Increased media misinformation is negatively impacting public understanding, negating the positive effects of accurate information. Misinformation affects human progress on a myriad of fronts: such as climate change, Covid-19, protest and political activism, or the proliferation of common misconceptions. This initiative will provide the public with an easy-to-use tool (browser-extension) to identify misinformed media articles, comparing it to other articles and placing the article’s claims on a spectrum of consistency, using modern NLP techniques. The scope of this project will focus on articles related to Climate Change, to achieve proof-of-concept before applying to varied other topics.

Our four main objectives can be summed up as: Context, Reputation, Evaluation and Dissent.\r\n\r\nContext\r\nOur browser extension aims to provide context to various media articles. With the click of a button, a user would be able to see a quantifiable figure, representing the similarity of claims to other articles. With the click of another button, the user could see detailed information regarding their article’s claims. It would include a line diagram, representing the spectrum of articles with various claims, for visual clarity; this would allow users to navigate the diverse landscape of opinions, in an accessible way. \r\nReputation\r\nOne use of this growing measure of article context is keeping track of media outlets’ leanings toward and from the mainstream Climate Change opinion. It would take the form of an emerging score for news outlets: a sort of accumulation of each article’s individual scores, but not reliant on any one particular outlying article. The intention for this score is a general guideline toward media outlets’ inclinations on Climate Change, yet may potentially be useful for users looking for mainstream or deviant opinions.\r\n\r\nEvidence\r\nOur initiative, ultimately, is a reaction to a growing culture of media misinformation. So naturally, evidence is important. In the process of accessing detailed information, users will be provided with full links to articles from which their context scores were derived. \r\n\r\nDissent\r\nA small but interesting part of the program is a Devil’s Advocate feature. That is, using the information on the correlation between articles, our program would recommend articles of a different point of view, or at least, the ‘most’ radically different view of the topic at hand. The system will be carefully designed to trail the fine line between contrarianism and fake news. \r\n\r\nThese objectives will be achieved through various NLP techniques, to identify claims, and correlate articles. Claim identification will use a mixture of POS tagging, and lexical centrality. In addition to Latent Semantic Analysis (LSA) and sentiment analysis, Latent Dirichlet Allocation (LDA), a topic modelling technique, will be used to specifically identify climate change related claims. For correlation, we plan to use LSA supplemented with processes derived from LexRank (lexical centrality), to generate claim/document similarities.

The curated news feed will be made available to the internet, for potential users to receive. The program itself will become available as an open-source project for those interested to use.\r\nThe hope is that these programs will be further expanded outside the scope of Climate Change in the future. The ability to measure the inter-article consistency of an article can become a powerful tool to mitigate the effects of misinformation. The program’s scope could possibly be expanded to include other major issues inclusive of public health, gender inequality, political misinformation etc. \r\nAnother possible expansion will be the introduction of multi-lingual sources. The scope of the current program is english articles, which does limit the extensibility of use to English-proficient users. The conversion to non-English NLP techniques is not a trivial problem; the simple use of google translate would not give sufficient conversion process.

ITS Education and Global Citizen Capital will be working with Jeff Kim as advisors. This will take the form of meetings on a weekly/bi-weekly basis to update on the progress of the initiative as well as take advice. Jeff Kim will be developing the program all throughout the duration of the initiative. \r\n\r\nTo scale the initiative, we will also be reaching out to local news organizations as well as global NGOs who regulate information distribution for some advice. Their input, based on expert experience in the field of journalism, will help steer the project development towards an outcome with more impact. This interaction will mainly take the form of email correspondence. \r\n\r\nIn terms of digital coordination, our program, run on cloud computing, will be a server for each user client. The chrome extension, installed on a user’s computer, will access this server for up-to-date information, as well as requesting a certain article for analysis. Our store of articles themselves will be archived on the cloud storage database.\r\n\r\nOur climate change article database will be updated daily, searching various news RSS feeds for new updates. This database currently includes mainstream news sources such as The New York Times, The Guardian, Fox News as well as less commonplace sources such as the Environmental Defense Fund’s Climate 411 blog, the Union of Concerned Scientist or NoTrickZone (a popular blog for Climate Change Contrarians). This is to convey the mediasphere in its most holistic sense.

ITS Education, Global Citizen Capital and Better Together Foundation