Automatic Question Generation in Education Domain Based on Ontology

Selvia Ferdiana Kusuma, Daniel O. Siahaan, Chastine Fatichah

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

2 Citations (Scopus)

Abstract

Generating questions at various difficulty levels require significant time and are challenging. It requires specific knowledge and skills. The quality of the generated questions would depend on the question maker's ability to relate information and represent it in the form of a question. Thus, it is difficult to maintain the quality consistency of a large set of questions. This study introduces an ontology-based approach for automating the generation of questions to maintain consistency of question quality. Information related to the ontology is broken down into information categories in the format of SPARQL queries. The queries are then converted into questions. Experts were asked to validate the generated questions. Based on our experiments, the accuracy of the generated questions reaches 86%.

Original languageEnglish
Title of host publicationCENIM 2020 - Proceeding
Subtitle of host publicationInternational Conference on Computer Engineering, Network, and Intelligent Multimedia 2020
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages251-256
Number of pages6
ISBN (Electronic)9781728182834
DOIs
Publication statusPublished - 17 Nov 2020
Event2020 International Conference on Computer Engineering, Network, and Intelligent Multimedia, CENIM 2020 - Virtual, Surabaya, Indonesia
Duration: 17 Nov 202018 Nov 2020

Publication series

NameCENIM 2020 - Proceeding: International Conference on Computer Engineering, Network, and Intelligent Multimedia 2020

Conference

Conference2020 International Conference on Computer Engineering, Network, and Intelligent Multimedia, CENIM 2020
Country/TerritoryIndonesia
CityVirtual, Surabaya
Period17/11/2018/11/20

Keywords

  • Ontology
  • Question Generation
  • SPARQL Query

Fingerprint

Dive into the research topics of 'Automatic Question Generation in Education Domain Based on Ontology'. Together they form a unique fingerprint.

Cite this