FICLE (Factual Inconsistency CLassification with Explanation)
Introduced by Raha et al. in Neural models for Factual Inconsistency Classification with Explanations
The FICLE dataset is a derivative of the FEVER dataset, which is a collection of 185,445 claims generated by modifying sentences obtained from Wikipedia. These claims were then verified without knowledge of the original sentences they were derived from. Each sample in the FEVER dataset consists of a claim sentence, a context sentence extracted from a Wikipedia URL as evidence, and a type label indicating whether the claim is supported, refuted, or lacks sufficient information.
Papers
| Paper | Code | Results | Date | Stars |
|---|
