structure of a database, and the use of open source tools for the database, such as, for example, an open source non-SQL document database for big data analysis. [0060] Sources of topic-related information may be deter- mined in any appropriate manner. For example, a source of topic-related information may comprise any appropriate source, such as, an online source accessible via a Uniform Resource Locator (URL), social media, TWITTER, YOU- TUBE, GOOGLE, FACEBOOK, YAHOO, BING, WIKI- PEDIA, LINKEDIN, FLICKR, INSTAGRAM, PINTER- EST, a photograph, an image, a publication, a forum, a blog, a forum, a video, a periodical, a research facility, an aca- demic site, an industry speci?c site, a web site, a university, a feed, an RSS feed, or the like, or any appropriate combi- nation thereof. A source(s) may be reviewed and the respec- tive intelligence channel accordingly may be updated, or curated, based on, for example, crowdsourcing data, research by the subject matter experts, input from the user, or the like, or any appropriate combination thereof. A source(s) may be reviewed periodically, aperiodically, responsive to a request, responsive to an occurrence of an event, continuously, or the like, or any appropriate combi- nation thereof. [0061] Sources may be searched for topic-related infor- mation. Search results may be analyzed for redundant, irrelevant, immaterial, or the like, information. For example, search results may be analyzed for profanity, duplicate headlines, duplicate content, time wasters such as tweets dealing with social comments rather than topic-related items of interest, posts that are not in the user?s language, posts that are not in a designated language or languages. dead hyperlinks to useless headlines, or the like, or any appro- priate combination thereof. Analyzed search results may be ?ltered. Analyzed search results may be ?ltered to remove information from a source, remove a source, edit informa- tion from a source, or the like, or any appropriate combi- nation thereof. In an example embodiment, sources and/or the level of ?ltering may be controlled by a user in real time, while topic-related information is being provided to the user. And ?ltered information may be refreshed and provided by the intelligence channel instantly, on demand. In an example Jun. 13, 2019 embodiment, ?ltering may comprise a ?exible noise-can- celing algorithm(s) which may be updated to limit future sources of noise based on, for example, an analysis of usage and crowd sourced comments and input. Filtering also may be based on crowd-sourced determinations of authority and reliability of sources or items, based on user statistics and preferences. [0062] FIG. 1 is ?ow diagram of an example process for generating a customizable, real time, intelligence channel and providing information via the intelligence channel. The example process depicted in FIG. 1, and as described in detail herein, may comprise any appropriate combination and/or sequence of charmel preproduction (step 12), channel development (step 14), aggregating, ?ltering, and quality testing (step 16), assembling information for delivery (step 18), providing topic-related information (step 20), and updating contents of the intelligence channel (step 22). [0063] FIG. 1 is described herein with reference to addi- tional ?gures providing detail associated with each step depicted in FIG. 1. To that goal, as described above with reference to FIG. 1, channel preproduction may be per- formed at step 12. And FIG. 2 is a ?ow diagram of an example process for channel preproduction comprising any appropriate combination and sequence of determining a topic at step 24, topic data mining and research at step 26, and determining search criteria at step 28. Atopic, or topics, may be determined at step 24. A topic may be selected from predetermined topics, a topic may be determined by a user, a topic may be a speci?c topic, a topic may be a speci?c topic for a speci?c user, or the like, or any appropriate combination thereof. Topics may be determined in any appropriate manner. A topic may comprise, for example, a company, an industry, a person, a group of persons, an asset, real property, intellectual property, a stock, a subject, a political entity, a city, a state, a country, a fashion trend, an activity, an event, an idea, an invention, a work, or the like, or any appropriate combination thereof. It is to be under- stood that there is no limitation on the type of a topic. In an example embodiment, a topic may use Boolean search expressions combining multiple topics; including some aspect(s) of a topic while excluding others. A Boolean topic search logic may link of ten or more Boolean terms, despite limitations on the number of search terms imposed by some search engines. It is to be understood however, that there is no limit on the number of terms included in a topic search. [0064] Topic data mining and research may be performed at step 26. Any appropriate information may be analyzed to determine and/or suggest a topic to the user. For example, a user?s assets, email, social media accounts, search engines, GOOGLE, BING, LINKEDIN, WIKIPEDIA, user pro?le information, a web site, a company web site, a government web site, an employer?s web site, or the like, or any appropriate combination thereof, may be utilized to deter- mine a topic. Predictive data mining may be utilized to identify likely appropriate areas of interest and channels for the user. In an example embodiment, predictive data mining based on a user?s email address may be utilized to identify likely appropriate areas of interest and channels for the user. An email domain may be analyzed to identify an association, from which interest(s) may be determined. For example, it may be determined and/or inferred that a user with an @aol.com extension has had the email account for a long time but has not kept up with the times. As another example, it may be determined and/ or inferred that a name like US johndoe@gmail.com indicates an early adopter, or person@bakerlaw.com would likely have an interest in legal-related information, or one withan@woodcock.com extension would have a particular interest in IP law. It may be determined and/or inferred that a user with an email address may prefer information in the French lan- guage. [0065] Search criteria may be generated, at step 28, to facilitate a determination of potential topic-related sources. Search criteria may be inclusive or exclusive. Inclusive search criteria may be utilized to search for information containing the inclusive search criteria. Exclusive search criteria may be utilized to exclude information containing the exclusive search criteria. Search criteria may comprise, for example, inclusive keywords, exclusive keywords, inclu- sive search terms, exclusive search terms, inclusive search phrases, exclusive search phrases, inclusive queries, exclu- sive queries, information indicative of a false positive search result, redundant information, or the like, or any appropriate combination thereof. Search criteria may be combined using Boolean operators to query many sources with dilfering application programming interfaces and semantic requirements. The versions, and semantic structure requirements associated with any source may be determined for each source from, for example, a database library. [0066] As described above, with reference to FIG. 1, channel development may be performed at step 14. And FIG. 3 is a ?ow diagram of an example process for channel development comprising any appropriate combination and sequence of determining a source, or sources at step 30, implementing search criteria at step 32, performing a test search or searches at step 34, determining a quality of results at step 36, removing a source from potential sources at step 38, determining if other sources exist at step 40, and determining a structure and format for search queries at step 42. [0067] A source, or sources, of topic-related information may be determined at step 30. A source of topic-related information may comprise any appropriate source, such as, for example, social media, TWITTER, YOUTUBE, GOOGLE, FACEBOOK, YAHOO, BING, WIKIPEDIA, LINKEDIN, FLICKER, INSTAGRAM, PINTEREST, a photograph, an image, a publication, a forum, a blog, a forum, a video, a periodical, a research facility, an academic site, an industry speci?c site, a web site, a university, a library, audio, radio, television, a feed, and RSS feed, or the like, or any appropriate combination thereof. A source may include any data accessible in digital form through a URL and may include non-textual data such as that developed by sensors in buildings, homes, vehicles, appliances, and other monitoring devices and available digitally through a URL. Topic-related information may be selected and curated from all available sources. A source may be reviewed and, as described in more detail herein, the respective intelligence channel may be updated based on, for example, crowdsourc- ing data, research by the subject matter experts, input from a user, or the like, or any appropriate combination thereof. [0068] In an example embodiment, an API for each source is determined. An API for a source may be unique for a respective source. That is, a source may have associated therewith a unique API, or any appropriate version thereof. A description of each API for each source may be stored, for example, in a database. Source may include individual Jun. 13, 2019 websites, and accordingly, a database may be established to store API variations, even when websites may appear to use a common API. [0069] Search criteria may be implemented, at step 32, to facilitate determination of a source, or sources, of topic- related information from potential sources. In an example embodiment, search criteria determined at step 28 may be implemented at step 32. Search criteria may be implemented on each prospective source. Search criteria may be imple- mented utilizing an appropriate semantic protocol, which may include appropriate Boolean terms and operators, required by a respective source. In an example embodiment, search criteria may be implemented utilizing an appropriate semantic structure and the digital format required by a respective source. [0070] A test search, or searches, to assess the quality of information from a prospective source may be performed at step 34. Search results may be used in an iterative process in which search criteria that control the search may be altered until the prospective source produces the appropriate quality results. If a prospective source does not produce appropriate quality results, the prospective source may not be selected as a source of topic-related information for the intelligence channel. It may be determined, at step 36, if the quality of information from a prospective source is appro- priate. If it is determined, at step 36, that the quality of the information from a prospective source is not appropriate, the process may proceed to step 32 to re?ne search criteria, and to implement re?ned search criteria, and proceed therefrom as described herein. If, after an appropriate attempt to re?ne search criteria and to adjust the quality of the information from a prospective source, it is not possible to improve the quality to an acceptable level, the prospective source may be remove from a list of prospective sources no longer considered). [0071] Quality initially may be determined during the channel development process to relate to the relevance of the returned items from a source to the topic. Keywords, include and exclude criteria as well as sources included or excluded may be adjusted to improve quality of an intelligence channel. Intelligence channel quality may be determined by crowdsourced information developed through the actions of intelligence channel users in visiting, deleting, or curating the source from the intelligence channel. As each user may rate an item, or delete an item from the user?s intelligence channel, aggregate actions by users may be used to deter- mine the quality of a source or particular item to the intelligence channel, as a source frequently visited may be deemed to be of high quality and one frequently curated out of the intelligence channel by users may be deemed to be of low quality. The quality determination within an intelligence channel may be applied in other intelligence channels, or not, as appropriate. [0072] If it is determined, at step 36, that the quality of information from a prospective source is appropriate, it may be detennined, at step 40, if another prospective source is to be tested. If it is determined, at step 38, that another prospective source is to be tested, the process may proceed to step 32 and proceed therefrom as described herein. If it is determined, at step 40, that there is not another source to be tested, a series of semantic protocol/digital format combi- nations for search queries for each source (selected from prospective sources) for a given channel topic may be determined at step 42. The semantic structure and digital US 2019/0179861Al format of each search query may be saved for future use. The ?nal semantic structures and digital formats may be avail- able for fusion with other intelligence channels in the development of new and/or additional intelligence channels. [0073] As described above, with reference to FIG. 1, aggregating, ?ltering, and quality testing may be performed at step 16. And FIG. 4 is a ?ow diagram of an example process for aggregating, ?ltering, and quality testing com- prising any appropriate combination and sequence of imple- menting a search query at step 44, aggregating information from sources at step 46, performing global ?ltering at step 48, determine information that has passed global ?ltering at step 50, removing a source and/or information that has not passed global ?ltering at step 52, performing channel based ?ltering at step 54, detemiine infomiation that has passed channel based ?ltering at step 56, removing a source and/or information that has not passed channel based ?ltering at step 58, performing clean up and enhancing formatting at step 60, and storing results at step 62. [0074] A search query may be implemented at step 44. The search query may be performed on each topic -related source of the intelligence channel. Each source may be queried with a search query comprising appropriate search parameters based on the user?s selected topic. A search query may comprise search criteria as previously described. Each search query may be formatted to meet needs of a respective source. In an example embodiment, search queries compris- ing appropriate search parameters may be provided to each source in a format appropriate for each source. Accordingly, results topic-related information) may be received from each source. For example, TWITTER allows a Bool- ean search structure which groups operations using paren- theses to indicate orders of operations, so that APPLE AND (iPhone OR iPad or iPod). GOOGLE does not allow the use of parentheses in the semantic structure or to indicate an order of operations, so the same search must be structured as: ?Apple iPhone?, ?Apple iPad?, ?Apple iPod?. [0075] Results may be aggregated at step 46. Aggregation may require data-normalization, as results may vary in their data structure. Sources may provide results in differing data structures or formats. Results may be mapped to a database based on the particular data structure employed by the source. Results may be aggregated using formatters which process incoming data using data mappings based on the source and may be stored in a standard data structure. Results may be stored in a database). As additional searches are conducted, results may be added to (inserted into) stored results. Aunique index to an item URL ?eld may be generated. During storage of results, if an item is deter- mined to be in the URL ?eld, the URL ?eld may be skipped (result not added to, or inserted into, stored results). During storage of results, if an item is determined not to be in the URL ?eld, the results are added to the stored results. [0076] Filtering may be performed prior to storage of results. Filtering may affect storage of results and aggrega- tion. For example, an in?uencer blacklist may be performed before the insertion process. For website resources, the in?uencer may be the website itself. For example, hu?ing- tonpost.com, for social media sources, the in?uencer is the actual user. Entire websites may be banned on domain level or top-level domain, .cn, .biz, .casino, etc., social media user on user level. Thus, items resulting from these in?u- encer ?lters may be skipped and not stored. In an example Jun. 13, 2019 embodiment, the results from each source may be combined and incorporated into a data stream. The data stream may be ?ltered. [007 7] The data stream may be globally ?ltered at step 48. Global ?ltering may comprise any appropriate combination of source ?ltering, duplicate ?ltering, noise ?ltering, or profanity ?ltering. Source ?ltering may be performed at step 48. Source ?ltering may be utilized to remove irrelevant sources and/or spam. In an example embodiment, source ?ltering may compare a dictionary of continuously updated domain names and social media usernames with the article source URL. If there is a match, the item may be tagged and identi?ed for further review. If information does not pass source ?ltering, as determined at step 50, the source and/or information may be removed from the data stream at step 52. If information does pass source ?ltering, as determined at step 50, upon completion of global ?ltering, the process may proceed to step 54. [0078] Duplicate ?ltering may be performed at step 48. Duplicate ?ltering may compare individual article headlines and/or articles to identify and remove duplicate articles. In an example embodiment, article may be compared in chronological order to identify and remove duplicates articles and/or article that are somewhat close (comprise duplicative information). If duplicates are identi?ed, they may be removed. If information does not pass duplicate ?ltering, as determined at step 50, the source and/or infor- mation may be removed from the data stream at step 52. If information does pass duplicate ?ltering, as determined at step 50, upon completion of global ?ltering, the process may proceed to step 54. For example, many news items may be repeated, and many press releases may be printed in multiple publications. To avoid wasting a user?s time, these duplicate items may be ?ltered out by headline and/or by content. [0079] Noise ?ltering may be performed at step 48. Noise ?ltering may remove items that do not add value, that are irrelevant to the topic, or the like, or any appropriate combination thereof. In an example embodiment, noise ?ltering may identify items of limited relevance and/or items that lack substantive content. Noise ?ltering may remove items of limited relevance and/ or items that lack substantive content from the data stream. If information does not pass noise ?ltering, as determined at step 50, the source and/or information may be removed from the data stream at step 52. If information does pass noise ?ltering, as determined at step 50, upon completion of global ?ltering, the process may proceed to step 54. [0080] Profanity ?ltering may be performed at step 48. Profanity ?ltering may compare keywords, phrases, images, video, etc. a dictionary) with articles obtained from a respective source to identity profanity in the article. In an example embodiment, the profanity ?lter the diction- ary) may be tailored for a particular user. For example, a parent may tailor the profanity ?lter to remove any infor- mation that the parent may determine to be inappropriate for children. If infomiation does not pass profanity ?ltering, as determined at step 50, the source and/ or information may be removed from the data stream at step 52. If information does pass profanity ?ltering, as determined at step 50, upon completion of global ?ltering, the