{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/csi-a-hybrid-deep-model-for-fake-news","title":"CSI: A Hybrid Deep Model for Fake News Detection","arxiv_id":"1703.06959","date":"2017-03-20","proceeding":null,"authors":["Natali Ruchansky","Sungyong Seo","Yan Liu"],"abstract":"The topic of fake news has drawn attention both from the public and the\nacademic communities. Such misinformation has the potential of affecting public\nopinion, providing an opportunity for malicious parties to manipulate the\noutcomes of public events such as elections. Because such high stakes are at\nplay, automatically detecting fake news is an important, yet challenging\nproblem that is not yet well understood. Nevertheless, there are three\ngenerally agreed upon characteristics of fake news: the text of an article, the\nuser response it receives, and the source users promoting it. Existing work has\nlargely focused on tailoring solutions to one particular characteristic which\nhas limited their success and generality. In this work, we propose a model that\ncombines all three characteristics for a more accurate and automated\nprediction. Specifically, we incorporate the behavior of both parties, users\nand articles, and the group behavior of users who propagate fake news.\nMotivated by the three characteristics, we propose a model called CSI which is\ncomposed of three modules: Capture, Score, and Integrate. The first module is\nbased on the response and text; it uses a Recurrent Neural Network to capture\nthe temporal pattern of user activity on a given article. The second module\nlearns the source characteristic based on the behavior of users, and the two\nare integrated with the third module to classify an article as fake or not.\nExperimental analysis on real-world data demonstrates that CSI achieves higher\naccuracy than existing models, and extracts meaningful latent representations\nof both users and articles.","url_abs":"http://arxiv.org/abs/1703.06959v4","url_pdf":"http://arxiv.org/pdf/1703.06959v4.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"csi-a-hybrid-deep-model-for-fake-news","repo_url":"https://github.com/soorism/CSI-Code","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"csi-a-hybrid-deep-model-for-fake-news","repo_url":"https://github.com/sungyongs/CSI-Code","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"articles","task_name":"Articles"},{"task_slug":"fake-news-detection","task_name":"Fake News Detection"},{"task_slug":"misinformation","task_name":"Misinformation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1703.06959","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}