Claim-Dissector: An Interpretable Fact-Checking System with Joint Re-ranking and Veracity Prediction

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F23%3APU149736" target="_blank" >RIV/00216305:26230/23:PU149736 - isvavai.cz</a>

Výsledek na webu

<a href="https://aclanthology.org/2023.findings-acl.647/" target="_blank" >https://aclanthology.org/2023.findings-acl.647/</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.18653/v1/2023.findings-acl.647" target="_blank" >10.18653/v1/2023.findings-acl.647</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Claim-Dissector: An Interpretable Fact-Checking System with Joint Re-ranking and Veracity Prediction

Popis výsledku v původním jazyce

We present Claim-Dissector: a novel latent variable model for fact-checking and analysis, which given a claim and a set of retrieved evidence jointly learns to identify: (i) the relevant evidences to the given claim (ii) the veracity of the claim. We propose to disentangle the per-evidence relevance probability and its contribution to the final veracity probability in an interpretable way - the final veracity probability is proportional to a linear ensemble of per-evidence relevance probabilities. In this way, the individual contributions of evidences towards the final predicted probability can be identified. In per-evidence relevance probability, our model can further distinguish whether each relevant evidence is supporting (S) or refuting (R) the claim. This allows to quantify how much the S/R probability contributes to final verdict or to detect disagreeing evidence. Despite its interpretable nature, our system achieves results competetive with state-of-the-art on the FEVER dataset, as compared to typical two-stage system pipelines, while using significantly fewer parameters. Furthermore, our analysis shows that our model can learn fine-grained relevance cues while using coarse-grained supervision and we demonstrate it in 2 ways. (i) We show that our model can achieve competitive sentence recall while using only paragraph-level relevance supervision. (ii) Traversing towards the finest granularity of relevance, we show that our model is capable of identifying relevance at the token level. To do this, we present a new benchmark TLR-FEVER focusing on token-level interpretability - humans annotate tokens in relevant evidences they considered essential when making their judgment. Then we measure how similar are these annotations to the tokens our model is focusing on.

Název v anglickém jazyce

Claim-Dissector: An Interpretable Fact-Checking System with Joint Re-ranking and Veracity Prediction

Popis výsledku anglicky

We present Claim-Dissector: a novel latent variable model for fact-checking and analysis, which given a claim and a set of retrieved evidence jointly learns to identify: (i) the relevant evidences to the given claim (ii) the veracity of the claim. We propose to disentangle the per-evidence relevance probability and its contribution to the final veracity probability in an interpretable way - the final veracity probability is proportional to a linear ensemble of per-evidence relevance probabilities. In this way, the individual contributions of evidences towards the final predicted probability can be identified. In per-evidence relevance probability, our model can further distinguish whether each relevant evidence is supporting (S) or refuting (R) the claim. This allows to quantify how much the S/R probability contributes to final verdict or to detect disagreeing evidence. Despite its interpretable nature, our system achieves results competetive with state-of-the-art on the FEVER dataset, as compared to typical two-stage system pipelines, while using significantly fewer parameters. Furthermore, our analysis shows that our model can learn fine-grained relevance cues while using coarse-grained supervision and we demonstrate it in 2 ways. (i) We show that our model can achieve competitive sentence recall while using only paragraph-level relevance supervision. (ii) Traversing towards the finest granularity of relevance, we show that our model is capable of identifying relevance at the token level. To do this, we present a new benchmark TLR-FEVER focusing on token-level interpretability - humans annotate tokens in relevant evidences they considered essential when making their judgment. Then we measure how similar are these annotations to the tokens our model is focusing on.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/FW03010656" target="_blank" >FW03010656: Multilingvální asistent pro hledání, analýzu a zpracování informací a podporu rozhodování</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2023

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Findings of the Association for Computational Linguistics: ACL 2023

ISBN

978-1-959429-62-3

ISSN

—

e-ISSN

—

Počet stran výsledku

22

Strana od-do

10184-10205

Název nakladatele

Association for Computational Linguistics

Místo vydání

Toronto

Místo konání akce

Toronto

Datum konání akce

9. 7. 2023

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—