process may proceed to step 54. [0081] Channel based ?ltering may be performed on the data stream at step 54. In an example embodiment, channel- based ?ltering may comprise any appropriate combination US of language ?ltering, source ?ltering, or noise ?ltering. As described herein, channel ?ltering operates strictly within a particular intelligence channel, while global ?ltering oper- ates universally. For example, assume an intelligence chan- nel in which the topic is high performance automobiles. The intelligence channel may include all manufacturers, but an intelligence charmel could be ?ltered for use by General Motors to remove all other manufacturers. [0082] Language ?ltering may be performed at step 54. Intelligence channels may be ?ltered by speci?c language. In an example embodiment, if a language is, or one or more speci?c languages are, designated, all nondesignated lan- guages, as determined at step 56, may be ?ltered out at step 58. That is, all articles in a language other than a designated language, or languages, may be removed from the data stream at step 58. And, upon completion of channel-based ?ltering, the process may proceed to step 60. In an example embodiment, a nondesignated language may be translated to a designated language. [0083] Source ?ltering may be performed at step 54. Source ?ltering may be utilized to remove irrelevant sources and/or spam. In an example embodiment, source ?ltering may compare a dictionary of continuously updated domain names and social media usemames with the article source URL. If there is a match, the item may be tagged and identi?ed to be excluded or for ?irther review. If information does not pass source ?ltering, as determined at step 56, the source and/or information may be removed from the data stream at step 58. If information does pass source ?ltering, as determined at step 56, upon completion of global ?ltering, the process may proceed to step 60. [0084] Noise ?ltering may be performed at step 54. Noise ?ltering may remove items that do not add value, that are irrelevant to the topic, or the like, or any appropriate combination thereof. In an example embodiment, noise ?ltering may identify items of limited relevance and/ or items that lack substantive content. Noise ?ltering may remove items of limited relevance and/ or items that lack substantive content from the data stream If information does not pass noise ?ltering, as determined at step 56, the source and/or information may be removed from the data stream at step 58. If information does pass noise ?ltering, as determined at step 56, upon completion of global ?ltering, the process may proceed to step 60. [0085] Profanity ?ltering may be performed at step 54. Profanity ?ltering may compare keywords, phrases, images, video, etc. a dictionary) with articles obtained from a respective source to identity profanity in the article. In an example embodiment, the profanity ?lter the diction- ary) may be tailored for a particular user. For example, a parent may tailor the profanity ?lter to remove any infor- mation that the parent may determine to be inappropriate for children. Any article determined to contain profanity, may be removed from the data stream If information does not pass profanity ?ltering, as determined at step 56, the source and/ or information may be removed from the data stream at step 58. If information does pass profanity ?ltering, as determined at step 56, upon completion of global ?ltering, the process may proceed to step 60. [0086] Article cleanup and enhanced formatting may be performed at step 60. Advertisements may be removed from article content. Content may be formatted as needed in order to provide information via an intelligence channel. Semantic analysis of article content may be performed in order to Jun. 13, 2019 select photographs, places, names, companies, addresses, and/or phone numbers in order to enhance channel de?ni- tion, generate word clouds, and deliver enhanced content. Natural language processing may be employed to highlight elements of interest, such as names of persons or entities, monetary values, locations, or the like. Final results may be may be stored at step 62. In an example embodiment, stored results may be available for delivery to any user selecting that channel. Stored results may be available for combina- tion with other channels. Stored results may be available for updating. In an example embodiment, if a result is combined and/or updated, a time associated with the combination and/or update may be stored. Users selecting the channel within a predetermined time interval may receive the stored information. A user inquiry received outside of the prede- termined time interval may trigger anew inquiry in order to obtain fresh and timely results. [0087] As described above, with reference to FIG. 1, assembly of information may be performed at step 18, topic-related information may be provided at step 20, and information may be updated at step 22. Assembled infor- mation may be provided via an intelligence channel via a customized interface. The interface may be in the form of a hub comprising a customized dashboard. Results received for each channel ordered by a user may be combined and delivered via the customized hub. [0088] FIG. 5 depicts an example interface 64 comprising a plurality of intelligence channels 66. Each channel depicted in FIG. 5 represents a different intelligence chan- nel. An intelligence channel may be a predetermined (pre- de?ned) channel provided by the system, an intelligence channel may be determined by a user, or any appropriate combination thereof. Thus, the plurality of intelligence channels 66 may represent a plurality of predetermined intelligence channels, a plurality of user-de?ned intelligence channels, a plurality of intelligence channels of a user?s favorite list, or any appropriate combination thereof. As depicted in FIG. 5, more intelligence channels may be accessed by selecting item 68 labeled as ?More Channels.? [0089] The interface 64 may be considered as a hub-like center, or e-store, via which a user may access intelligence channels. Channels and news may be delivered to the user through the hub like interface 64. The hub-like interface 64 may be presented in any appropriate manner and/ or format. In an example embodiment, the hub-like interface 64 may be presented as ?my Channels? page, or the like, which may be accessible by being displayed on all website pages. The hub-like center may be applied to channels already selected by a user or to additional channels which may be available. [0090] An intelligence channel as rendered via the inter- face 64 may comprise any appropriate information that identi?es the intelligence channel. For example, an intelli- gence channel may comprise an icon, text, video, sound, or any appropriate combination thereof that identi?es the intel- ligence channel. [0091] When an intelligence channel is selected, topic- related information may be provided via the intelligence channel. An intelligence charmel may be selected in any appropriate manner. As described herein, selection of a rendering on an interface may be accomplished in any appropriate manner, such as, for example, clicking on a rendering, tapping a rendering, touching a rendering, mak- ing a gesture over a rendering, making a gesture proximate to a, providing an audio command, or the like, or any US appropriate combination thereof. Accordingly, in an example embodiment, an intelligence channel may be selected via the interface 64 by clicking on the rendering of the intelligence channel, by tapping the rendering of the intelligence channel, by touching the rendering of the intel- ligence channel, by making a gesture over the rendering of the intelligence channel, by making a gesture proximate to the rendering of the intelligence channel, by providing an audio command, or the like, or any appropriate combination thereof. [0092] When an intelligence channel is selected, topic- related information may be provided in real time. FIG. 6 depicts an example illustration of a user friendly, interactive, intelligence channel interface 70. In an example embodi- ment, interface 70 may comprise an interactive display of an apparatus, device, server, computer, or the like. In an example embodiment, interface 70 may comprise any appro- priate combination of a selectable source portion or region 78, a portion or region allowing selection of more intelli- gence channels 74, a portion or region allowing sharing of the intelligence channel 76, an adjustable ?lter portion or region 80, and dynamically con?gurable content portion or region 82. In an example embodiment, the interface 70 may comprise a web page or the like. The topic may be rendered via the interface 70 as depicted by item 72. More channels may be accessed by selecting item 74. [0093] The selectable source region 78 may provide indi- vidually selectable access to a plurality of sources of topic- related information Source 1, Source 2, Source 3, Source 4, . . . Source N), wherein each source of the plurality of topic-related sources may comprise information related to a topic and may be individually selectable. The content region 82 may render topic-related information Article 1, Article 2, Article 3, Article 4 . . . and provide access to a source of the rendered topic-related information. The adjustable ?lter region 80 may provide adjustable ?ltering of the dynamically con?gurable content region 82, wherein content of the content region 82 may be dynamically modi- ?ed, in real time, based on the adjustable ?ltering of the ?lter render in ?lter region 80. [0094] An intelligence channel may be shared by selecting item 76. An intelligence channel and/or information pro- vided via an intelligence channel may be shared via, for example, email, social media, TWITTER, LINKEDIN, FACEBOOK, a social network, a news article, a test message, or the like, or any appropriate combination thereof [0095] Sources 78 of topic-related information may be rendered on the interface 70. Each rendering of a source may be selectable in a toggle-like fashion, wherein selection of a rendering of a source may allow information to be received from the source, and a subsequent selection of the source may prevent information from being received from the source. For example, Source 1 may represent TWITTER, Source 2 may represent GOOGLE, and Source 3 may represent a user-de?ned source. Selecting Source 1 may allow topic-related information from TWITTER to be received and rendered in the plurality of articles 82. A subsequent selection of Source 1 may prevent information from being received from TWITTER, and previously ren- dered articles from TWITTER would be removed from the plurality of articles 82. Similarly, selecting Source 2 may allow topic-related information from GOOGLE to be received and rendered in the plurality of articles 82. A Jun. 13, 2019 subsequent selection of Source 1 may prevent information from being received from GOOGLE, and previously ren- dered articles from GOOGLE would be removed from the plurality of articles 82. And, selecting Source 3 may allow topic-related information from the user-de?ned source to be received and rendered in the plurality of articles 82. A subsequent selection of Source 3 may prevent information from being received from the user-de?ned source, and previously rendered articles from the user-de?ned source would be removed from the plurality of articles 82. [0096] In an example embodiment, ?lter of topic-related information may be adjustable via the interface 70. For example, slider 80 may function as a ?lter. Selecting and moving control 86 to the left or to the right of slider 80 may more or less ?lter topic-related information based on in?u- ence of the source and/ or relevance of the item. For example, selecting and moving the control 86 to the left end of slider 80 may allow less in?uential and/or less relevant topic- related information to be received. And selecting and mov- ing the control 86 to the right end of slider 80 may allow only more in?uential and/or more relevant topic-related information to be received. Thus, in this example scenario, the left edge of slider 80 represents less restriction and the right edge of slider 80 represent more restriction. As slider 80 is adjusted via control 86, the information rendered in the plurality of articles 82 may accordingly be adjusted. [0097] The functionality invoked via slider 80 may deter- mine a combined ranking of each item by relevance and in?uence (referred to herein as myCREOrank). In an example embodiment, the slider 80 scale may allow posi- tioning from left to right, from a position designated ?All? which returns all items to the channel to the far-right position which may be designated ?myCREOpicks?, and returns only the most relevant items from the most in?uen- tial sources. Slide 80 may be positioned at intermediate points to allow more or less relevant items from more or less in?uential sources to be shown. The myCREOrank ?ltering by in?uence and relevance may be based on an algorithm which may be applied to all items in the database which have been initially selected for inclusion in the channel, based on the search for topic-related keywords. [0098] The myCREOrank may be calculated based on in?uencer online presence metrics and/or channel context. For in?uencer online presence metrics (In?uencer), metrics may be obtained from social networks and/or various third- party services for each in?uencer, or the like. In?uencer metrics may vary depending on In?uencer typeia web site or a social network user. Example In?uencer metrics may include TWITTERinumber of posts user has made, num- ber of followers, retweets etc., of user videos, number of views each video has got, average number of views etc., Blog?number of posts, number of follower, number of reposts, links etc. Web pageiGoogle Page rank, or the like, or any appropriate combination thereof. [0099] To compare di?erent In?uencer metrics, In?uencer metrics may be weighted based on various characteristics and returned as Score (0-100). The weighting methodology may be based on academic research, other publications, information available online, or the like, or any appropriate combination thereof. The weighting methodology may be subject to adjustment based on crowd-sourced information the actions of users of the channel such as views and curation of items) and re?nement. Thus, Score may repre- sent online popularity. US [0100] Channel context (Relevance) may be indicative of highly ranked items related closely to the channel topic. A post by a highly ranked influencer may not be as relevant to a particular channel as a post from a generally less highly ranked in?uencer. Therefore, a relevance score may be determined by considering such items as the number of mentions of topic-related keywords in an item, a ratio of topic-related keywords to other words in the item, the number of mentions of topic-related keywords in the item headline, the ratio of topic-related keywords to other words in the headline, the number of views or Page rank of the item, and other factors. Relevance may consider crowd- sourced data such as views and curation by users of a particular channel on particular items or items sourced from particular sources or authors. [0101] In an example embodiment, slider 80 may ?lter items returned by a search for a topic by utilizing an algorithm based on a weighted scale which may consider the perceived reliability of the source the New York Times or Wall Street Journal would be deemed more reliable than an anonymous TWITTERTM post) and the authority of the author, which may be determined based on the number of previous items posted by the author. The slider 80 may adjust the weighting of the algorithm based on crowd- sourced data as some items are liked, not liked, curated out, recommended to be deleted, or viewed, or the like. [0102] Topic-related information may be provided via a rendering of the plurality of articles 82. An article may be rendered in any appropriate manner. In an example embodi- ment, as depicted in FIG. 6, a rendering of an article may comprise an icon and/or a description. Selection of an article may provide a feed to topic-related information form the article. For example, article 1 may provide a link to a YOUTUBE video. Selection of article 1 may result in a rendering of the YOUTUBE video. The rendering of the YOUTUBE video may be ?ltered as described above. The description of the article may comprise any appropriate description. For example, as depicted in FIG. 6, the descrip- tion article may comprise a title of the article, and a time and a description of the source. For example, in accordance with the foregoing example scenario, the description may com- prise a title of the YOUTUBE video, the time the video was obtained, and the name of the source YOUTUBE). [0103] The interface 70 may provide a mechanism for updating an intelligence channel via selection of item 84, depicted as ?Curate? in FIG. 6. Selection of item 84 may allow an intelligence channel to be updated in any appro- priate manner. For example, selection of item 84 may provide a link to a web page, or the like, wherein the intelligence channel may be updated, by providing a source, removing a source, combining source, adjusting ?ltering, adjusting ?ltering criteria, or the like, or any appropriate combination thereof. [0104] FIG. 7 illustrates an example interface for provid- ing access to a plurality of intelligence channels. As shown in FIG. 7, each intelligence channel may be represented by an icon and/or a description of the intelligence charmel. For example, intelligence channel 86 is depicted by a graphic logo and textual description ?Incubators,? indicating the topic of the intelligence channel. Selection of intelligence channel 86 may provide link to an interface that provided topic related information. [0105] FIG. 8 illustrates an example interface that pro- vides topic related information pertaining to incubators. Jun. 13, 2019 Labels 78, 80, 82, and 86 on FIG. 8 correspond to labels 78, 80, 82, and 86 of FIG. 6, respectively, to identify items that ?inction as previously described. [0106] Referring again to FIG. 1, an intelligence channel may be updated at step 22 in any appropriate manner. As previously described, an intelligence channel may be updated based on ?ltering. An intelligence channel may be updated in an iterative fashion (self-learning closed loop) wherein problems found from quality testing and user expe- rience may be fed back to improve channel development and re?ne the chaImel. This manner of updating may occur continuously, periodically, aperiodically, based on the occur- rence of an event, or any appropriate combination thereof. In an example embodiment, an intelligence channel may be updated based on a user of an intelligence channel desig- nating unwanted channels or designate channels to be com- bined or fused. [0107] Crowdsourcing may be utilized to update an intel- ligence channel. An intelligence channel and/ or information provided by an intelligence channel may be distributed and comments and/or suggestions may be received. The com- ments and/or suggests clicks on a like or dislike button, clicks to remove an article, clicks through to review the article, etc.) may be utilized to update an intelligence channel. For example, an intelligence channel and/or infor- mation provided by an intelligence channel may be provided via for example, email, social media, TWITTER, LINKE- DIN, FACEBOOK, GOOGLE, a social net- work, news articles, or the like, or any appropriate combi- nation thereof. Recipients thereof may ?like? a chaImel, article, source, or the like. A recipient thereof may request that a channel, article, source, or the like, be removed. This type of feedback may be utilized to update an intelligence channel. In an example embodiment, a database or the like, of such feedback may be generated and utilized for subse- quent updates. [0108] FIG. 9 is an example functional block