Time-aware evidence ranking for fact-checking
Research output: Contribution to journal › Journal article › Research › peer-review
Standard
Time-aware evidence ranking for fact-checking. / Allein, Liesbeth; Augenstein, Isabelle; Moens, Marie Francine.
In: Journal of Web Semantics, Vol. 71, 100663, 2021.Research output: Contribution to journal › Journal article › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - JOUR
T1 - Time-aware evidence ranking for fact-checking
AU - Allein, Liesbeth
AU - Augenstein, Isabelle
AU - Moens, Marie Francine
N1 - Publisher Copyright: © 2021 The Author(s)
PY - 2021
Y1 - 2021
N2 - Truth can vary over time. Fact-checking decisions on claim veracity should therefore take into account temporal information of both the claim and supporting or refuting evidence. In this work, we investigate the hypothesis that the timestamp of a Web page is crucial to how it should be ranked for a given claim. We delineate four temporal ranking methods that constrain evidence ranking differently and simulate hypothesis-specific evidence rankings given the evidence timestamps as gold standard. Evidence ranking in three fact-checking models is ultimately optimized using a learning-to-rank loss function. Our study reveals that time-aware evidence ranking not only surpasses relevance assumptions based purely on semantic similarity or position in a search results list, but also improves veracity predictions of time-sensitive claims in particular.
AB - Truth can vary over time. Fact-checking decisions on claim veracity should therefore take into account temporal information of both the claim and supporting or refuting evidence. In this work, we investigate the hypothesis that the timestamp of a Web page is crucial to how it should be ranked for a given claim. We delineate four temporal ranking methods that constrain evidence ranking differently and simulate hypothesis-specific evidence rankings given the evidence timestamps as gold standard. Evidence ranking in three fact-checking models is ultimately optimized using a learning-to-rank loss function. Our study reveals that time-aware evidence ranking not only surpasses relevance assumptions based purely on semantic similarity or position in a search results list, but also improves veracity predictions of time-sensitive claims in particular.
KW - Automated fact-checking
KW - Document ranking
KW - Learning to rank
KW - Temporal relevance
KW - Temporal semantics
UR - http://www.scopus.com/inward/record.url?scp=85117218307&partnerID=8YFLogxK
U2 - 10.1016/j.websem.2021.100663
DO - 10.1016/j.websem.2021.100663
M3 - Journal article
AN - SCOPUS:85117218307
VL - 71
JO - Web Semantics
JF - Web Semantics
SN - 1570-8268
M1 - 100663
ER -
ID: 284630783