Predicate constraints based question answering over knowledge graph

Sangjin Shin, Xiongnan Jin, Jooik Jung, Kyong Ho Lee

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Generally, QA systems suffer from the structural difference where a question is composed of unstructured data, while its answer is made up of structured data in a Knowledge Graph (KG). To bridge this gap, most approaches use lexicons to cover data that are represented differently. However, the existing lexicons merely deal with representations for entity and relation mentions rather than consulting the comprehensive meaning of the question. To resolve this, we design a novel predicate constraints lexicon which restricts subject and object types for a predicate. It facilitates a comprehensive validation of a subject, predicate and object simultaneously. In this paper, we propose Predicate Constraints based Question Answering (PCQA). Our method prunes inappropriate entity/relation matchings to reduce search space, thus leading to an improvement of accuracy. Unlike the existing QA systems, we do not use any templates but generates query graphs to cover diverse types of questions. In query graph generation, we put more focus on matching relations rather than linking entities. This is well-suited to the use of predicate constraints. Our experimental results prove the validity of our approach and demonstrate a reasonable performance compared to other methods which target WebQuestions and Free917 benchmarks.

Original languageEnglish
Pages (from-to)445-462
Number of pages18
JournalInformation Processing and Management
Volume56
Issue number3
DOIs
Publication statusPublished - 2019 May 1

Fingerprint

management counsulting
performance
Question answering
Graph
Query
Benchmark
Consulting
Template

All Science Journal Classification (ASJC) codes

  • Information Systems
  • Media Technology
  • Computer Science Applications
  • Management Science and Operations Research
  • Library and Information Sciences

Cite this

Shin, Sangjin ; Jin, Xiongnan ; Jung, Jooik ; Lee, Kyong Ho. / Predicate constraints based question answering over knowledge graph. In: Information Processing and Management. 2019 ; Vol. 56, No. 3. pp. 445-462.
@article{ebc38bbf22214f3286cf724737aaff96,
title = "Predicate constraints based question answering over knowledge graph",
abstract = "Generally, QA systems suffer from the structural difference where a question is composed of unstructured data, while its answer is made up of structured data in a Knowledge Graph (KG). To bridge this gap, most approaches use lexicons to cover data that are represented differently. However, the existing lexicons merely deal with representations for entity and relation mentions rather than consulting the comprehensive meaning of the question. To resolve this, we design a novel predicate constraints lexicon which restricts subject and object types for a predicate. It facilitates a comprehensive validation of a subject, predicate and object simultaneously. In this paper, we propose Predicate Constraints based Question Answering (PCQA). Our method prunes inappropriate entity/relation matchings to reduce search space, thus leading to an improvement of accuracy. Unlike the existing QA systems, we do not use any templates but generates query graphs to cover diverse types of questions. In query graph generation, we put more focus on matching relations rather than linking entities. This is well-suited to the use of predicate constraints. Our experimental results prove the validity of our approach and demonstrate a reasonable performance compared to other methods which target WebQuestions and Free917 benchmarks.",
author = "Sangjin Shin and Xiongnan Jin and Jooik Jung and Lee, {Kyong Ho}",
year = "2019",
month = "5",
day = "1",
doi = "10.1016/j.ipm.2018.12.003",
language = "English",
volume = "56",
pages = "445--462",
journal = "Information Processing and Management",
issn = "0306-4573",
publisher = "Elsevier Limited",
number = "3",

}

Predicate constraints based question answering over knowledge graph. / Shin, Sangjin; Jin, Xiongnan; Jung, Jooik; Lee, Kyong Ho.

In: Information Processing and Management, Vol. 56, No. 3, 01.05.2019, p. 445-462.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Predicate constraints based question answering over knowledge graph

AU - Shin, Sangjin

AU - Jin, Xiongnan

AU - Jung, Jooik

AU - Lee, Kyong Ho

PY - 2019/5/1

Y1 - 2019/5/1

N2 - Generally, QA systems suffer from the structural difference where a question is composed of unstructured data, while its answer is made up of structured data in a Knowledge Graph (KG). To bridge this gap, most approaches use lexicons to cover data that are represented differently. However, the existing lexicons merely deal with representations for entity and relation mentions rather than consulting the comprehensive meaning of the question. To resolve this, we design a novel predicate constraints lexicon which restricts subject and object types for a predicate. It facilitates a comprehensive validation of a subject, predicate and object simultaneously. In this paper, we propose Predicate Constraints based Question Answering (PCQA). Our method prunes inappropriate entity/relation matchings to reduce search space, thus leading to an improvement of accuracy. Unlike the existing QA systems, we do not use any templates but generates query graphs to cover diverse types of questions. In query graph generation, we put more focus on matching relations rather than linking entities. This is well-suited to the use of predicate constraints. Our experimental results prove the validity of our approach and demonstrate a reasonable performance compared to other methods which target WebQuestions and Free917 benchmarks.

AB - Generally, QA systems suffer from the structural difference where a question is composed of unstructured data, while its answer is made up of structured data in a Knowledge Graph (KG). To bridge this gap, most approaches use lexicons to cover data that are represented differently. However, the existing lexicons merely deal with representations for entity and relation mentions rather than consulting the comprehensive meaning of the question. To resolve this, we design a novel predicate constraints lexicon which restricts subject and object types for a predicate. It facilitates a comprehensive validation of a subject, predicate and object simultaneously. In this paper, we propose Predicate Constraints based Question Answering (PCQA). Our method prunes inappropriate entity/relation matchings to reduce search space, thus leading to an improvement of accuracy. Unlike the existing QA systems, we do not use any templates but generates query graphs to cover diverse types of questions. In query graph generation, we put more focus on matching relations rather than linking entities. This is well-suited to the use of predicate constraints. Our experimental results prove the validity of our approach and demonstrate a reasonable performance compared to other methods which target WebQuestions and Free917 benchmarks.

UR - http://www.scopus.com/inward/record.url?scp=85059611094&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85059611094&partnerID=8YFLogxK

U2 - 10.1016/j.ipm.2018.12.003

DO - 10.1016/j.ipm.2018.12.003

M3 - Article

AN - SCOPUS:85059611094

VL - 56

SP - 445

EP - 462

JO - Information Processing and Management

JF - Information Processing and Management

SN - 0306-4573

IS - 3

ER -