diagram for developing an intelligence channel. Available sources 90 may be provided to multiple ?mctional paths, depicted in FIG. 9 as columns 91, 93, and 95. Each functional path may perform operations on a speci?c source. A speci?c source may comprise any appropriate source. For example, speci?c source 92 may represent TWITTER, GOOGLE, GOOGLE NEWS, GOOGLE BLOGS, YOUTUBE, FACEBOOK, BING, YAHOO, WIKIPEDIA, a direct feed, or the like. Note, functional components are labeled with numbers only in functional path 91 for the sake of simplicity. However, functions performed in functional path 91 may be performed in other functional paths 93, 95) in a similar manner on the respective speci?c source of the functional path. [0109] A speci?c source may be searched for keywords via a custom search at functional block 94. The custom search may be customized for the speci?c source based on the topic and/or by the user. For example, a custom search on TWITTER may comprise the URL (Uniform Resource Locator) a custom search on GOOGLE NEWS may comprise the URL http:// news.google.com, a custom search on GOOGLE BLOGS may comprise the URL a cus- tom search on YOUTUBE may comprise the URL http: a custom search on FACEBOOK may search a company or the like FACEBOOK page, or a custom search for a direct feed RSS feed) may comprise US 2019/0179861 A1 searching for topic-related information via the direct feed, company web sites, news feeds, or the like. [0110] Custom search phrases for topic related informa- tion using speci?c source search parameters may by gener- ated at function block 96. For example, TWITTER may be searched using TWITTER speci?c search parameters and semantic structure, GOOGLE NEWS may be searched using GOOGLE speci?c search parameters and semantic struc- ture, GOOGLE BLOGS may be searched using GOOGLE BLOGS speci?c search parameters and semantic structure, YOUTUBE may be searched using YOUTUBE speci?c parameters and semantic structure, or FACEBOOK may be searched using a FACEBOOK ID, user name, other FACE- BOOK appropriate speci?c parameters and semantic struc- ture, or the like. [0111] Custom test phrases may be tested at functional block 98. Custom test phrases may be tested on the respec- tive speci?c source utilizing an appropriate URL, and semantic structure for the speci?c source. For example, a custom phrase for TWITTER may be tested by using the URL a custom phrase for GOOGLE NEWS may be tested by using the URL a custom phrase for GOOGLE BLOGS may be tested by using the URL http://blogsearch. googlecom, a custom phrase for YOUTUBE may be tested by using the URL or a custom phrase for FACEBOOK may be tested by using the URL page.php?id, or the like. [0112] Search phrases may be encoded at functional block 100. Search phrase encoding may handle special characters that are not allowed in URLs, which are limited to the ASCII character set and cannot contain spaces. As temis often contain characters outside the ASCII character set or spaces, the URL may be converted to a valid ASCII format. URL encoding may replace characters with a followed by two hexadecimal digits. For example, the normal text Michelle Gunter may be encoded as Michelle%20G%C3%BCnter Encoded search phrases may be tested at on custom feed at functional block 102. And custom feeds and search phrased may be stored at functional block 104. [0113] FIG. 10 is an example block diagram ofa system 106 for generating an intelligence channel and providing information via an intelligence channel. In an example embodiment, the system 106 may comprise an application layer, 110, a database layer 112, and a web layer 114. The application layer 110 may comprise any appropriate com- bination of an updates processor 116 and a data processor 118. The database layer 112 may comprise any appropriate combination of a database 120 and a search engine 122. The web layer 114 may comprise any appropriate combination of an administration interface 124 and a user interface 126, each of which may be coupled to a network 128. [0114] In an example embodiment, the database 120 may be a mongo db, an open source program, and open sourced tools. The data processor 118 may generate the search parameters which may be implemented through the search engine, which may be GOOGLE or another commercially available search engine. The data processor handles the raw item data processing, analysis and enrichment, and interacts with the database by requesting raw items and storing enriched items to the database. The search engine delivers ?ltered keyword results and interacts with the database by Jun. 13, 2019 requesting raw data and with the user interface by returning ?ltered data. The administration interface 124 may be uti- lized to de?ne and re?ne the search parameters and key- words, includes and excludes, which operate through the search engine to generate results delivered to the database which are delivered to users through the user interface. The updates processor 118 may repeat the process periodically to update all results by scheduling source updates, requesting load balancing, requesting search preparation and execution, data aggregation, and low-level ?ltering. The database 120 may track each article retumed as associated with a search during the update process. The database 120 also may maintain all user information for purposes including pay- ment processing and history, data mining, and crowd- sourced article or source ratings and determinations. The database is where all data is stored, including all searches, items retrieved, user information, including all clicks, likes, dislikes, curations, click-throughs, click-through destina- tions, curation information, billing information and the like. All other components interact with the database by request- ing data for processing or display and sending data for storage. [0115] FIG. 11 is an example block diagram of the updates processor 116. In an example embodiment, the updates processor 116 may comprise any appropriate combination of a load balancer 130, a request manager 132, a metadata extractor 138, data ?lters 140, a data normalizer 142, a processor 144, a query boulder 134, a data aggregator 146. The updates processor schedules source updates and requests load balancing, request preparation and execution, data aggregation, and low-level ?ltering. The updates pro- cessor interacts with the database by requesting data about channels and sources and storing raw items in the database. The query builder 134 and the data aggregator may be in communications with sources 136, which a may be in communications with a network 148. [0116] The system may be designed to be horizontally scalable. The database and search engine(s) may be scaled as demand requires by adding additional servers (nodes). The updates processor and data processor may also be scaled horizontally by adding additional servers (nodes), provided that each server (node) acts as an individual instance of the updates processor and data processor. Load balancing is provided on a system database level. [0117] In an example embodiment, the load balancer 130 may optimize usage of hardware and communications resources to minimize processing time and allow multiple searches to proceed simultaneously to reduce response time. The request manager 132 may interact with the load bal- ancer 130 so requests from sources or to search engines are e?iciently managed. The metadata extractor 138 may iden- tify the appropriate metadata for identifying and indexing each article within the database and may identify API versions and settings associated with particular sources. Data ?lters 140 may be used to remove irrelevant items identi?ed through exclude statements, as including profanity or as duplicates. The query builder 134 may comprise a listing of the de?ning sources, keywords, includes, excludes, and ?lters for an intelligence channel. The data aggregator 146 may assemble data for delivery to the user through the user interface and ?lters duplicates. [0118] FIG. 12 is an example block diagram of the data processor 118. In an example embodiment, the data proces- sor 118 may comprise any appropriate combination of a load US balancer 150, a processor 152, a miscellaneous helper 154, a ranking algorithm processor 156, an in?uencer ranking processor 158, a content ranking processor 160, a content enrichment processor 162, a content extraction processor 164, an images, video, etc. processor 166, a content analysis processor 168, a language detection processor 170, and a nature language processing (NLP) processor 172. The load balancer may monitor available request limits and schedul- ing requests. The ranking algorithm processor may comprise a system structural unit including ?CREOrank?, which may calculate the in?uence of a source based on prior instances of the source items and may be crowd-sourced, and content ranking, which may calculate item content relevance based on characteristics such as headline length, ration of key- words/stopwords to all words in item, occurrence of #hashtags and @usernames in the headline or item, content length, occurrence of keywords/stopwords in content, and the like. A content enrichment processor is a system struc- tural unit which may extract content from text to provide structured data and extracts media elements (images, video, sound, etc.) from text. The content analysis module is a system structural element that includes language detection, which may include or exclude an item based on its language (French, German, English, Mandarin, Hebrew etc.) and natural language processing, which may detect and highlight certain types of data elements such as names of individuals, names of entities monetary values, times or dates or the like. [0119] FIG. 13 is a block diagram of an example apparatus 180 that may be utilized to implement and/or facilitate an intelligence channel as described herein. The apparatus 180 may comprise hardware or a combination of hardware and software. The apparatus 180 depicted in FIG. 13 may represent any appropriate apparatus, device, processor, server, a gateway, a node, a database, or the like, or any appropriate combination thereof. For example, the apparatus 180 may comprise an apparatus, a device, a processor, a server, a gateway, a node, a database, the updates processor 116, the data processor 118, the database 120, the search engine 122, the administration interface 124, the user inter- face 126, each of which may be coupled to a network 128, the load balancer 130, the request manager 132, the meta- data extractor 138, the data ?lters 140, the data nonnalizer 142, the processor 144, the query boulder 134, the data aggregator 146, or the like, or any appropriate combination thereof. It is emphasized that the block diagram depicted in FIG. 13 is exemplary and not intended to imply a speci?c implementation or con?guration. Thus, the apparatus 180 may be implemented in a single apparatus or multiple apparatuses single server or multiple servers, single gateway or multiple gateways, single apparatus or multiple apparatuses, single node or multiple nodes, single processor or multiple processors, single database or multiple data- bases, single device or multiple devices, etc.) Multiple apparatuses may be distributed or centrally located. Multiple apparatuses may communicate wirelessly, via hard wire, or any appropriate combination thereof. [0120] In an example embodiment, apparatus 180 may comprise a processor and memory coupled to the processor. The memory may comprise executable instructions that when executed by the processor cause the processor to effectuate operations associated with an intelligence channel as described herein. As evident from the herein description apparatus 180 is not to be construed as software per se. Jun. 13, 2019 [0121] In an example con?guration, apparatus 180 may comprise a processing portion 182, a memory portion 184, and an input/output portion 186. The processing portion 182, memory portion 184, and input/output portion 186 may be coupled together (coupling not shown in FIG. 13) to allow communications therebetween. Each portion of the appara- tus 180 may comprise circuitry for performing functions associated with an intelligence as described herein. Thus, each portion may comprise hardware, or a combination of hardware and software. Accordingly, each portion of the apparatus 180 is not to be construed as software per se. That is, processing portion 182 is not to be construed as software per se. Memory portion 184 is not to be construed as software per se. Input/output portion 186 is not to be construed as software per se. Volatile memory portion 188 is not to be construed as software per se. Non-volatile memory portion 190 is not to be construed as software per se. Removal storage portion 192 is not to be construed as software per se. Non-removal storage portion 194 is not to be construed as software per se. Input device(s) portion 196 is not to be construed as software per se. Input device(s) portion 198 is not to be construed as software per se. And communication connection(s) portion 200 is not to be con- strued as software per se. Each portion of apparatus 180 may comprise any appropriate con?guration of hardware and software as would be ascertainable by those of skill in the art to perform respective functions of an intelligence channel. [0122] The input/output portion 186 may be capable of receiving and/or providing information from/to a commu- nications device and/or other apparatuses con?gured to generate and/or utilize an intelligence channel as described herein. For example, the input/output portion 186 may include a wireless communications card. The input/output portion 186 may be capable of receiving and/or sending video information, audio informa- tion, control information, image information, data, or any combination thereof. In an example embodiment, the input/ output portion 186 may be capable of receiving and/or sending information to determine a location of the apparatus 180 and/or a communications device. In an example con- ?guration, the input\output portion 186 may comprise a GPS receiver. In an example con?guration, the apparatus 180 may determine its own geographical location and/or the geographical location of a communications device through any type of location determination system including, for example, the Global Positioning System (GPS), assisted GPS (A-GPS), time difference of arrival calculations, con- ?gured constant location (in the case of non-moving devices), any combination thereof, or any other appropriate means. In various con?gurations, the input/output portion 186 may receive and/or provide information via any appro- priate means, such as, for example, optical means infrared), electromagnetic means RF, WI-FI, BLU- ETOOTH, ZIGBEE, etc.), acoustic means speaker, microphone, ultrasonic receiver, ultrasonic transmitter), or a combination thereof. In an example con?guration, the input/ output portion may comprise a WIFI ?nder, a two-way GPS chipset or equivalent, or the like, or a combination thereof. [0123] The processing portion 182 may be capable of performing functions associated with an intelligence chan- nel as described herein. In an example embodiment, the processing portion 182 may be capable of, in conjunction with any other portion of the apparatus 180, installing an application for an intelligence channel as described herein. US 2019/0179861 A1 [0124] In a basic con?guration, the apparatus 180 may include at least one memory portion 184. The memory portion 184 may comprise a storage medium having a concrete, tangible, physical structure. Thus, the memory portion 184, as well as any computer-readable storage medium described herein, is not to be construed as a transient signal. The memory portion 184, as well as any computer-readable storage medium described herein, is not to be construed as a propagating signal. The memory portion 184, as well as any computer-readable storage medium described herein, is to be construed as an article of manu- facture. The memory portion 184 may store any information utilized in conjunction with an intelligence charmel as described herein. Depending upon the exact con?guration and type of processor, the memory portion 184 may be volatile 188 (such as some types of RAM), non-volatile 190 (such as ROM, ?ash memory, etc.), or a combination thereof. The apparatus 180 may include additional storage removable storage 192 and/or non-removable storage 194) such as, for example, tape, ?ash memory, smart cards, CD-ROM, digital versatile disks (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, universal serial bus (USB) compatible memory, or any other medium which can be used to store information and which can be accessed by the apparatus 180. [0125] The apparatus 180 also may contain communica- tions connection(s) 200 that allow the apparatus 180 to communicate with other apparatuses, devices, network enti- ties, or the like. A communications connection(s) may comprise communication media. Communication media may typically embody computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mecha- nism and includes any information delivery media. By way of example, and not limitation, communication media may include wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infra- red, and other wireless media. The term computer readable media as used herein includes both storage media and communication media. The apparatus 180 also may include input device(s) 196 such as keyboard, mouse, pen, voice input device, touch input device, etc. Output device(s) 198 such as a display, speakers, printer, etc. also may be included. [0126] FIG. 14 is a block diagram of an example device 220 that may be utilized to generate and/or implement an intelligence channel as described herein. The device 220 may comprise and/or be incorporated into any appropriate device, examples of which may include, a mobile device, a mobile communications device, a cellular phone, a portable computing device, such as a laptop, a personal digital assistant a portable phone a cell phone or the like, a smart phone, a video phone), a portable email device, a portable gaming device, a TV, a DVD player, portable media player, a portable music player, such as an MP3 player, a Walkman, etc.), a portable navigation device GPS compatible device, A-GPS compatible device, etc.), or a combination thereof. The device 220 can include devices that are not typically thought of as portable, such as, for example, a public computing device, a navigation device installed in-vehicle, a set top box, or the like. The mobile device 220 can include non-conventional computing devices, such as, for example, a kitchen appliance, a motor Jun. 13, 2019 vehicle control steering wheel), etc., or the like. As evident from the herein description device 220 is not to be construed as software per se. [0127] The device 220 may include any appropriate device, mechanism, software, and/or hardware for facilitat- ing and/ or implementing an intelligence channel as described herein. In an example embodiment, the ability to generate and/or implement an intelligence channel is a feature of the device 220 that may be turned on and off. Thus, in an example embodiment, an owner and/or user of the device 220 may opt-in or opt-out of this capability. [0128] In an example embodiment, the device 220 may comprise a processor and memory coupled to the processor. The memory may comprise executable instructions that when executed by the processor cause the processor to effectuate operations associated with an intelligence channel as described herein. [0129] In an example con?guration, the device 220 may comprise a processing portion 222, a memory portion 224, an input/output portion 226, and a user interface (U1) portion 228. Each portion of the device 220 may comprise circuitry for performing functions associated with each respective portion. Thus, each portion may comprise hardware, or a combination of hardware and software. Accordingly, each portion of the device 220 is not to be