Exploring and Analyzing the “Web of Claims”: Characteristics, Biases, and Trends in Fact-Checked Statements

The Research Center's lectures are designed to raise awareness within the community about new topics and technological innovations. 

Designed for a broad audience—students and researchers from all disciplines—these initiatives draw on the multidisciplinary expertise of the MIPS Cluster, partner clusters and organizations, as well as national and international teams.

They are held on the first Friday of every month at 11:30 a.m. and are followed by a buffet, providing an opportunity to continue the discussion.

Contact

BROWSING AND ANALYSING THE WEB OF CLAIMS: CHARACTERISTICS, BIASES, AND EVOLUTION OF FACT-CHECKED CLAIMS

EXPLORING AND ANALYZING THE "WEB OF CLAIMS": CHARACTERISTICS, BIASES, AND EVOLUTION OF FACT-CHECKED STATEMENTS

Konstantin Todorov, LIRMM, Montpellier

How do falsehoods spread on the web? This question, linked to the spread of fake news and biased discourse, has garnered interest across fields ranging from the social sciences to AI. Online discourse—claims and opinions shared online along with their contexts (authors, entities, events)—offers valuable insights. However, today’s knowledge graphs lack information about claims and related discourse, limiting their usefulness for accessing and reusing this data. In this talk, I will introduce ClaimsKG, a knowledge graph of fact-checked claims enriched with metadata such as truth labels, authors, topics, and references. Built from fact-checking portals and structured using a purpose-built data model, ClaimsKG provides ground truth data for analyzing societal debates on the web. I will demonstrate tools developed for ClaimsKG that facilitate claim exploration and statistical analysis, exploring the biases and evolution of fact-checked claims on the web. Finally, I will place a specific emphasis on ongoing work in modeling and tracing science-related discourse, showcasing the importance of this topic and the need for novel models specifically tailored to scientific discourse on online platforms.