Cite-worthiness Detection and Scientific Reference Disambiguation for Fighting Scientific Misinformation Online
Abstract
Scientific web claims, seen as science-related claims on social media and in news articles, have been shown to compromise the accuracy of scientific findings. Complex scientific claims are uttered in the form of short, digestible snippets containing ”implicit references” (seen as references to scientific publications where the URLs to the actual studies are never cited). This phenomenon has led to online scientific debates that contain misinformation, controversy, and polarization. Examples include social media discussions about global health pandemics or climate change. As part of the AI4Sci (AI for Science) project, we aim to detect claims from social media which lack explicit references and to retrieve their original scientific publications. In this abstract, we briefly summarize the project’s context and ongoing contributions.
Our work has introduced novel problem formalizations, ground-truth corpora, and baseline models and algorithms for fighting scientific misinformation online. We also present remaining challenges to be explored in future work.
Auteur(s) : Salim Hafid, Sandra Bringay and Konstantin Todorov
Lien vers l'article :