construed as software per se. That is, processing portion 222 is not to be construed as software per se. Memory portion 224 is not to be construed as software per se. Input/ output portion 226 is not to be construed as software per se. And user interface portion 228 is not to be construed as software per se. Each portion of device 220 may comprise any appropriate con?guration of hardware and software as would be ascertainable by those of skill in the art to perform respective functions of an intelligence channel as described herein. It is emphasized that the block diagram depiction of device 220 is exemplary and not intended to imply a speci?c implementation and/or con?guration. For example, in an example con?guration, the device 220 may comprise a cellular communications tech- nology and the processing portion 222 and/or the memory portion 224 may be implemented, in part or in total, on a subscriber identity module (SIM) of the device 220. In another example con?guration, the device 220 may com- prise a laptop computer and/or tablet device (laptop/tablet). The laptop/tablet may include a SIM, and various portions of the processing portion 222 and/or the memory portion 224 may be implemented on the SIM, on the laptop/tablet other than the SIM, or any combination thereof. [0130] The processing portion 222, memory portion 224, and input/output portion 226 may be coupled together to allow communications therebetween. In various embodi- ments, the input/ output portion 226 may comprise a receiver of the device 220, a transmitter of the device 220, or a combination thereof. The input/output portion 226 may be capable of receiving and/or providing information pertain- ing to an intelligence channel as described herein. In various con?gurations, the input/output portion 226 may receive and/or provide information via any appropriate means, such as, for example, optical means infrared), electromag- netic means RF, WI-FI, BLUETOOTH, ZIGBEE, etc.), acoustic means speaker, microphone, ultrasonic receiver, ultrasonic transmitter), or any appropriate combi- nation thereof. [0131] The processing portion 222 may be capable of performing functions pertaining to an intelligence channel as US described herein. In a basic con?guration, the device 220 may include at least one memory portion 224. The memory portion 224 may comprise a storage medium having a concrete, tangible, physical structure. Thus, the memory portion 224, as well as any computer-readable storage medium described herein, is not to be construed as a transient signal. Further, the memory portion 224, as well as any computer-readable storage medium described herein, is not to be construed as a propagating signal. The memory portion 224, as well as any computer-readable storage medium described herein, is to be construed as an article of manufacture. The memory portion 224 may store any infor- mation utilized in conjunction with an intelligence channel as described herein. Depending upon the exact con?guration and type of processor, the memory portion 224 may be volatile (such as some types of RAM), non-volatile (such as ROM, ?ash memory, etc.), or a combination thereof. The mobile device 220 may include additional storage removable storage and/or non-removable storage) such as, for example, tape, ?ash memory, smart cards, CD-ROM, digital versatile disks (DVD) or other optical storage, mag- netic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, universal serial bus (USB) com- patible memory, or any other medium which can be used to store information and which can be accessed by the mobile device 220. [0132] The device 220 also may contain a user interface (U1) portion 228 allowing a user to communicate with the device 220. The U1 portion 228 may be capable of rendering any information utilized in conjunction with an intelligence channel as described herein. The UI portion 228 may provide the ability to control the device 220, via, for example, buttons, soft keys, voice actuated controls, a touch screen, movement of the mobile device 220, visual cues moving a hand in front of a camera on the mobile device 220), or the like. The U1 portion 228 may provide visual information via a display), audio information via speaker), mechanically via a vibrating mechanism), or a combination thereof. In various con?gu- rations, the UI portion 228 may comprise a display, a touch screen, a keyboard, an accelerometer, a motion detector, a speaker, a microphone, a camera, a tilt sensor, or any combination thereof. The U1 portion 228 may comprise means for inputting biometric information, such as, for example, ?ngerprint information, retinal information, voice information, and/or facial characteristic information. [0133] The U1 portion 228 may include a display for displaying multimedia such as, for example, application graphical user interfaces (GU1s), text, images, video, tele- phony functions such as Caller ID data, setup functions, menus, music, metadata, messages, wallpaper, graphics, Internet content, device status, preferences settings, map and location data, routes and other directions, points of interest (P01), and the like. [0134] In some embodiments, the U1 portion may com- prise a user interface (UI) application. The U1 application may interface with a client or operating system (OS) to, for example, facilitate user interaction with device functionality and data. The U1 application may aid a user in entering message content, viewing received messages, answering/ initiating calls, entering/deleting data, entering and setting user IDs and passwords, con?guring settings, manipulating content and/or settings, interacting with other applications, Jun. 13, 2019 or the like, and may aid the user in inputting selections associated with an intelligence channel as described herein. [0135] In accordance with the herein description of intel- ligence chaimels, an intelligence chaimel may be utilized in various applications. In various example embodiments, an intelligence channel may comprise compilations or fusions of related topics into a single channel. The intelligence channel may comprise channels for a topic as a single ?thought? channel encompassing a variety of opinions and ideas. An intelligence channel may comprise a ?pack? of individual intelligence channels directed to topics in selected market segments, regions, and other areas of interest. The mechanisms for interfacing with intelligence channels may provide a user to search within an intelligence channel for subtopics of interest, and to generate intelligence channels directed to the subtopics. A user may de?ne a private intelligence channel that is available to only designated recipients. Locations, addresses, or the like, from informa- tion provided via an intelligence channel may be automati- cally inserted into a map GOOGLE MAPS, geocoding utilities, etc.). Automatic translation of information provided via an intelligence channel may be accomplished. A word cloud display may be generated from information provided via an intelligence channel for use in a presentation, display, posting, or the like. In an example embodiment, a word cloud display is a visualization of the frequency of use of the words in a document, in which the size of the font for each word is related to its frequency or use. Words may be scrambled and may be shown as in a cloud. [0136] In various example embodiments, use of an intel- ligence chaimel may be monitored to observe user behavior to predict events that may be of interest to a user, to provide alerts that may be applicable to a user, to provide informa- tion of interest to a user, to generate a custom intelligence channel, to combine intelligence channels, or any appropri- ate combination thereof. A data feed may be provided RSS, or other data feeds) comprising user requested intel- ligence channels for use on websites and displays, and potentially customized to employ user graphic formats, and user system feed requirements. Information provided via an intelligence channel may be viewed via any appropriate mechanism, such as, for example, o?line viewing, standard readers, GOOGLE READER, KINDLE, or the like. Containing Disinfomiation Spread Using Customizable Intelligence Channels [0137] As will be explained below, techniques are pro- vided for rating the veracity of content distributed via digital communications sources by creating an ontology and select- ing keywords for a topic of the content and creating cus- tomizable intelligence channels of the type described above to identify and extract from the customizable intelligence channel a ?rst list of potential experts on the topic sorted by at least relevance and in?uence. The list of experts may be supplemented by mining trusted media sources to extract a second list of potential experts or witnesses on the topic. The ?rst and second lists of potential experts are evaluated as a ?inction of at least one of professionalism, reliability, prox- imity, experience, responsiveness, and lack of self-interest in the topic to identify a short list of experts. The content is provided to the short list of experts for evaluation. The experts are then polled about the veracity of the content to create a veracity score for delivery with the content. By evaluating and delivering the content with the US veracity score as well as ?fake news? warnings and metrics relating to the content as appropriate, the spread of disin- formation may be contained. [0138] Several features of sample embodiments are described below in connection with features illustrated in FIG. 15. It will be appreciated that each of these features forms a part of a system and process for containing the spread of disinformation in the sample embodiments. Feature 1: Providing an Interactive Interface that Aggregates and Summarizes News Content Veracity Signals into a Creos Core with Possible Fake News Warnings and Rebut- tals [0139] A hyperlink to a custom-built web page called (see, for example, FIG. 16) is created that includes the following: [0140] 1. An introduction summarizing bene?ts; [0141] 2. The initial headline ?Robot was killed by a self-driving Tesla cm?%ES Jan. 6, 2019 with hyperlink) from the breaking news story with photo or video (including possibly deepfake); [0142] 3. A color-coded warning (red: proceed with caution, orange: undecided, green: OK, with accom- modations for color blind persons by, for example, including some blue in the green light); however, a 3-color sign may be replaced by a large warning if/when enough experts opine 80% of the experts feel there is an 80% or more chance that the presented new is fake news); [0143] 4. A new measure such as ?Unlikely fake based on a CREOscore of 86 on a scale of 1-100? (the actual wording may be tested by experts); [0144] 5. The corresponding CREOscore color coded chart (being updated in near real time); [0145] 6. Background about the originating media, including for example whether it is a site that comes with a clear satire disclaimer; [0146] 7. Learn how: Hyperlinks are provided to fre- quently asked questions including the methodology applied; [0147] 8. Selected rebuttals from a few credible experts debunking the falsehood, or testimonials from wit- nesses on the scene or others who con?rm the claim; [0148] 9. An intelligence channel with unique interface/ toggle ?separating the wheat from the chaff? where rather than a slider from myCREOpicks to All Sources, a user could conveniently switch between 4 positions to observe various modes of news reports. ?The good?, ?the friendly?, ?the bad?, and ?the ugly? would be isolated in different news feeds derived from: [0149] a) Reliable experts who provided original quality content; [0150] b) My friends (where the user could toggle to enable duplicates to show the ?lter bubble); [0151] 0) Generally reliable sources who were fooled by the fake news; and [0152] d) Known fake news sources and other bad actors extremist sites, Russian bots, Macedo- nian teenagers, existence of long dormant accounts suddenly activated, accounts posting similar content and/or using same hashtags). [0153] 10. An icon could also be used to graphically show how progressively fake the news stories are. [0154] As soon as the CREOpoint system is activated it displays information disclosing when the results will be Jun. 13, 2019 ready show an image that 14 out of 30 minutes of wait time has elapsed until results will be reported) and the percentage of experts who have been asked or weighed in with a spinning Mac type wheel of death to show experts are at work). [0155] To encourage persons who know something to say something, crowdsourced input is provided from users that would have to provide their LinkedIn pro?le as initial validation click to a simple form to be ?lled out or access con?dential communications channel). [0156] Preliminary information about the source of the news story also may be provided from credible press asso- ciation directories. If no Wikipedia page is available for the topic of the news story, this could be useful as a red ?ag. Depending on the topic vaccine hesitancy, one of the top 10 threats to global health per the World Health Orga- nization), especially if relevant to possibly ?real world harm,? a link to a corresponding Britannica entry may provide initial insights to users (including for example fraudulent claims that have been falsely made about the supposed dangers of the HPV cervical cancer or MMR measles, mumps and rubella vaccines). Other links to the source the news source Facebook page and Twitter handle) are added so that the user can also access and rate the quality of their other content. [0157] A temporary warning ?Rated false by fact checking organizations? with top worst fake news stories about that industry sector may be pre-populated into the interface. For example, assuming the CREOpoint system was activated ?irther to the ?news? breaking Sep. 5, 2018 that ?Michael Jordan Resigns From The Board At Nike?Takes ?Air Jordan?s With Him,? while predetermined experts are con- tacted, the interface displays proven fake news. Such articles would have included ?Nike CEO Resigns After Massive Kaepemick Blunder,? ?Nike Stock Drops 260 Points In A Single Day Thanks To Colin Kaepemick,? ?Florida Man Accidentally Burns Home Down After Lighting Nike Shoes On Fire In Protest Of Nike?s Colin Kaepemick Ad,? ?Fed- eral Government Cancels $80 Million Nike Contract,? ?Nike Fires Colin Kaepemick After Arrest.? An icon could be selected to graphically show how quickly these fake news stories spread. [0158] From curation of the CREOpoint Intelligence Channel, the CREOpoint system curates top real news about the topic or client Nike) such as the following story from Forbes ?Continued success of Nike?s Jordan Brand has pushed Michael Jordan?s net worth to As soon as available, the CREOpoint system would also post rebuttals from such sources as the Associated Press It is expected the CREOpoint system activated immediately would have created an immediate warning and decreased the close to one million Facebook engagements initiated by a known online satire site about Michael Jordan parting ways with Nike. [0159] From a link making a possible fake statement, the user may scroll down or open anew page with corresponding data and the following intro: [0160] [0161] 1. Known fake news sources and bad actors or others with similar characteristics, remote from the subject matter, unknown in ?eld, etc.; [0162] 2. Your friends/?lter bubble; So, ?who ya gonna trust?? US [0163] 3. Generally reliable sources who have been fooled by fake news; and [0164] 4. Independent experts with high reliability scores. [0165] This is followed by a customized intelligence chan- nel with unique interface/ 4 position toggle rather than the slider from myCREOpicks to All Sources (see, for example, FIG. 17). [0166] To better inform users by giving them more context on the information they see, users are provided additional information, by sharing more details on the articles and quality of the sources with various levels of credibility. [0167] A user could conveniently switch among the fol- lowing 4 positions to observe various concurrent ?realities? from: [0168] 1. The ugly [0169] 2. The bad [0170] 3. The friendly [0171] 4. The good Case Study: Autonomous Vehicles Safety and Trust [0172] On Jan. 7, 2019, a group of automated vehicle developers, suppliers, and advocacy groups announced the Partnership for Automated Vehicle Education (PAVE), a new coalition for public education on automated vehicles as follows: ?Media interest is picking up, public attention is dialing in, and people are, understandably, a bit con?ised . . . . There?s a lot misinformation ?oating around, and it?s on all of us, especially this group here, to help correct that,? said PAVE member Kyle Vogt, chief technology of?cer at the autonomous vehicle developer Cruise. The CREOpoint system described herein makes it possible to access experts to take action to debunk misinformation. For example, assuming PAVE, GM, Allianz (who insures the autonomous industry), or Tesla were clients, the CREOpoint system described herein would be activated to prevent the fake news from spreading. A headline published during CES on Jan. 6, 2019, was actually a PR stunt, and it had an impact on trusted companies. From such as statement, the following categories of news stories followed: [0173] 1. The fake/ugly [0174] In fact, the CREOpoint system would have detected in near real-time (after noise ?ltering described above) that the breaking ?news? was a Russian disinfomia- tion/PR stunt from a robot company on Jan. 6, 2019. mobot-v-las-vegase/ [0175] 2. The bad [017 6] The Verge was not fooled by the story: 1082367428330434560), but many other news outlets pub- lished this as news on Jan. 7, 2019, including: [0177] The Daily Mail (noting the possible PR stunt but still publishing details) the next day on Jan. 7, 2019: [0178] The Washington Times (with the wrong picture of a destroyed Tesla): [0179] TheDaily Show on Jan. 8, 2019: Jun. 13, 2019 [0180] The Consumer Editor of USA Today shared the press release as is: 1082380540584517632 [0181] The Las Vegas Post (Las Vegas news you can trust!): [0182] AutoWeek: [0183] ?The AV and robot wars of 2019 have begun? [0184] 3. The friendly [0185] Ideally, the face of the friend, group, or page who initially shared the post is provided noting that ?friends? shared way beyond ?the end of times is upon us?: BOX ?keywords:CES%20Tesla%20R obot&origin:SWITCH_ [0186] 4. The good (unfortunately Jan. 8, 2019, 2 days after the press release): [0187] Derek Kessler, Managing Editor at Mobile Nations: vegas.139856/ [0188] Wired: [0189] Electrek: crash-robot-pr?stunt-media [0190] Polygraph: [0191] Electrive: 1085467176042221568 Feature 2: Pre-Determining Experts who Could Best Crowd Source Veracity Signals [0192] The ?rst step is to aggregate a long list of prospec- tive experts based on their recognition and skills. The objective is to constantly improve the quality of the veracity signals by including the CREOpoint Expert Trust Ratings (CREOExpert Rating), which will only be as effective as the experts selected (garbage in, garbage out), the level of their engagement, and how exhaustive and diversi?ed the list is. [0193] Experts are selected for a set of high potential crisis situations. A number of relevant use cases are preselected autonomous vehicle crashes or other fatality from an accident, ?nancial scandal, sexual harassment by executives, data breaches, etc.). [0194] Next, an ontology and most used keywords are built for each set of crisis situations. For example, in the case of autonomous vehicle ?crashes? at CES leading to fake news that implies AI will kill us all, see: -crash- vegas-account driving-education [0195] The CREOpoint system extracts from the most relevant articles words such as (?self? driving OR autono- mous) AND (Tesla OR Uber OR Cruise OR Ford OR US 2019/0179861Al NAVYA OR Car etc.) AND (crash OR hit OR struck OR victim OR injured OR killed OR fatality OR Accident). [0196] Sources of experts