Multilingual Previously Fact-Checked Claim Retrieval
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2023
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916279758094336 |
|---|---|
| author | Pikuliak, Matúš Srba, Ivan Moro, Robert Hromadka, Timo Smolen, Timotej Melisek, Martin Vykopal, Ivan Simko, Jakub Podrouzek, Juraj Bielikova, Maria |
| author_facet | Pikuliak, Matúš Srba, Ivan Moro, Robert Hromadka, Timo Smolen, Timotej Melisek, Martin Vykopal, Ivan Simko, Jakub Podrouzek, Juraj Bielikova, Maria |
| contents | Fact-checkers are often hampered by the sheer amount of online content that needs to be fact-checked. NLP can help them by retrieving already existing fact-checks relevant to the content being investigated. This paper introduces a new multilingual dataset -- MultiClaim -- for previously fact-checked claim retrieval. We collected 28k posts in 27 languages from social media, 206k fact-checks in 39 languages written by professional fact-checkers, as well as 31k connections between these two groups. This is the most extensive and the most linguistically diverse dataset of this kind to date. We evaluated how different unsupervised methods fare on this dataset and its various dimensions. We show that evaluating such a diverse dataset has its complexities and proper care needs to be taken before interpreting the results. We also evaluated a supervised fine-tuning approach, improving upon the unsupervised method significantly. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2305_07991 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Multilingual Previously Fact-Checked Claim Retrieval Pikuliak, Matúš Srba, Ivan Moro, Robert Hromadka, Timo Smolen, Timotej Melisek, Martin Vykopal, Ivan Simko, Jakub Podrouzek, Juraj Bielikova, Maria Computation and Language Fact-checkers are often hampered by the sheer amount of online content that needs to be fact-checked. NLP can help them by retrieving already existing fact-checks relevant to the content being investigated. This paper introduces a new multilingual dataset -- MultiClaim -- for previously fact-checked claim retrieval. We collected 28k posts in 27 languages from social media, 206k fact-checks in 39 languages written by professional fact-checkers, as well as 31k connections between these two groups. This is the most extensive and the most linguistically diverse dataset of this kind to date. We evaluated how different unsupervised methods fare on this dataset and its various dimensions. We show that evaluating such a diverse dataset has its complexities and proper care needs to be taken before interpreting the results. We also evaluated a supervised fine-tuning approach, improving upon the unsupervised method significantly. |
| title | Multilingual Previously Fact-Checked Claim Retrieval |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2305.07991 |