We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CL

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computation and Language

Title: CLIMATE-FEVER: A Dataset for Verification of Real-World Climate Claims

Abstract: We introduce CLIMATE-FEVER, a new publicly available dataset for verification of climate change-related claims. By providing a dataset for the research community, we aim to facilitate and encourage work on improving algorithms for retrieving evidential support for climate-specific claims, addressing the underlying language understanding challenges, and ultimately help alleviate the impact of misinformation on climate change. We adapt the methodology of FEVER [1], the largest dataset of artificially designed claims, to real-life claims collected from the Internet. While during this process, we could rely on the expertise of renowned climate scientists, it turned out to be no easy task. We discuss the surprising, subtle complexity of modeling real-world climate-related claims within the \textsc{fever} framework, which we believe provides a valuable challenge for general natural language understanding. We hope that our work will mark the beginning of a new exciting long-term joint effort by the climate science and AI community.
Comments: Accepted for the Tackling Climate Change with Machine Learning Workshop at NeurIPS 2020
Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
Cite as: arXiv:2012.00614 [cs.CL]
  (or arXiv:2012.00614v2 [cs.CL] for this version)

Submission history

From: Markus Leippold [view email]
[v1] Tue, 1 Dec 2020 16:32:54 GMT (2803kb,D)
[v2] Sat, 2 Jan 2021 16:07:48 GMT (2803kb,D)

Link back to: arXiv, form interface, contact.