for the selected topic are mined by creating a CREOpoint channel for each possible crisis situation and extracting corresponding experts. Correspond- ing CREOpoint customizable real-time intelligence chan- nels and natural language processing (NLP) processes are used to identify authors of traditional and social media news items sorted by relevance, influence and when they last covered this topic, and experts quoted in articles from the authors of these most relevant and in?uent niche sources Ars Technica). Top media sources generally accepted as quoting real experts (rather than opinionated TV news show commentators) are also leveraged. The CREOpoint system starts by mining trusted media sources such as AFP, AP, Axios, BBC, Bloomberg, CBS News, CIO, Christian Science Monitor, Le Monde, The Economist, The Guardian, Information Week, The New York Times, The Wall Street Journal (US, Europe, and Asia), The Washington Post, NPR, Politico, Propublica, and Reuters (not Reuters Plus spon- sored content). Other sources could be added based on the reliability of experts they quote. The objective is to identify the relevant experts the top media sources quote the most. For example, using natural language processing (NLP), the CREOpoint system would have pre-identi?ed experts quoted in top publications during previously reported seri- ous crashes. Examples include: Tesla: [0197] Smith, a University of South Carolina law profes- sor who studies self-driving cars and is quoted in: 8-03-31/tesla-says-driv- [0198] National Transportation Safety Board agency spokesman Chris O?Neil 43617752); Uber: [0199] Anthony Foxx, who served as US Secretary of Transportation under former President Barack Obama quoted in [0200] Carla Bailo, president of the Center for Automotive Research (also in and Michael Bennett, an associate research professor at Arizona State University (quoted in 03/ [0201] To extract names, organizations and quotes from articles, the following tools could be used: [0202] The CREOpoint system also mines other search- able media sources to extract expert information from such sources as: [0203] 1. Google Scholar [0204] 2. Patent databases [0205] 3. Professional social networks such as LinkedIn and Twitter [0206] 4. Relevant stock [0207] 5. Expert networks like the GLG Group [0208] 6. References in Wikipedia [0209] 7. Awards like Nobel Prize laureates Jun. 13, 2019 [0210] For example Google Scholar will help identify the most cited Professors such as: [0211] Starting from the top, the ?rst search item identi?es the author/expert Alexander Herd, and links to 5 people citing the article. The CREOpoint system is set up to continue the discovery process by, for example, identifying one of the experts citing Alexander Herd as Sarah Fox in Georgetownedu ?Planning for density in a driverless world? by 20177HeinOnline. This paper was cited by 8 people, so she would also be identi?ed as an expert and exported to a database. [0212] Other sources mentioned above are searchable such as Microsoft Linkedln and Sales Navigator, which offer ?ltering by keyword, profession, industry, location These sources enable credentialization of experts and their networks of influence. [0213] Clients or their service providers (such as public relations, government relations, or technology consulting ?rm) also could provide a list of known experts within their networks to expand the expert population. They would not have to meet the short-listing screening criteria (de?ned below) to be polled, except lack of self-interest. [0214] The CREOpoint system may supplement the results with fact checkers and experts in ?ghting fake content ?news,? fake ?fact-checks? discrediting authentic news stories, doctored photos and deepfake videos among other fakes). A database of the most frequently quoted experts on key topics including, for example, cyber- security and deepfakes (arti?cial intelligence has allowed malicious actors to make convincingly fake videos of key politicians, distorting the truth and raising issues of digital impersonation spreading online). With changing expecta- tions of the public, and the potential for legislation and greater regulation of cybersecurity and weaponized disin- formation on social media, deepfakes may need to be actively monitored and taken down by social networks even faster than fake news. CREOpoint envisions quickly polling deepfakes experts so they can provide a quick signal and an ability to ?ag these videos in areas beyond cyberdefense. Imagine, for example, the stakes given a fake video of a limping New England Patriots quarterback minutes before the close of betting on the Super Bowl, or pornographic videos that realistically superimpose the face of his celebrity model wife when the spotlight is on the red carpet events at the Oscars or the Cannes Film Festival. [0215] The results may also be supplemented with a list of AI trained subject matter expert machines. In addition to human expertise, it is anticipated that machine learning will help AI experts do well in some subject matter areas, so the CREOpoint system also identi?es and selects such situations where there is su?icient training data to provide useful responses. It is possible that some of these algorithms would be black boxes with no way to check for bias, so these expert machines would skip the short-listing ?lters below, but their CREOExpert Trust ratings would be computed to see how they are doing over time. [0216] The resulting list of experts is exported to a data- base with ?elds including, for example, ?rst name, last name, organization, title, profession, industry, location, key- words, relevance of times they author on topic), in?u- ence, times quoted in last year, last quoted, social media US metrics, Rating (see below), topics, Linkedln skills, universities, degrees, associations, corporate a?ilia- tions, noted possible con?icts, tags with crisis situations where their expertise is relevant and the source of the name quote in Bloomberg, skill in Linkedln, client). Regard- ing their CREOpoint Reliability score, sources who reported a news story that was ultimately found to be fake, or merely a public relations or stock manipulation stunt, would see their CREOpoint Reliability score penalized, and be red ?agged as Spreaders of fake news. Note that focusing on only rating sources with a nutrition or other public labels is not enough and could be misleading and oifer a false sense of security. For example, when Bloomberg News unknow- ingly published a fake press release in November, 2016 about a ?nancial scandal allegedly a?ecting Vinci, the market capitalization of the $40 Bn European construction company declined by 18%. Reputable French news agency AFP also apologized for a mistake in reporting the death of Martin Bouygues, the CEO of another large multinational. The accuracy of this rating would improve based on vali- dated facts over a signi?cant time (possibly with uncertainty percentage). [0217] The CREOpoint system described herein also includes a rating mechanism to rate and rank experts based on how they are trusted. The [0218] CREOExpert Trust Rating is a function of Profes- sionalism, Reliability, Proximity, Experience, Responsive- ness and Lack of Self Interest, where: [0219] 1. Professionalism =qualities that characterize an expert related to his or her mastery of specialized knowledge and competence, and deep personal com- mitment to develop and improve their skills and cre- dentials; [0220] 2. Reliability:degree of consistency of the expert? prediction versus what turns out to be the truth; [0221] 3. Proximity:closeness in space, time, or rela- tionship; [0222] 4. Experience?the process of getting knowledge or skill from practical contact with and observation of subject matter facts or events; [0223] 5. Responsiveness:how quickly an expert reacts to question(s) posed by the CREOpoint system; and [0224] 6. Lack of self-interest:focus on other people and problem at hand, rather than primarily on his or her own interests. [0225] The weights assigned to each factor are adjusted over time based on the use case and CREOpoint system performance. For example, some experts with self-interest have the ability to make objective judgements. At times experience and responsiveness could negatively correlate. Also, a reporter with 20 years of experience covering automobiles who responds does not make him or her an expert in breaking fake news about how a discovery in new battery science a?ects Tesla. [0226] The CREOpoint system described herein improves the rating instruments and weight for each of these screening criteria over time since poorly designed rating instruments may be useless or actually do more harm than good (see A. R. adad and A. Gagliardi, ?Rating health information on the Internet: Navigating to knowledge or to Babel??, JAMA, 279 (1998), pp. Jun. 13, 2019 Professionalism Filter [0227] For an extra layer of professionalism to increase its e?ectiveness and completeness, the CREOpoint system mines resumes and other online sources to add further ?elds to the expert database. These include: [0228] a. Top awards Nobel prize laureate) [0229] b. Patent o?ices of patents granted or applied for) [0230] 0. Top universities MIT, Penn State or SUPAERO) [0231] (1. Top associations IPCC, USGBC) [0232] e. Research papers and citations in top respected peer-reviewed scienti?c publications in the ?eld see but not predatory or hijacked ones) [0233] f. Top think tanks [0234] g. Board memberships [0235] h. Social media recognitions Linkedln in?uencer, veri?ed Twitter handle) [0236] i. For journalists one might look at how well they meet existing or future standards Reliability Filter [0237] The CREOpoint system starts with sources whose predictions have been consistently accurate. Rather than search online only for examples where prospective experts were clearly wrong, for real time due diligence the CREO- point system creates and leverages a customizable intelli- gence channel about reliability. The search query for this Intelligence Channel is perfected over time to remove false positives (such as the expert himself or herself using these words) and include key words indicating negative qualities such as ?wrong, fake, lie, fail, failed, negative, screwed up, over con?dent, biased, misleading, unfounded, incorrect, mistaken, careless, inept, erroneous, inaccurate, invalid, untrue, fallacious, false, mistaken, in error, incorrect, falla- cious, untrue, unsound, prejudiced, erratic, deceiving, ungrounded, off target? etc. The CREOpoint system then searches for expert name matches that would ?ag an expert in a long list database whom may have been reported wrong or have been associated with signi?cant controversy. These names would be subjected to further human review and curation, with contentious experts not making the list. [0238] Individual experts on the long list are asked to vote on the probability of a chosen news story being fake. A proprietary dynamic reliability score is then generated based on how their predictions fared versus the consensus of experts and the reality. Proximity Filter [0239] Proximity is de?ned as closeness in space, time, or relationship. Proximity information is collected and used as a factor in evaluating the extraction of the experts: [0240] 1. Home location (extracted from Linkedln for example). [0241] 2. Location at the time of the crisis incident (extracted from Twitter if available depending on loca- tion privacy setting, otherwise possibly from early reporting just after the crisis breaks). For example, one can assume a UK-based soccer referee or security manager working at the 2022 FIFA World Cup in Qatar during an incident may be more likely to be credible and available than someone with the same profession- US alism but located on the other side of the world and with less access to witnesses. [0242] 3. The time zone is also noted in the database since availability to respond quickly to a crisis/poll is key. [0243] 4. Friends and family relationships some- one closer to a VIP like Michael Jordan or Martin Bouyguesis more likely to know how fake the rumor about that person is). [0244] Referring to the previous example, one can assume an expert with a home base in Silicon Valley and attending CES during the fake Tesla crash may be more likely to be credible and available than someone with the same profes- sionalism but located on the other side of the world and with less access to witnesses. Experience Filter [0245] Experience is the process of acquiring knowledge or skill from practical contact with and observation of subject matter facts or events. This factor is less important and will have a lower weight in the weighting formula. As this information may also be of interest to people question- ing the results, the CREOpoint system nevertheless extracts the number of years of experience since last year of univer- sity direct from LinkedIn), whether the person is being quoted in the press or oumals, directorships in associations, and the like. The CREOpoint system may also take into account the time this expert has been in the database (account age), assuming that an expert who has been con- sulted a number of times would have seen its reliability score become a better indicator. Responsiveness [0246] The expert database further includes whether the expert indicates they are prepared to respond quickly if and when polled as well as how quickly the expert actually reacts to a polling question(s) after they receive it, or at least the promptness of their response to the initial polling question. The CREOpoint system attempts to take into account where they are when the question is sent (they might be sleeping or in a plane). Tools that could be used with expert permission include Twitter location, Snap?s Zen.ly and possibly Internet of Things (IOT) health sensors detecting sleep. Avoiding Con?icts of Interest [0247] Con?icts of interest are minimized by short listing experts with high integrity who would focus on other people and the problem at hand, rather than primarily on their own interests. For example, note the backlash when Elon Musk, concerned and frustrated with sensational negative reviews driving market value down, proposed PRAVDA, his own ?Yelp for Journalists.? [0248] Also, rather than search online once for examples where prospective experts were clearly self-interested, for real time due diligence, the CREOpoint system creates and leverages a customizable intelligence channel about inde- pendence. The search query for this Intelligence Charmel is perfected over time, and, for the expert population, indi- vidual names are matched to such key words as ?activism, activist, aligned, accused, bias, biased, ideologies, con?ict, con?icted, censured, guilty, delinquent, felonious, dishonest, dishonorable, corrupt, violation, infringement, infraction, improper, incongruous, transparency, indicted, convicted, Jun. 13, 2019 suspect, con?dence, honesty, dishonesty, integrity, immoral, unlawful values, autonomy, independence, ethics, belliger- ent, blame, accountable, objectivity, ?Jnding sources, dis- closures, and promotional talks.? For example, the CREO- point system may identify and reduce the reliance on any potentially interested or compromised experts, such as doc- tors known or credibly accused of being paid by medical device or pharmaceutical companies, in connection with any related content. [0249] The CREOpoint system may then search for expert name matches that would ?ag that an expert in the long list database may have been reported to be associated with any of these qualities. These names may be subjected to further human review and curation, but if there is any doubt, individuals in the long list of experts would not make the short list. When people lose objectivity or become overly invested in their predictions, their judgement may be impaired?no matter how experienced or educated or intel- ligent they are. [0250] Once the breaking news and the companies in question are known, the shortlisted experts would be required to: [0251] 1. Con?rm they have no direct or indirect self- interest in their answer; [0252] 2. Decline to opine about matters pertaining to their current employer (whether public or private), govemment agency or any company where they are a major shareholder; [0253] 3. May not discuss his/her employer and con?- dential infomlation concerning its relationship with a customer or supplier; and [0254] 4. Respond only to qualifying questions in the survey questionnaire when they do not have a ?nancial interest for a company subject to a possible fake news item. [0255] The CREOpoint system takes into account best practices. For example, the Good Housekeeping Institute (GHI) has been around for almost a century and has built a stellar reputation for independence despite the fact that many of Good Housekeeping advertisers apply for GHI accreditation. Consequently, the GHI Seal Of Approval is seen by consumers as valuable and authentic. In order to maintain independence (and the value of the GHI brand), Hearst strenuously upholds a wall between the GHI and (and all other Hearst entity?s) commercial and editorial teams and ensures that GHI operates as a separate, independent organization. For example, access to the GHI lab in New York is passcode restricted and by appointment only. [0256] The CREOExperts may also be prepared to answer questions about perceived con?icts due to compensation for their participation in the program. If experts are indirectly paid by CREOpoint or its client company, it is expected that experts on both sides of the question will be compensated the same, so they are independent since their shared com- pensation is not linked to their opinion. [0257] To decrease chances of experts lying and given the increase in wearables, cues to deception such as an elevated heart rate, verbal and textual cues excessive quantity, reduced complexity) may be measured using and devices, with the prior consent of the expert involved. [0258] The CREOpoint system may also leverage recent research including ?ndings by Cornell University about two US new approaches for developing classi?er-based empirical trust-sensor models that speci?cally use electroencephalog- raphy and galvanic skin response measurements Clas- si?cation Model for Sensing Human Trust in Machines Using EEG and March 2018). [0259] Information, fake or not, will evolve beyond real- life, via virtual technologies such as in 3D, and other immersive environments. Words with verbal cues should help, and the CREOpoint expert networks from around the world are likely to be able to collaborate more dynamically with better context. [0260] Per ?Behind the Stars: The Effects of User and Expert Reputation Ratings on Users? Belief in Fake News on Social Media? (Department of Operations and Decision Technologies, Indiana University, August 2018). ?it is pos- sible to erode people?s trust in experts by labeling them as biased with an agenda. Hence, experts need to manage their reputation, and the entity proving the ratings must be judi- cious in choosing the experts.? Using the CREOpoint sys- tem described herein, experts who have a consistent bias will see their reliability rankings drop as their predictions will likely tend to be consistently erring in a way that is system- atic in one direction. The CREOpoint system identi?es such bias by cataloging errors of experts to determine if these errors are systematic and therefore indicative of biasi indicating either prejudged thoughts or con?icts of inter- est?or random and therefore just incorrect without bias. It is understood that most experts build knowledge through a combination of education, specialization, problem solving, career development, experience, extensive investigations into past research, articles and practices and by evaluating current news sources and breaking stories based on sound facts, experience, and evidence. [0261] Given that the experts are being asked for a quick ?judgment call? based on their ?educated guess? and without much time