Результаты исследований: Публикации в книгах, отчётах, сборниках, трудах конференций › статья в сборнике материалов конференции › научная › Рецензирование
FactDebug at SemEval-2025 Task 7: Hybrid Retrieval Pipeline for Identifying Previously Fact-Checked Claims Across Multiple Languages. / Nikolaev, Evgenii; Бондаренко, Иван ; Aushev, Islam и др.
Proceedings of the 19th International Workshop on Semantic Evaluation (SemEval-2025). ред. / Sara Rosenthal; Aiala Rosá; Debanjan Ghosh; Marcos Zampieri. Vienna, Austria : Association for Computational Linguistics, 2025. стр. 2190–2196 2025.semeval-1.284.Результаты исследований: Публикации в книгах, отчётах, сборниках, трудах конференций › статья в сборнике материалов конференции › научная › Рецензирование
}
TY - GEN
T1 - FactDebug at SemEval-2025 Task 7: Hybrid Retrieval Pipeline for Identifying Previously Fact-Checked Claims Across Multiple Languages
AU - Nikolaev, Evgenii
AU - Бондаренко, Иван
AU - Aushev, Islam
AU - Krikunov, Vasilii
AU - Glinskii, Andrei
AU - Konovalov, Vasily
AU - Belikova, Julia
N1 - Conference code: 19
PY - 2025/7
Y1 - 2025/7
N2 - The proliferation of multilingual misinformation demands robust systems for crosslingual fact-checked claim retrieval. This paper addresses SemEval-2025 Shared Task 7, which challenges participants to retrieve fact-checks for social media posts across 14 languages, even when posts and fact-checks are in different languages. We propose a hybrid retrieval pipeline that combines sparse lexical matching (BM25, BGE-m3) and dense semantic retrieval (pretrained and fine-tuned BGE-m3) with dynamic fusion and curriculum-trained rerankers. Our system achieves 67.2% crosslingual and 86.01% monolingual accuracy on the Shared Task MultiClaim dataset.
AB - The proliferation of multilingual misinformation demands robust systems for crosslingual fact-checked claim retrieval. This paper addresses SemEval-2025 Shared Task 7, which challenges participants to retrieve fact-checks for social media posts across 14 languages, even when posts and fact-checks are in different languages. We propose a hybrid retrieval pipeline that combines sparse lexical matching (BM25, BGE-m3) and dense semantic retrieval (pretrained and fine-tuned BGE-m3) with dynamic fusion and curriculum-trained rerankers. Our system achieves 67.2% crosslingual and 86.01% monolingual accuracy on the Shared Task MultiClaim dataset.
UR - https://www.scopus.com/pages/publications/105055480562
M3 - Conference contribution
SP - 2190
EP - 2196
BT - Proceedings of the 19th International Workshop on Semantic Evaluation (SemEval-2025)
A2 - Rosenthal, Sara
A2 - Rosá, Aiala
A2 - Ghosh, Debanjan
A2 - Zampieri, Marcos
PB - Association for Computational Linguistics
CY - Vienna, Austria
T2 - The 19th International Workshop on Semantic Evaluation
Y2 - 31 July 2025 through 1 August 2025
ER -
ID: 85312678