Automated Question Generating Method Based on Derived Keyphrase Structures from Bloom's Taxonomy

Wijanarko, Bambang Dwi and Heryadi, Yaya and Toba, Hapnes and Budiharto, Widodo (2020) Automated Question Generating Method Based on Derived Keyphrase Structures from Bloom's Taxonomy. ICIC Express Letters, 14 (11). pp. 1059-1067. ISSN 1881-803X

	Text 2020-Automated Question_Lengkap.pdf Download (2323Kb)
	Text 12. Turnitin_Automated Question.pdf Download (2554Kb)

Abstract

Learning systems for higher education have improved significantly in the past decade. A central research topic is how to stimulate learning and improve the quali�ty of questions for learning evaluation by generating relevant questions with objective of learning and cognitive skill levels. This study presents an automated question generating method based on keyphrase to improve learning system in such way that the generated questions comply with both predetermined learning outcomes and Bloom’s taxonomy of hierarchical cognitive skill levels. Unlike the prominent methods, this study proposed the learning materials written in local languages transformed into English which make the method applicable to learning material written in various languages given the language translation to English is available. In addition, the expert knowledge about keyphrase pat�terns in this study is represented in a set of context-free grammar rules making it possible to extend the pattern’s rules to capture various keyphrase structures. The study which used 1,432 sentences from learning material written in Bahasa Indonesia can generate about 60,000 questions. From the experimentation, the average of Bilingual Evaluation Understudy (BLEU) score which is computed from the generated questions is 0.9566. This result indicates that the generated questions are well understood by human asses�sors. From sentiment analysis results, the generated questions achieve 0.73 classified as neutral, and 0.68 classified as objective. The Cohen-Kappa coefficient of 0.34589 means there is a difference between human evaluators in understanding the generated questions.

Item Type:	Article
Contributors:	Contribution Contributors NIDN/NIDK Email UNSPECIFIED Wijanarko, Bambang Dwi UNSPECIFIED UNSPECIFIED UNSPECIFIED Heryadi, Yaya UNSPECIFIED UNSPECIFIED UNSPECIFIED Toba, Hapnes UNSPECIFIED UNSPECIFIED UNSPECIFIED Budiharto, Widodo UNSPECIFIED UNSPECIFIED
Uncontrolled Keywords:	Question generating, Keyphrase extraction, Keyphrase template, Bloom’s taxonomy, Context-free grammar
Subjects:	T Technology > T Technology (General)
Depositing User:	Perpustakaan Maranatha
Date Deposited:	30 Aug 2023 09:02
Last Modified:	30 Aug 2023 09:02
URI:	http://repository.maranatha.edu/id/eprint/32091

Actions (login required)

View Item