for their due diligence investigation in order to develop a quick News Veracity Score, the experts are likely to be subjected to their own biases. For instance, experts who believe that people are going to lose their jobs or die as a result of the advance of AI may be better at detecting fake news articles that praise AI in cars but may do poorly at detecting fake articles that are negative towards AI in cars. This is taken into account by the CREOpoint system as their reliability rankings would drop. Feature 3: Preparing to Complement the Predetermined Expert List as soon as the News Breaks [0262] The CREOpoint system may expand the list of predetermined experts based their proximity to items extracted from the breaking story. For example: BREAKING NEW87TF1 France Television: ?Tragic ?re in a crucial SNCF high-voltage substation near Paris expected to bring weekend railway traffic at Montpamasse Station to a standstill shortly. Investigators believe the ?re was started by a 53-year-old SNCF conductor (photo) who was found dead near the scene. No suicide note has been found yet. Jean-Eric Drouaix, reported to be an active ?Yellow Jacket? movement leader, had been vocal about SNCF refusal to pay train tickets for the activists as well as upcoming job loss due to autonomous trains being tested. Drouaix had recently been arrested and released following a group ?ght with journalists in Rouen. EDF and RTE released no ?irther details Friday night. No comments yet either from authorities as Elizabeth Bome, French Transport Minister and Guillaume Pepy, CEO of SNCF railway company are in Jun. 13, 2019 a plane on their way to a conference in Qatar. For more from his Deputy Agnes Ogier, follow the exclusive interview by TF1 Julien Beaumont also on my TF1.fr, Facebook Live and [0263] Unfortunately, almost anyone can create and instantly spread this type of tragic news around the world. Is it fake or fact? The perceived veracity of a news item such as this is enhanced by local events such as the French ?Yellow Jacket? movement?s actions in early February 2019; frequent train service disruptions in France; transit employee suicides reported in France; previous strikes; and growing opposition to autonomous vehicles. When such a news item is detected, the CREOpoint system considers the following factors: [0264] 1. Location in question (Paris, France): For example, a local authority, a Public Information Officer with the Police or Fire Department previously quoted as well as reported eye-witnesses would be added to the expert list since the CREOpoint system would closely track their reported reactions. They are likely to be one of the ?rst quoted in a rebuttal article, but not necessarily (see SNCF- ?inded Navy a shuttle [0265] 2. Narrower topics relevant in this example and that could have been pre?identi?ed ahead of time, especially if a major transportation company had previously engaged an entity using the CREOpoint system. In this example, topics would have pre-identi?ed with such terms as French ?Yellow Jackets?, SNCF employee suicides, French Trans- port Industry VIPs, activism against autonomous vehicles, ?res at critical sites, and disruptions of train stations. The CREOpoint system described herein would have pre-iden- ti?ed reporters who covered major events such as prominent TV anchorman Julien Beaumont of Les Echos Lionel Steinman who routinely covers SNCF, and William Audu- reau who covers fake news about the ?Gilets Jaunes? at the French daily ?Le Monde.? The time savings resulting from pre-identi?cation would be crucial since the CREOpoint system would immediately monitor and take into account whether they retweeted the story or warned their readers about the possibility that the news item could be fake news. [0266] 3. Relationship to the companies in question (SNCF, Keolis, Mairie de Lyon, Ministere de l?Interieur [0267] 4. Type of fake news: Misleading news not based on facts, fabricated to be intentionally deceptive, deepfake, fabricated picture, PR stunt, hoax, satire or parody, sloppy agenda-driven reporting, misleading news without context, origin of the fake news local activists, Russia), ?nan- cial scandal, sexual harassment, etc. The CREOpoint system indirectly provides readily available fake news detection and debunking tools thanks to and experts who are well informed and routinely and quickly fact check various types of fake news. For example, local French experts evaluating French fake news ?Les Decodeurs? from ?Le Monde?, CheckNews from ?Liberation?, CrossCheck from First- Draft) would have a higher weight in the overall scoring mechanism than a predetermined expert ?irther from the incident/county. [0268] It is anticipated that, as anticipated also by the former Chief Technologist at the University of Michigan?s Center for Social Media Responsibility, the CREOpoint proprietary database of screened experts could also be made available (for a fee or not depending on situations) to an US organization helping reporters and others looking to improve their work and sources they quote as they respond to breaking news. Feature 4: Developing Mechanisms to Incentivize these Experts to be Active in near Real-Time when Consulted [0269] In the CREOpoint system described herein, the short-listed experts would: [027 0] 1. Be contacted early in the process to establish a personal contact, possibly by mail, e-mail (which helps verify the address), text, secure apps such as WhatsApp, phone or sometimes face to face. Studies have shown that invitation letters by e-mail with links to the poll tend to have lower response rates compared to an invitation sent out before the actual poll is carried out (Andrews 2003; Cook 2000). [0271] 2. Understand the program?s aims and rationale and communicate to the experts that they would be among the few asked to answer a simple question about the veracity of a breaking news story relevant to their expertise or an important organization in their sector. [0272] 3. Be asked how they prefer to be contacted by CREOpoint (email, Messenger, Slack, WhatsApp, Twitter Direct Messaging . . . and possibly by other experts. [0273] 4. Receive a simple, single and focused sample question so they can immediately see what would be expected of them, including the ability to respond on the go from their mobile phone. [0274] 5. Have the opportunity to access FAQs (Fre- quently Asked Questions) and ask follow-up questions if needed. [0275] 6. Understand that communications will be via secured communications (and that they will be reminded by a call if their response/participation is delayed). [0276] 7. Ensure that their privacy can be protected. Although false identities will not be allowed, to ensure anonymity if required, an address register replaces e-mail addresses by identi?cation numbers, available only to a single administrator. [0277] 8. Receive drill noti?cations when relevant so the experts recognize that the system inquiry is not spam. [0278] 9. Not be surprised with new frameworks, de?ni- tions, formats, survey questions, or anything new without a proper well-explained business reason. [0279] 10. Have assurance that their contact or personal information will not be resold or shared with any other parties or entities without their express permission. [0280] Since the credibility of the CREOpoint system is likely to be criticized on whatever criteria, scores and experts are chosen, few details are made public. Instead, with their permission, few experts who provided the most robust rebuttals/testimonials are featured. [0281] Experts would also bene?t from unique advantages including: [0282] 1. Unique content such as fake news in their sector provided 24/7 thanks to the prede?ned Intelligence Channel and/ or Knowledge Board spreadsheet with curated intelligence news). [0283] 2. Being part of a partial list of influencers CREO- point and/or a media partner would publish with comments about fake news in the industry like what BuzzFeed did for 2016-2018 but with more focus on corporates and impactful fake news that hurt a company?s equity valuation. Jun. 13, 2019 [0284] 3. When the experts provide a response (percentage probability news is fake+supporting comment), the experts receive early and more detailed access to the CREOscore and real time commentary from other experts (assuming they have signed a NDA and will not release early incom- plete and unvetted information). [0285] 4. Experts will be provided the opportunity to use the data and the research when the study is ?nished. [0286] 5. CREOpoint will seek alliances with trusted institutions and other organizations that may display the and other deliverables. [0287] 6. Any compensation to the expert may be desig- nated to a charity most relevant to the expert or victims involved in a crisis news event. [0288] Experts may also be ranked in a leaderboard and advanced on the leaderboard as they get closer to and on the right side of the ?nal CREO Veracity Rating 14% true per consensus of experts), and then to the truth or reality (100% true after fact checking). These rankings based on corresponding trust-driven CREOpoints may be accompa- nied by unique badges that could be shared on social media for even more recognition and transparency. [0289] An example of how the CREOExpert scoring sys- tem works is as follows. When a news item is determined to be suitable for rating the veracity of the story, the CREO- point experts are polled. When poll results are ?nally compiled, out of 100 experts surveyed, 30 experts have responded and therefore they are credited with CREOpoints with the pie distributed based on some pro rata distance (positive or negative) to average veracity score (say con- sensus is 86% fake). People who responded the furthest (?Absolutely no way this is fake?) would see their CREO- Expert Trust rating penalized the most since their opinion was wrong. The experts who responded 100, the furthest positive distance to 86 would be given a higher rating than the expert who responded 86. Some who consistently have a bias anti Al, anti-technology) could see their rating re?ned based on the relative accuracy of their predictions. Eventually the veracity of the news item is resolved, and it is either fake or factual, or grey somewhere in between, and the CREOpoint system automatically adjusts the pert Trust Ratings. [0290] It is noted that since reliability is one factor in CREOpoint expert trust rating, in this example, some of the 70 non-responders scores would decline if they could have physically responded but, in this instance, they did not respond to the CREOpoint poll. [0291] As blockchain applications mature in the market- place, the CREOpoint system may use a distributed, shared and secure ledger to enable transparency and trust in the polling and commentary process. To provide opinion trace- ability, each step of the polling process is registered and tracked on the blockchain. Each of the expert?s votes, and any additional individuals identi?ed by the CREOpoint system as eyewitnesses or indicating signi?cant objective knowledge of the situation will have a unique digital token, enabling the CREOpoint system to verify every step of the distribution of potentially fake news items to the expert group, the subsequent vote by each expert, and any com- mentary provided. The ledger will provide a digital history of this information including data on location, source, con- tent, timestamps, origin and path of messaging, expert opinions, scores, rankings and how the CREOScores were derived, all of which is presented and available to users via US an interface they can access through Quick Response (QR) codes, Near Field Communication-enabled (NFC) labels, or location-based services, among others. A blockchain may also be applied to protect the original opinion, votes and comments from validated sources from ?invisible? tamper- ing, hacking, or cheating, including voting by non-experts posing as experts. [0292] Expert compensation may also be accomplished using smart contracts leveraging or altcoin platforms such as Etherium, Litecoin, Bitcoin, and other global payment networks employing digital currency and/or blockchain technology. Among other bene?ts, this would: [0293] 1) Help allocate the compensation of experts based on how good their predictions are; [0294] 2) Make the register including the expert scores, ranking and how we got there not cheatable/hackable; [0295] 3) Enable the experts to be compensated on a timely basis; and [0296] 4) Decentralize or diversify trust and allow CREOpoint to not be central point. [0297] Oracles, in the context of blockchains and smart contracts, are agents that ?nd and verify real-world actions and occurrences and submit this information to a blockchain to be used by smart contracts. Smart contracts only unlock value if certain prede?ned conditions are met. Within the CREOpoint system, oracles may be deployed to ensure the satisfaction of the temis of the smart contracts Feature 5: Preparing to Quickly Identify Potentially Fake News [0298] As noted above, the CREOpoint system extracts keywords and develops a search query such as (?self-driving OR autonomous OR AV) AND (Tesla OR Uber OR Cruise OR Ford OR NAVYA OR Car etc.) AND (crash OR hit OR struck OR victim OR injured OR killed OR fatality OR Accident) for a topic of interest. Relevant hashtags may be identi?ed and used to develop an intelligence channel relevant to the topic using the techniques described above. One of the Channel Interface settings may allow the display of all news from traditional and social media sources (called Setting All). On the other hand, predetermined short listed experts would be expected to be noti?ed early and serve as potential early identi?ers of fake news about a topic within their area of expertise. One setting would display only posts from these experts (call Setting Ex). [0299] Since fake news is often likely to be published in known fake news sources such as ?SpicyAmericaNeszu? and other sources/ super Spreaders and bots who have been known to frequently post and amplify fake news previously, the Channel may be constrained to these Fake News Sources rather than All Sources. [0300] The CREOpoint system may also further leverage the Interface Capability described above to adapt a toggle so that a user could conveniently switch between All Sources or only the most in?uential and relevant ones. Among others, the adaptable interface may be adapted so that a user could conveniently toggle between All Sources to Only Fake News Sources (called Channel FN, or to an XFN channel displaying all sources except those likely to contain disin- formation. [0301] It is noted that the Channel TFN may at times include some limited content from a few previously unknown fake news sources. This is due to constantly new fake news sources as well as ?domain cycling? where bad Jun. 13, 2019 actors launch sites with one or more fake stories and then quickly shift to new domain names once the story has been flagged by fact-checkers and/or blacklisted by Facebook or other platforms. [0302] To quickly develop coordinated countermeasures, CREOpoint will work closely in partnerships between social media platforms who, among others, have expert investiga- tors manually searching for sophisticated bad actor net- works, fact checkers, academic researchers ASU), foundations and software companies MetaCert). Fake news sources can be identi?ed and aggregated from such sites as: Perennial_sources tirical-stories/ Hoaxy project List by Grinberg et al. (2018, 490 sites) PolitiFact?s guide to fake news websites and what they peddle (Gillin 2017, 325 sites) Snopes [0303] BuzzFeed on fake news Silverman 2016; Silverman et al. 2017a; Silverman et al. 2017b; 223 sites+ over 10,000 news items December 2018) An academic paper by Guess et al. (2018, 92 sites) FactCheck?s article titled ?Websites that post fake and satirical stories? (Schaedel 2017, 61 sites) Lists assembled by blogger Brayton (2016) Media studies scholar Zimdars (2016) [0304] It is noted that the MetaCert Protocol already detects fake news by checking domains, websites, and news sources against the world?s largest network of fact-checking databases. The classi?cation of each news source is veri?ed by fact-checking organizations and stored in the MetaCert Protocol registry to provide unbiased, democratically assessed information on the integrity of each website and news source. The CREOpoint system described herein may aggregate signals indicating possible fake news sources to add to the above list. [0305] In addition, to complete the list of fake news sources, signals such as from ?The Manipulation of Social Media Metadata Data Society? (Data Society, Novem- ber 2018) indicate that often fake news sources can be flagged given their ?Double consonants, default avatars, random numbers, Screen name di?erent from user name, Name contains the words ?O?icial? or ?Real?, recent account creation followed by large numbers of suspicious accounts, nonsense comments from followers, sudden growth in followers or following, replies are automated messages, reshares, lack of transparency about the account, previous dormancy followed by sudden reactivation, use of similar hashtags or links, or responses with links etc. [0306] The CREOpoint system may also explore the most ?engaged? content relative to a given topic or domain (Facebook, Twitter and/or tools such as Buzzsomo, get. trendolizercom, or CrowdTangleinow owned by Face- book). A custom noti?cation (text, slack alert) may be programmed if the online reactions is at levels times above the average reactions expected. US [0307] It is also noted that, given anti-big business move- ments, extremists and toxic business models of engagement on social networks, fake news centering on a well-known multinational corporation spreads quickly and an unusual amount of negative online buzz is created, threatening the reputation of the company and their brands. Partnerships also may be developed with social networks such as Face- book, Instagram, WhatsApp, Twitter, YouTube, Snapchat, Reddit, LinkedIn, Tencent?s onne, and Sina Weibo to surface speci?c news relevant to a corporate client and their major brands. Social networks, corporations, and brands have a major stake in attenuating fake news. [0308] International Fact-Checking Networks provide a place for collaboration between fact checkers worldwide. Potential fake news stories are flagged by social and tradi- tional media readers and reviewed by Fact Checkers in key countries, sometimes as a service to Social Networks. The CREOpoint system may access this early data from fact- checkers who may also identify stories to review on their own. [0309] The CREOpoint system may also use the block- chain to track users who ?rst reported to CREOpoint a breaking story relevant to a client or topic covered by the CREOpoint system. Abounty system, similar to what?s now used to report bugs, may be implemented. For example, to mark a post as a possible false news a user may share on Twitter the post #CREOFakeNews and that would be a signal to start analyzing it and possibly rewarding the user given the corresponding time stamp. [0310] The impacted party may con?rm an attack by fake news such as ?Carnival cruise ship overturned off the coast of Mexico Nov. 5, 2018 resulting in 32 deaths The issue is then acknowledged at least internally and with technical advisors with expertise in such disciplines as public relations, public affairs, government relations, and external communications. The CREOpoint system and human intelligence would be applied to deter- mine if this were a situation where the CREOpoint experts should be polled to help determine the veracity of the story in the near-term. Feature 6: Deciding how Bene?cial Activating the CREO- point System could be as soon as the Problem News is Identi?ed [0311] A decision matrix may be used to score how valuable the CREOpoint system is based on: [0312] 1. The nature of crisis (see examples below); [0313] 2. Whether initiated by a usually reliable source