Multilingual Evidence Retrieval and Fact Verification to Combat Global Disinformation: The Power of Polyglotism

TL;DR

This paper explores multilingual evidence retrieval and fact verification to combat disinformation across languages, demonstrating transfer learning capabilities and providing a new dataset for cross-lingual evaluation.

Contribution

It introduces a novel multilingual verification system and a mixed-language dataset, addressing the challenge of verifying claims in evidence-poor languages.

Findings

EnmBERT shows transfer learning ability in multilingual verification.

A 400-example mixed English-Romanian dataset is created for evaluation.

Multilingual systems can verify claims in disinformation-prone languages.

Abstract

This article investigates multilingual evidence retrieval and fact verification as a step to combat global disinformation, a first effort of this kind, to the best of our knowledge. The goal is building multilingual systems that retrieve in evidence-rich languages to verify claims in evidence-poor languages that are more commonly targeted by disinformation. To this end, our EnmBERT fact verification system shows evidence of transfer learning ability and 400 example mixed English-Romanian dataset is made available for cross-lingual transfer learning evaluation.