Reddit is a very large discussion organized into multiple discussion communities (known as \u201csubreddits\u201d), which can have widely varying discursive cultures\u2013and information quality. The Boston University Climate Disinformation Initiative is exploring the prevalence and distribution of climate disinformation on Reddit using computational tools ranging from keywords to machine learning.<\/p>\n
Focus #1: High-level overview of Climate change discussion on Reddit<\/h4>\n
To develop a general impression of climate change-related discussions on Reddit, the research team identified three significant subreddits devoted to climate change discussions. This includes a climate skepticism-driven community (\/r\/climateskeptics<\/em>), and two climate science-related communities that are science-based (\/r\/climatechange, \/r\/climate<\/em>) and prohibit climate skepticism under their community guidelines. Using the same topics identified, the researchers then estimate what these communities (both the science-driven, as well as conspiracy driven) are discussing, based on their topical focus, and how these communities are different from each other:<\/p>\n Topic models are useful for revealing the topics being discussed in a discussion space; but they are less able to assess the quality of the information being exchanged. In other words, they do not tell us much about whether disinformation or misinformation is present.<\/p>\n Researchers are currently in the midst of refining machine learning tools for detecting disinformation about climate present in Reddit discussions. Because statements on Reddit are often nuanced and filled with dense argumentative details, their current approach uses a grounded entailment application of machine learning. That is, each claim about climate change detected in the Reddit data is evaluated against a set of statements about the scientific consensus, and corresponding climate denial claims. The goal is to train the machine to make accurate assessments about whether a given claim more closely resembles the scientific claim, or the skeptical one. At present, the model is able to identify climate skepticism in the Reddit communities with an 83% accuracy (F1 score of 85%).<\/p>\n The researchers use this methodology to evaluate the stance of Reddit posts discussing the 10 most popular claims from Skeptical Science<\/a>. From a random sample of 9,842 discussions automatically classified through the model, they have found that:<\/p>\n The skeptic claim that carbon dioxide is plant food and therefore not harmful<\/a> had the most skeptical claims, with 81.44% of posts related to that claim refuting the scientific consensus, while the claim that global warming stopped in 1998<\/a> was second with 71.24% of skeptical tone.<\/p>\n The claims with the least skeptical discussion were the claim glaciers are growing all over the world<\/a>, with only 44% of the posts as skeptical, and hurricanes are getting more intense due to global warming<\/a> with 42% of the discussion as skeptical.<\/p>\n Interestingly, the researchers find that science-based claims and climate skepticism are intermingled in both scientific and skeptical Reddit communities:<\/p>\n In other words, there is not strong evidence in the Reddit data for the idea that subreddits are \u201cecho chambers\u201d: readers of these subreddits are likely to encounter both scientific and skeptical ideas, and posters are likely to have their ideas challenged in either case.<\/p>\n Assistant Professor, Department of Administrative Sciences, Metropolitan College<\/p>\t<\/a>\n\n\t\n<\/li>\n\t\t\t\t\t\n Assistant Professor, Department of Electrical and Computer Engineering, College of Engineering<\/p>\t<\/a>\n\n\t\n<\/li>\n\t\t\t\t\t\n Affiliated Faculty, IGS; Boston University Climate Disinformation Initiative Research Theme | How climate disinformation spreads through social media sites: Reddit Research goal: To build a catalog of prominent narratives of disinformation about climate change circulating on Twitter and Reddit; to identify major accounts promoting disinformation and the networks of which they are a part; to model shifting patterns […]<\/p>\n","protected":false},"author":15764,"featured_media":0,"parent":15109,"menu_order":1,"comment_status":"closed","ping_status":"closed","template":"","meta":[],"_links":{"self":[{"href":"https:\/\/www.bu.edu\/igs\/wp-json\/wp\/v2\/pages\/15065"}],"collection":[{"href":"https:\/\/www.bu.edu\/igs\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.bu.edu\/igs\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.bu.edu\/igs\/wp-json\/wp\/v2\/users\/15764"}],"replies":[{"embeddable":true,"href":"https:\/\/www.bu.edu\/igs\/wp-json\/wp\/v2\/comments?post=15065"}],"version-history":[{"count":16,"href":"https:\/\/www.bu.edu\/igs\/wp-json\/wp\/v2\/pages\/15065\/revisions"}],"predecessor-version":[{"id":15156,"href":"https:\/\/www.bu.edu\/igs\/wp-json\/wp\/v2\/pages\/15065\/revisions\/15156"}],"up":[{"embeddable":true,"href":"https:\/\/www.bu.edu\/igs\/wp-json\/wp\/v2\/pages\/15109"}],"wp:attachment":[{"href":"https:\/\/www.bu.edu\/igs\/wp-json\/wp\/v2\/media?parent=15065"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
\nFrom more than 150,000 posts within these communities over more than a decade, researchers built a Structural Topic Model to identify 48 themes of discussion. They trace how these topics have evolved in prevalence over time:<\/p>\n\n
\n![\"\"](\"\/igs\/files\/2023\/05\/Evolution-of-climate-consequences.png\")
<\/li>\n
\n![\"\"](\"\/igs\/files\/2023\/05\/Evolution-of-climate-science-discussion.png\")
<\/li>\n
\n![\"\"](\"\/igs\/files\/2023\/05\/Evolution-of-climate-action_policy-1.png\")
<\/li>\n<\/ol>\n<\/ol>\n
\n
Focus #2: Automatic Identification of Climate Skepticism on Reddit using grounded stance detection models.<\/h4>\n
\n
\n
Research Theme Champions<\/h4>\n\n\t
\n\t\t\t\t\t\n
![\"Headshot](\"\/igs\/files\/2022\/05\/Kim-thumb.jpg\")
<\/figure>\t\t\t\tHyunuk Kim<\/h6>\n\t\t
![\"Headshot](\"\/igs\/files\/2022\/05\/Screenshot-2024-04-08-at-1.37.16\u202fPM-300x300.png\")
<\/figure>\t\t\t\tGianluca Stringhini<\/h6>\n\t\t
![\"Headshot](\"\/igs\/files\/2022\/05\/Screenshot-2024-04-08-at-1.53.59\u202fPM-300x300.png\")
<\/figure>\t\t\t\tIrena Vodenska<\/h6>\n\t\t
\r\nProfessor of Finance, Director of Finance Programs, and Chair of Administrative Sciences, Metropolitan College<\/p>\t<\/a>\n\n\t\n<\/li>\n\t\t\t<\/ul>\n\t\n","protected":false},"excerpt":{"rendered":"