or via a known fake news source that will not be taken seriously; [0314] 3. Whether the crisis has been identi?ed as a possible event in advance with the subject or entity where the CREOpoint system is being applied; [0315] 4. Whether a relevant population of experts has been identi?ed in advance; [0316] 5. How likely a new population of credible independent experts more likely to have speci?c insights could be quickly identi?ed to complement the predetermined expert list; [0317] 6. The likelihood, speed and quantity of proper responses by a panel of experts compared to credible journalists or other experts in the ?eld; [0318] 7. Deciding if a proper and useful poll question can be crafted. Say the news item is a report that a Jun. 13, 2019 corporate executive assaulted 12 women, but the truth is 5 only and 6 cannot be proven. Fake or fact? The CREOpoint system may activate the team survey pro- fessional to carefully craft the question such as ?What? the probability of veracity of this breaking news that the corporate executive is a sexual predator (with de?nition)?? [0319] For illustration, three different types of crises can be distinguished, assuming the news is not fake. I. Accidental Crisis (Technical Disruption) [0320] 1. Rumor of crash or industrial accident [0321] Small scale: This is likely the most suitable assum- ing a small scale; for example, the capability of an autono- mous vehicle fatality, since the stakes are high. The industry requires public trust and transport travel companies almost always put safety ?rst. [0322] Medium scale: A train accident such as 2581127525.html would be quickly covered by local TV if not fake. Live coverage, on the scene, would tend to con?rm the facts of the event and subsequent reports. The CREO- point system may be helpful if it is di?cult for camera crews to access the site a ship sinking in a remote area). [0323] Large scale: In a tragedy such as the World Trade Center attack or Fukashima breaking story, the CREOpoint system has little application since TV footage would go on all TV Channels almost immediately. [0324] 2. Product recallsiThere are always opposing viewpoints and arguments, e.g. in talcum powder litigation, there are big gaps between the company and the litigators, and various interests and opinions of the public. [0325] 3. Threat of additional litigation II. Preventable Crisis (Human Failure, sometimes Self- Inflicted) [0326] 1. Financial scandal?Such as fake news about Vinci shared by [0327] Bloomberg News that resulted in an 18% decline in Vinci?s market value before the company?s denial. The advantage is that the CREOpoint system could be integrated into broader systems from Big 4 risk mitigation ?rms and Financial Communications crisis-focused ?rms. The CREO- point system described herein could also be used to detect stock price manipulation or assist in hostile takeover situa- tions. For example, small biotech ?rms are often subjected to rumors or disinformation conceming the results of drug trials or patient reactions, for the purpose of manipulating stock prices. [0328] 2. Sexual harassment or serious abuse by a top executiveiFor example, for claims and accusations tagged as #me too (real and fake) some cases took place for many years before becoming public, even though people with insider knowledge had reasons to be aware. The veracity of some crisis news items may not be resolved quickly sexual harassment, oceans rising due to climate change), so the CREOpoint system focuses on use cases where the response to a fake news story could have a high impact within minutes or hours. [0329] 3. Brand safety and celebrity reputation?Beyond fake products and fake news about brands and stars, there is a growing industry of ?influencers? (including some with fake followers) who push brand products. Brands, their contracted ambassador stars, mega YouTubers and Insta- grammers, or even rogue micro in?uencers could directly or US indirectly pay for the CREOpoint system. For example, Chanel and Burberry ?red brand ambassador Kate Moss over cocaine allegations. Tiger Woods lost his Accenture sponsorship due to extramarital affairs, with investors in the three sports-related companies (Nike, Gatorade, and Tiger Woods PGA Tour Golf) faring the worst, a 2009 UC Davis found. They experienced a 4.3% scandal-generated drop in stock value, equivalent to about $6 billion. Fast forwarding to a recent situation one can imagine a situation where a cannabis brand whose $5 Mn Super Bowl TV ad was rejected by CBS would re-allocate its advertising budget to social media micro-in?uencers promoting its retail outlets. Would established brands like Philip Morris or Unilever want to be unwittingly caught up in fake news spambot warfare about the danger of medical cannabis? In the above high stakes situations, CREOpoint experts would know or could contact their sources, just as detectives and journalists might do. In the case of deepfakes doctored videos or ?virtual in?uencers? like Dior?s doll-like CGI in?uencer Noonoouri going rogue and sharing damaging fake LVMH news, analyzing social media signals supplemented by a CREOpoint network of polled experts is likely to provide faster better technical judgements. Meanwhile, with spam- bots, the cost of the false positive rate with arti?cial intel- ligence is high making a mistake in mistakenly block- ing regular accounts is not acceptable even if 1% of hundreds of millions of users). [0330] 4. Executive rumor involving sensitive matters [0331] For example, executives being accused of compro- mising situations leading to divorce of a billionaire e-com- merce CEO and spouse; a serial tech entrepreneur accused of smoking pot or rumored to be quitting, or an executive being arrested in a foreign country. [0332] 5. High pro?le closures and redundancies [0333] 6. Regulatory or media investigations [0334] 7. Allegations from political leadersiThis use case is unlikely to be used immediately. Political experts often cherry pick information favorable to their position or point of view. Also, note expert predictions in this ?eld are often wrong (starting with magazine/ 2005/ 1 2/ 05/ everybodys-an-expert) [0335] 8. Company contributes to a cause that does not align with company?s core valuesiprobably not something experts would de?nitely know/can opine about, and there- fore this use case is a lower priority use case. 111. Victim Crisis [0336] 1. Major hacking and data breaches?easy to iden- tify companies like Equifax and Marriott that would seek to stop the spread of fake news or further data breaches. [0337] 2. Malicious rumors like the death of a political or business leader or celebrity. Given the high stakes on commodity or security prices, it is possible to be asked to prepare a group of experts to be polled on Syria, oil, etc. and to rate Syrian President death rumors and similar news reports. [0338] [0339] 4. Terrorism and major crimesiThis is unlikely to be a practical use case since usually these events are very visible by design from terrorists. On large scale news reports, TV networks quickly go live with evidence. 3. Activist attacks and accusations Jun. 13, 2019 Feature 7: Polling all Selected Experts in Near Real-Time to Quickly Develop a Veracity Score for the Breaking News Content [0340] In the always-on information overload mode of modem society, it does not help that some people tend to be time-starved, lazy, do not take time to apply critical thinking and/or, among others, suffer from con?rmation bias (the headline supports the person?s initial position) and some- times delusionality, dogmatism, fundamentalism. Some people believe blatantly inaccurate news headlines such as ?The Pope endorses Trump? because they have unfortu- nately already seen shared a number of times from their ?friends? (Eli Pariser ?Filter Bubble? 2011). The customi- zable intelligence channel described above provides NLP- powered ways to only display a news item headline once. This allows the CREOpoint system to not only de?ate the devastating and polarizing ?lter bubble effect but also con- tributes to important health and wellness ?Time Well Spent? initiatives. [0341] The CREOpoint system described herein attenu- ates disinformation by seeking to stop the falsehood in its tracks with robust content including a warning. David Rand has also showed the bene?ts of care?illy worded warnings (?The implied truth effect: Attaching warnings to a subset of fake news stories increases perceived accuracy of stories without warnings? 2017). [0342] Among others, Facebook has been experimenting with various warnings and pop-ups such as ?disputed? (Adam Mosseri, VP News Feed December 2016 and Tessa Lyons, Product Manager. December 2017). For example, affecting [0343] Google?s Android, it took days to get to the facts and the ?disputed? warning. On 26 Feb. 2017, the Seattle Tribune website published an article concerning President Donald Trump, reporting that his Android cell phone was believed to be the source of recent intelligence leaks from the White House. That article cited two non-existent ?intel- ligence agencies,? A.R.H. Intelligence and le3 Security, as its primary source, and hijacked the social media hashtag which has been used to encourage cell phone users to spend more time away from their tech. On March 2, Snopes declared this as a bogus news story. On March 3 at 4:28 pm, Politifact labelled this as a false news story. Finally, using these two signals, Facebook added a ?disputed? ?ag to the post. [0344] Speed of response matters. Also due to sometimes deeply held user beliefs, Facebook found that showing ?Related Articles? rather than ?Disputed Flags? was found to help give people better context. ?Related articles? is the foundation of the customizable intelligence channel described above as curating articles related to a topic, organization, personality or event is an important part of the customization of the intelligence channels. The CREOpoint system described herein re?nes a single and simple score in a user interface providing people justi?cation not only for a strong warning if needed but only related input from vetted CREOpoint experts who did not fall for the fake news. [0345] Since fake news may commonly mix true state- ments with falsehoods, rather than a binary value, the likelihood that the news is fake or fact is predicted 86% likely accurate). A quick CREOpoint system poll of these trusted experts instantly refreshes a weighted CREOpoint Veracity Score that is posted on a pre-prepared and private dark site. US [0346] Although the CREOpoint system may be used by clients, interested parties and survey experts may draft and evaluate the pros and cons of other questions to elicit ideal results. The expert panel would typically be asked to opine on a single and simple poll question as illustrated in FIG. 18, for example. testing may be performed to optimize a question such as: ?Please respond to the question below on a scale of to 10, considering all you know right now, and the impact on your reputation: How likely are you to vouch for its veracity by recommending this breaking story to a colleague by sharing it as is?? This question grounds the CREOpoint system in the existing knowledge of NPS (net promoter score): Promoter into a way to see how comfortable the experts are in recommending the news content. [0347] In the interest of turnaround time and to maximize expert participation, automated phone surveys may be added as a systematic, lower cost and faster collection mechanism. Calls may be made automatically to the preset list of experts with the aim of collecting information and gain feedback via the telephone. Interactive voice response, or IVR, technolautomate telephone contact between humans and machines. Teleservice company call centers, market research, expert networks and other partners may also be used to warn experts that a written poll is going their way, or to follow up if they are not fast enough in responding to it. [0348] CREOExperts would either be detractors, passive/ neutral, or promoters super sharers on social media): [0349] 1. Low (not at all likely to share):Detractors: Experts who feel the news item is fake might even be active detractors by retweeting with a factual debunk- ing comment (Marked as RED in interface); [0350] 2. Medium (neutral/undecided yet): People who are not sure whether fake or fact are unlikely to repo st, given the risk of being wrong; and [0351] 3. High (extremely likely/sharers as is): Experts who feel it is valuable news will feel the urge to share it, possibly even with enthusiastic reinforcing com- ments. [0352] The CREOVeracity Score of the news item being questioned may also be re?ned by offering a forum for experts who are con?dent in their disagreements. Responses may form a normal Gaussian distribution, looking like a bell-shaped curve, and it is possible that during a poll, values will follow a normal distribution with an equal number of measurements above and below the mean value. To improve the dynamic CREOVeracity Score, as a re?nement to the initial answer, the CREOpoint system may ask experts who gave divergent opinions and answered high to the question ?What level of con?dence do you have in your answer? (Low Con?dence to Near Certainty)? to try to convince either other one way or another in a dedicated forum. [0353] Adequate time should be allowed so the experts can form an informed opinion but not too long to avoid their consensus being obsolete when it becomes public because other sources have proven the news fake or fact. Experts have some time to think about the breaking news, see who is saying what on the provided CREOpoint Intelligence Channels, Twitter, call their sources, etc. To prevent a herd mentality, they would not have access to the dynamic score until after they have opined (similar to Airbnb reviews). A certain statistically-valid critical mass of expert responses needs to be reached before the results are deemed usable. Jun. 13, 2019 [0354] To complete the list of sources to be considered for addition, the CREOpoint system may also ask the experts which other priority sources should also be monitored. [0355] Also, to improve the dynamic as a refmement to the scaling system applied by the experts, the CREOpoint system may be re?ned into a predictive market system. Predictive markets are collections of people specu- lating on future events or outcomes. Prediction markets essentially are event derivatives, where the value of the derivative will in most cases re?ect the probability of an outcome actually occurring. The technology to implement a prediction market is under development by: Augur, an open-source, decentralized, peer-to-peer oracle and predic- tion market platfom1; and Gnosis, a prediction market plat- form. Both platfomis are built on the Ethereum blockchain. [0356] It is anticipated that CREOpoint system may accommodate clients or licensees requiring technical and UX adaptation of the timing, ratings or others to align with their brands CBS may require the best expert judge- ment available after ?60 Minutes? and Amazon may prefer to display expert ratings in a format similar to their 1-5-star customer reviews). Optimization models for the question, polling protocol, modeling, noti?cations, work?ows, forum and real time conference coordination will be re?ned thanks to drills and real KPIs from fake news in a safe client drill area. This would also allow advocates who could also be pre-identi?ed and tracked over time before they might be helped to augment the message more authentically. The resulting CREOscore could be a number between 1-100 or a number of stars 1 to 5 stars). Feature 8: Adding other Veracity Signals from the Behavior of Experts and Crowds Sourcing other Trusted Sources Sharing or not Sharing the Breaking News on Social Media Social Media Signal Analysis [0357] Tracking what CREO point experts and trusted media sources see earlier list including AP, Reuters and Bloomberg) post or repost will allow the CREOpoint system to provide an indication, leaning towards fake or fact, that is helpful to rate the veracity of the news content. The CREO- point system may then analyze the text of retweets, Face- book posts or other messages from detractors (still from this expert group, not from laypeople) who post comments like can?t believe this! #fakenews?, or similar to John McEn- roe famous ?You cannot be serious.? That rebuttal would be computed as a negative in the calculation of the news item veracity score. For example, Cedric Ingrand from reputable LCI French TV debunked the Promobot PR stunt at CES as fake news. [0358] The CREOpoint system also monitors key sources who take their comments down. For example, when contro- versial meal replacement drink company Soylent claimed ?Furloughed Federal Workers Surviving on Soylent? report- ers from BuzzFeed and Gimiodo did not accept its validity. Note how their tweets referenced in this article employees-government-shutdown have now been removed. The CREOpoint system may count these as a signal of fake news. New Sources [0359] After con?rming the high relevance and in?uence of a commentating source, the CREOpoint system may add US the commentator to the trusted media source list and count the commentator?s insight. These types of new sources are tagged in the CREOpoint database with classi?ers such as activists, fans, persuaders, clari?ers, Spreaders as is, or debunkers identi?ed post incident. [0360] To identify people who could help identify and/or debunk fake news, the CREOpoint system may also use NLP to create a corpus by analyzing what words and hashtags are used by the ?rst people debunking fake news such as in the BuzzFeed of 10,000 fake news and hoaxes and would include words like fake, false, hoax, mis- information, disinformation, lie, debunk, correction, no basis in fact, lie, said no, no fact at all, chain mail, fallacious, bullshit, BS, etc)? [0361] For benchmarking purposes and to measure the e?ectiveness of the CREOpoint system, the ratio of the rumor posts divided by the news items and one of the keywords would be measured. The lower the ratio, the more effective the performance of the CREOpoint system. [0362] The sources to be added to CREOpoint monitoring would be identi?ed from the initial list of experts and trusted media sources who they recommended as other potentially useful sources to be considered for inclusion in the database and short listing. This information may be added to the CREOpoint system to complement the database by short listing these possibly useful sources based on metrics such as in?uence and relevance. New Intelligence Channel Searching for the News Item Being Questioned [0363] As noted above, the CREOpoint system also may develop a customizable intelligence channel using the search query including the news item being questioned. For example, thanks to a query checking on the early February 2019 fake news about the death of a European movie star: ?Gerard Jugnot? AND ?mort?) followed by AND NOT (hoax OR canular OR ?fake news? OR lie OR debunk etc. based on the ontology described earlier), the CREOpoint system would identify many people including traditional media reporters on social media who doubt the veracity of this statement. The CREOpoint system also may be supple- mented by a human curator (to avoid false positives due to sarcasm) who would add evidence, counter-evidence, and record all sources in an online spreadsheet or database. Sources who are consistently found to report fake news may be considered for addition to the CREOpoint list. Enriched Database of Reposts and Commentaries [0364] Any high-value relevant content may be curated in a proprietary online library to create an enriched database of reposts and commentaries. Relevant reposts ?this is fake and here?s why?) would be saved by offering the CREOpoint system user the ability to simply click on a browser extension that enable the post to be tagged in one of the above categories super spreader as is, debunker) and saved. Users are also allowed to enrich