Claim2Source at CheckThat! 2026: Improving Multilingual Scientific Claim-Source Retrieval with Verification-based Re-Ranking

2026-07-04

AI Summary

This work addresses the challenge of weak alignment between multilingual scientific claims and their source literature due to discrepancies in language, phrasing, and detail, which hinders cross-lingual provenance tracing. To tackle this, the authors propose a multi-stage retrieval framework: first, structured representations of claims and source documents are constructed through bilingual representation learning enhanced with metadata, followed by initial candidate recall using a language-adapted dense retrieval model; subsequently, a verification-signal-driven re-ranking mechanism refines results by identifying the most supportive documents beyond mere similarity. Evaluated on English, German, and French, the approach achieves an average MRR@5 of 0.7628, ranking first in the CheckThat! 2026 shared task and significantly advancing the accuracy of multilingual scientific claim attribution.

Abstract

Multilingual scientific claim-source retrieval aims to identify the scientific publication supporting a claim shared on social media. This task is challenging because claims often differ from source publications in terms of language, wording, and level of detail, which weakens the connection between claims and their underlying evidence. In this paper, we present our approach for the CheckThat! 2026 Lab Task 1: Source Retrieval for Scientific Web Claims. We propose a multi-stage retrieval framework for multilingual scientific claim-source retrieval that combines structured claim and source representations with progressive candidate refinement. To address multilingual retrieval challenges, the framework employs bilingual claim representations, metadata-enhanced source representations, and language-specific adaptation of dense retrieval models. Building on this setup, a first-stage retriever generates an initial pool of candidate sources, after which similarity-based re-ranking improves the ranking of highly relevant sources and verification-based re-ranking identifies the candidate source that best supports the claim using verification signals. Our approach achieves an average MRR@5 score of 0.7628 across English, German, and French claims, ranking first on the CheckThat! 2026 leaderboard.

Problem

• multilingual
• scientific claim
• source retrieval
• claim-source alignment
• social media

Innovation

• multilingual retrieval
• verification-based re-ranking
• dense retrieval
• claim-source alignment
• metadata-enhanced representation