Exfakt: A framework for explaining facts over knowledge graphs and text

Mohamed H. Gad-Elrab, Jacopo Urbani, Daria Stepanova, Gerhard Weikum

Research output: Chapter in Book / Report / Conference proceedingConference contributionAcademicpeer-review

Abstract

Fact-checking is a crucial task for accurately populating, updating and curating knowledge graphs. Manually validating candidate facts is time-consuming. Prior work on automating this task focuses on estimating truthfulness using numerical scores which are not human-interpretable. Others extract explicit mentions of the candidate fact in the text as an evidence for the candidate fact, which can be hard to directly spot. In our work, we introduce ExFaKT, a framework focused on generating human-comprehensible explanations for candidate facts. ExFaKT uses background knowledge encoded in the form of Horn clauses to rewrite the fact in question into a set of other easier-to-spot facts. The final output of our framework is a set of semantic traces for the candidate fact from both text and knowledge graphs. The experiments demonstrate that our rewritings significantly increase the recall of fact-spotting while preserving high precision. Moreover, we show that the explanations effectively help humans to perform fact-checking and can also be exploited for automating this task.

LanguageEnglish
Title of host publicationWSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining
PublisherAssociation for Computing Machinery, Inc
Pages87-95
Number of pages9
ISBN (Electronic)9781450359405
DOIs
Publication statusPublished - 30 Jan 2019
Event12th ACM International Conference on Web Search and Data Mining, WSDM 2019 - Melbourne, Australia
Duration: 11 Feb 201915 Feb 2019

Conference

Conference12th ACM International Conference on Web Search and Data Mining, WSDM 2019
CountryAustralia
CityMelbourne
Period11/02/1915/02/19

Fingerprint

Semantics
Experiments

Keywords

  • Explainable evidence
  • Fact-checking
  • Knowledge graph
  • Reasoning

Cite this

Gad-Elrab, M. H., Urbani, J., Stepanova, D., & Weikum, G. (2019). Exfakt: A framework for explaining facts over knowledge graphs and text. In WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining (pp. 87-95). Association for Computing Machinery, Inc. https://doi.org/10.1145/3289600.3290996
Gad-Elrab, Mohamed H. ; Urbani, Jacopo ; Stepanova, Daria ; Weikum, Gerhard. / Exfakt : A framework for explaining facts over knowledge graphs and text. WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining. Association for Computing Machinery, Inc, 2019. pp. 87-95
@inproceedings{04480f82af1e479d9bd4ca76cd49b5e1,
title = "Exfakt: A framework for explaining facts over knowledge graphs and text",
abstract = "Fact-checking is a crucial task for accurately populating, updating and curating knowledge graphs. Manually validating candidate facts is time-consuming. Prior work on automating this task focuses on estimating truthfulness using numerical scores which are not human-interpretable. Others extract explicit mentions of the candidate fact in the text as an evidence for the candidate fact, which can be hard to directly spot. In our work, we introduce ExFaKT, a framework focused on generating human-comprehensible explanations for candidate facts. ExFaKT uses background knowledge encoded in the form of Horn clauses to rewrite the fact in question into a set of other easier-to-spot facts. The final output of our framework is a set of semantic traces for the candidate fact from both text and knowledge graphs. The experiments demonstrate that our rewritings significantly increase the recall of fact-spotting while preserving high precision. Moreover, we show that the explanations effectively help humans to perform fact-checking and can also be exploited for automating this task.",
keywords = "Explainable evidence, Fact-checking, Knowledge graph, Reasoning",
author = "Gad-Elrab, {Mohamed H.} and Jacopo Urbani and Daria Stepanova and Gerhard Weikum",
year = "2019",
month = "1",
day = "30",
doi = "10.1145/3289600.3290996",
language = "English",
pages = "87--95",
booktitle = "WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining",
publisher = "Association for Computing Machinery, Inc",

}

Gad-Elrab, MH, Urbani, J, Stepanova, D & Weikum, G 2019, Exfakt: A framework for explaining facts over knowledge graphs and text. in WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining. Association for Computing Machinery, Inc, pp. 87-95, 12th ACM International Conference on Web Search and Data Mining, WSDM 2019, Melbourne, Australia, 11/02/19. https://doi.org/10.1145/3289600.3290996

Exfakt : A framework for explaining facts over knowledge graphs and text. / Gad-Elrab, Mohamed H.; Urbani, Jacopo; Stepanova, Daria; Weikum, Gerhard.

WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining. Association for Computing Machinery, Inc, 2019. p. 87-95.

Research output: Chapter in Book / Report / Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - Exfakt

T2 - A framework for explaining facts over knowledge graphs and text

AU - Gad-Elrab, Mohamed H.

AU - Urbani, Jacopo

AU - Stepanova, Daria

AU - Weikum, Gerhard

PY - 2019/1/30

Y1 - 2019/1/30

N2 - Fact-checking is a crucial task for accurately populating, updating and curating knowledge graphs. Manually validating candidate facts is time-consuming. Prior work on automating this task focuses on estimating truthfulness using numerical scores which are not human-interpretable. Others extract explicit mentions of the candidate fact in the text as an evidence for the candidate fact, which can be hard to directly spot. In our work, we introduce ExFaKT, a framework focused on generating human-comprehensible explanations for candidate facts. ExFaKT uses background knowledge encoded in the form of Horn clauses to rewrite the fact in question into a set of other easier-to-spot facts. The final output of our framework is a set of semantic traces for the candidate fact from both text and knowledge graphs. The experiments demonstrate that our rewritings significantly increase the recall of fact-spotting while preserving high precision. Moreover, we show that the explanations effectively help humans to perform fact-checking and can also be exploited for automating this task.

AB - Fact-checking is a crucial task for accurately populating, updating and curating knowledge graphs. Manually validating candidate facts is time-consuming. Prior work on automating this task focuses on estimating truthfulness using numerical scores which are not human-interpretable. Others extract explicit mentions of the candidate fact in the text as an evidence for the candidate fact, which can be hard to directly spot. In our work, we introduce ExFaKT, a framework focused on generating human-comprehensible explanations for candidate facts. ExFaKT uses background knowledge encoded in the form of Horn clauses to rewrite the fact in question into a set of other easier-to-spot facts. The final output of our framework is a set of semantic traces for the candidate fact from both text and knowledge graphs. The experiments demonstrate that our rewritings significantly increase the recall of fact-spotting while preserving high precision. Moreover, we show that the explanations effectively help humans to perform fact-checking and can also be exploited for automating this task.

KW - Explainable evidence

KW - Fact-checking

KW - Knowledge graph

KW - Reasoning

UR - http://www.scopus.com/inward/record.url?scp=85061711876&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85061711876&partnerID=8YFLogxK

U2 - 10.1145/3289600.3290996

DO - 10.1145/3289600.3290996

M3 - Conference contribution

SP - 87

EP - 95

BT - WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining

PB - Association for Computing Machinery, Inc

ER -

Gad-Elrab MH, Urbani J, Stepanova D, Weikum G. Exfakt: A framework for explaining facts over knowledge graphs and text. In WSDM 2019 - Proceedings of the 12th ACM International Conference on Web Search and Data Mining. Association for Computing Machinery, Inc. 2019. p. 87-95 https://doi.org/10.1145/3289600.3290996