posts on the ?y using notes and highlights. [0365] The CREOpoint system may allow enriched posts to be shared through: [0366] A comprehensive online spreadsheet (including columns such as post headline, date, URL, of engage- ments to date, source, in?uence and relevance of the source, CREOpoint Trust Rating if the source is an Jun. 13, 2019 expert in the database, type of sources, team highlights, of posts supporting and debunking the claim, team comment, recommended actions, etc); [0367] Dedicated emails to an internal team member; [0368] Newsletters integrating with Mailchimp or similar, for intemal and/or external purpose for example automated transmittal of top debunking tweets from other reporters going to other reporters); [0369] Twitter direct message; [0370] Slack; and [0371] Dedicated ?Creme de la creme? CREOpoint Intelligence channels (or separate interface toggle set- ting in the user interface) with just the best content post AI +human curation (possibly from a team). Visualization [0372] The CREOpoint system may further leverage a tool such as Table 2 Net (from leading French University Sci- ences Po MediaLab) to extract a scientometrics network. The scientometrics network could then be visualized with a tool like Gephi Feature 9: Alerting In?uencers and the Public about the Low Veracity of the News Content and Reducing the Spread of the Fake News [0373] As part of risk management and mitigation plans, the CREOpoint system may use a hybrid human+AI designed for a worst-case scenario with ?weakest links in the chain? in mind: [0374] Support stalf available on call 24/7 in regional centers in the Americas, Europe and AsiaPac, including Account Executives, Project Managers as well as trained CREOpoint Curators and quick access fact checkers in key languages; [0375] Multiple accessible interfaces and visible spokespeople for the client organization and their advi- sors; [0376] A capability developed to contact and poll key stakeholders quickly and directly; [0377] Possibly working with industry organizations with capability to back the messaging; and [0378] Other technology companies, researchers, aca- demics, law enforcement, regulators, election commis- sions, and civil society groups. [0379] Knowing in advance the countries that will be the theater of the fake news is important for preparation and a quick reaction. Some countries and upcoming events can be anticipated as a future hotbed of fake news and attacks at scale from ideologically-motivated networks of anonymous trolls the spotlight on Qatar and the 2022 World Cup). Others will be more important to a client than others due to the location of their people, operations, customers, and supply chains. [0380] In other countries outside the United States, like places such as Brazil where intemet access is very expen- sive, fake news can have a signi?cant impact depending on the economic stakes, Net Neutrality debates focusing on the impact of price discrimination. Zero Rating models (free intemet access under certain conditions) are mainly implemented on mobile networks and are based on subsi- dizing a limited set of sponsored applications, whose data consumption is not counted against the users? data allow- ance (often in exchange for user data and low quality/ fake news). US [0381] The immediate objective of the CREOpoint system is to contain the infectious agent locally if possible. To do so, the CREOpoint system may purchase the most relevant keywords in advance for Google adwords, organize Face- book and Twitter campaigns, prepopulated with major cities where key journalists are based (New York, Washington DC, San Francisco, London, Paris, etc.). The campaign may be adapted by adding several keywords and locations even more directly relevant to the breaking news story. For example, in the case of the Russian PR stunt damaging to Tesla at CES, Las Vegas would be added so anyone in the area checking on early rumors would be directed to the dedicated webpage. The proprietor of the CREOpoint sys- tem may also coordinate with a corporate client for other types of counter-messaging and counter-narrative commu- nications and advertising should the crisis and crisis avoid- ance requirements be very serious. [0382] The CREOpoint system may also produce and offer access to more positive and loud messages to overcompen- sate and attenuate the negatives of fake news sources by, for example, featuring anew page called a CREOboard explain- ing: [0383] Upcoming CREOscore; [0384] Expert comments; and [0385] Meanwhile informing people by disclosing the top 5 worst fake news about that industry sector. This CREOboard may have its own identify/URL such as [0386] Fact checkers may be helped by being provided with relevant Intelligence Channels to help ?lter out the problematic noise (duplicates, false positives, low value sources) inherent in news reports, enabling the fact checkers to improve productivity and turnaround time. [0387] To prevent spread to local and regional super spreaders (network in?uencers spreading ideas, informa- tion), nano in?uencers and in?uencers, the CREOpoint system proprietor or licensee may make contacts in local and regional expert and in?uencer communities and develop such relationships before they are needed. A predetermined database of local and regional in?uencers and experts (as well as other people they recommend) may also be cross- referenced with the list of people on site (available from a conference attendee list, self-check ins on Twitter, etc.) and add key local reporters and regional press focused on the event CES) press list (they would be the most likely to care and have good local sources that could help). The key possible fake news spreaders closest to the incident would receive a short message adapted from a template, such as ?(First name) did you see (add link)?! I know you think twice about sharing this type of video? (message to be tested). [0388] Proven solutions to push voice, text/SMS, and email noti?cations may be integrated with the CREOpoint system. This allows the next focus to quickly be on a wider in?uencer network, including key in-country in?uencers who may be provided early and exclusive approved quotes from CREOExperts. [0389] The CREOboard may be strategically shared with a select group of highly connected in?uencers who typically act as ?rewalls to fake news and more authentically spread the CREOpoint-provided expert content. Given that these ?in the know? influencers may have previously visited a CREOpoint website such as a CREOboard, the CREOpoint system may trigger retargeting, the process of following or Jun. 13, 2019 tracking website visitors online after they have left a web- site. When a person visits a CREOboard, the browser drops a cookie (a piece of data that embeds itself in the browser of the user, enabling tracking of the websites they visit). The cookie then implements the retargeting strategy by placing CREOpoint ads on other sites that the user is visiting, enticing them to click through and return to the CREOboard. These ads are only targeted at users who have previously visited a CREOpoint site shown interest in a previous CREOboard) but have not yet completed a call to action such as Share a Fake News Warning. This is why it is called ?retargeting??the website is targeting prior visitors again. The CREOpoint system also may shortlist this select group based on their in?uence, relevance and popularity and other criteria relating to authentically communicating the truth. Metrics that may populate the algorithm may ?irther include uniques on TV in the last 12 months, number of followers on Linkedln, Twitter, YouTube, Facebook, lnstagram and others. [0390] The overall CREOpoint system is designed to attenuate the ampli?cation of harmful content from speci?c relevant actors from the ultra-densely connected core of heavily followed accounts that repeatedly link to fake or conspiracy news sites, knowing that ?80% of the ones active in 2016 are still active as of the October 2018 writing of ?Seven ways misinformation spread during the 2016 elec- tion? George Washington University and Graphika for the Knight Foundation. After an outbreak of fake news to some users in countries beyond the country of incident, the contamination may begin to spread internationally from the Western United States CES Jan. 6, 2019 evening to the UK when people wake up in the morning of Jan. 7, 2019 to the Daily Mail story). The objective is to contain fake news in key countries, regions, cities by engaging with key stakeholders and possibly government authorities in these key areas. The time would then come to provide a tool to people wherever they get their news, including in their email as well as on social network Facebook, Twitter) or Apps Google News, Apple News, BBC News, Flipboard, Feedly, Reddit, SmartNews, News Break or others including CREO- point powering intelligence chan- nels about various topics, brands, personalities and events. [0391] A mobile application may be used to share CREO- point Veracity Scores for news items directly on user?s phones. For example, news items that have been identi?ed as possible fake news, and subjected to CREOpoint expert polling and scoring, may be ?agged with a CREOpoint Veracity Score and a link to the page describing the basis for the score and explaining why caution is required. [0392] The fake news should not only be constrained to key in country influencers, but they should be engaged to stop spreading fake news. To address this issue, perts may be engaged to gather some early quotes that could be shared with key in?uencers. The experts would focus on a wider in?uencer network, including key in country in?u- encers. The CREOpoint page would be shared with highly connected and smart people who typically act as ?rewalls to fake news and spread the CREO truth. The CREOpoint system would look at influence, relevance and popularity. Although it does not increase the professionalism criteria, it would be better if experts can also leverage their social US in?uence to authentically communicate the truth. Metrics that would populate the expert rating algorithm would further include: [0393] uniques on TV in the last 12 months [0394] of followers on Linkedln [0395] of followers on Twitter [0396] of followers on Facebook [0397] of followers on Instagram [0398] Emails may go out to pre-rented mailing lists with a message like ?Tell your friends about this fake news: Hit reply to this email, add three friends to the CC line, and write a line or two of background if you?d like. People may also be directed to copy the hyperlink to the to his or her clipboard and be prompted to click on a provided hyperlink to a CREOpoint system generated search query on Facebook BOX), Twitter c=typd), Google, etc. The user may be allowed to conve- niently add the CREOpoint hyperlink in the Facebook comments or retweet of friends or others sharing the news. This would provide for the warning and supporting content to be shared in friend?s news feeds. [0399] As and when necessary, a pre-populated press release may be adapted and sent. Such a press release may include: [0400] The breaking ?news? [0401] Corresponding veracity metrics and warnings [0402] Denial from the affected CREOpoint client com- pany [0403] More authentic quotes from a couple perts [0404] Quote from a local witness if available [0405] Quote from a relevant in?uencer [0406] Supportive audio video [0407] All these steps would be tested in crisis simulations ?re drills) where company personnel and crisis/PR ?rms could improve their response to a fake news situation. Feature 10: Monitoring the Impact of the CREOpoint System and Iterating to Re?ne [0408] The CREOpoint system provides a countermeasure against smart and potentially well-funded adversaries who are constantly adapting and changing their tactics. The CREOpoint system is continuously improved by developing metrics and benchmarks to improve its predictive and attenuation model and its impact, such as for example: [0409] Independent surveys of the perceptions and behav- ior of users before and after seeing the [0410] Changes in the number of fake news items about a client, or a person, a subject or organization; [0411] A change in the number of fake news items quickly reported; [0412] A change in the time needed to verify a relevant fake news item; [0413] Average time to ?kill the fake news in the egg? from people dominating the topic before it resonates with in?uencers who propagate the message, possibly leading some traditional media to pick it up; [0414] Independent surveys of CREOpoint experts; [0415] Reliability of CREOpoint experts versus others quoted in the media; [0416] Number of mentions for CREOscore; Jun. 13, 2019 [0417] Number of sources and social networks leveraging the CREOpoint system; [0418] The number of propagators; [0419] The number of Shares and Likes on social media; [0420] Applying predictive market concepts to increase the engagement of experts and the accuracy of their predic- tions and polling; [0421] Applying blockchain and smart contracts to the news item distribution, polling, communication and com- pensation process; [0422] Impact on customer trust and sales volume, reten- tion of clients or stock value; [0423] Impacts on brand image, perception, and equity; and [0424] How many influential nodes and information bro- kers are connected to various sets of networks (influencers are enormously powerful and may easily di?use misinfor- mation and disinformation with limited consequence to their reputation within a network). See as early as in 1992 with Burt, R. S. ?Structural holes: the social structure of compe- tition? Harvard University Press. [0425] The CREOpoint system further envisions incorpo- rating best practices in success measurements and constant improvements from research in social media, neurocogni- tion and other ?elds including aircraft maintenance as well as ?re and pandemics and other health risk propagation. Ideas to be evaluated include, for example: [0426] ?Inoculating the Public against Misinformation? (Dr. S. van der Linden et a1 Department of University of Cambridge 2017); [0427] If tech addiction and viral engagement around fake news are ?the new cigarettes?, warnings could be tested leveraging ?earnings from the tested mandatory warning ?Smoking seriously harms you and others around you;? [0428] Finding Outbreaks Faster National Institutes of Health 2017 Apr ticles/PMC5404242; and [0429] Providing a dashboard for say 30 companies in a Dow Jones Index rather than 30 counties in: http://emer- gency.vic.gov.au/prepare/# understanding-wamings (see other tabs). [0430] While intelligence channels and uses thereof have been described in connection with the various embodiments of the various ?gures, it is to be understood that other similar embodiments may be used, or modi?cations and additions may be made to the described embodiments of an intelli- gence channel without deviating therefrom. For example, one skilled in the art will recognize that embodiments and application of an intelligence channel as described in the instant application may apply to any environment, whether wired or wireless, and may be applied to any number of such devices connected via a communications network and inter- acting across the network. Therefore, an intelligence channel as described herein should not be limited to any single embodiment, but rather should be construed in breadth and scope in accordance with the appended claims. What is claimed is: 1. A computer-implemented method of rating the veracity of content distributed via digital communications sources, comprising: creating an ontology and selecting keywords for at least one topic of the content; creating a customizable intelligence channel for the at least one topic of the content and extracting from the US customizable intelligence channel a ?rst list of potential experts on the at least one topic of the content sorted by at least relevance and in?uence; mining trusted media sources for the at least one topic of the content to extract a second list of potential experts on the at least one topic of the content; providing the ?rst and second lists of potential experts on the at least one topic of the content to a database; rating and ranking the potential experts as a function of at least one of professionalism, reliability, proximity, experience, responsiveness, and lack of self-interest in the at least one topic of the content to identify a short list of experts; providing the content to the short list of experts for evaluation; polling the short list of experts about the veracity of the content to create a veracity score; and delivering the veracity score with the content. 2. The method of claim 1, further comprising creating a third list of potential experts and any local witnesses on the at least one topic of the content based on at least one of a relationship and a proximity of the potential experts to a breaking story on the at least one topic of the content and providing the third list of potential experts to the database to complement the polling. 3. The method of claim 1, wherein delivering the veracity score with the content comprises at least one of issuing a pre-populated press release, initiating a social media and press campaign including the veracity score and at least one of a warning and denial if the content is not completely true, issuing a quote from an expert from the short list of experts, issuing a quote from a local witness to the at least one topic of the content, and issuing a quote from an in?uencer on the at least one topic of the content and related reassuring metrics including infomiation about trustworthiness of sources of the content. 4. The method of claim 1, wherein a fake news warning is presented with the veracity score and content along with insights and metrics relating to the content. 5. The method of claim 4, wherein the veracity score, content, fake news warning, insights and metrics relating to the content are delivered via an interactive interface 28 Jun. 13, 2019 enabling a user to select the types of sources by level of trust or proximity to the news content or user. 6. The method of claim 1, wherein providing the ?rst and second list of potential experts on the at least one topic of the content to a database includes predeterrnining a list of experts who could best crowdsource veracity signals for a given topic of content. 7. The method of claim 1, wherein mining trusted media sources for the at least one topic of the content to extract a second list of potential experts on the at least one topic of the content is performed upon the release of a new story. 8. The method of claim 1, further comprising incentiviz- ing experts to be active in near real time when consulted by compensating experts based on how accurate their predic- tions are and creating a decentralized register including expert trust ratings. 9. The method of claim 1, further comprising creating at least one customizable intelligence channel for at least one topic of the content relating to potential sources of fake news and the semantics of fake news content. 10. The method of claim 1, further comprising creating a decision matrix to evaluate a breaking news story to decide whether the news story is a candidate for detemiining a veracity rating based on at least one of the nature of the breaking news story, a source of the news story, and whether a relevant population of experts readily exists. 11. The method of claim 1, wherein polling the short list of experts about the veracity of the content to create a veracity score occurs in near real-time. 12. The method of claim 1, further comprising modifying the veracity score to reflect the behavior of additional experts and trusted sources in sharing and commenting upon a breaking new story. 13. The method of claim 1, further comprising bench- marking the veracity scores to create a predictive fake news spread containment model and iterating to revise the model and overall performance of the model over time.