BROWSE
Detail
Conference Paper
A Study on Performance Analysis of Question Generation based on Korean Pretrained Language Model
Cited 0 time in 
Share 
Share
- Authors
- HongYeon Yu, Jiwon Yang, Seunghun Oh, Donghoon Son, Aram Lee, Jeongeun Kim
- Issue Date
- 2023-10
- Citation
- International Conference on Information and Communication Technology Convergence (ICTC) 2023, pp.1239-1241
- Publisher
- IEEE
- Language
- English
- Type
- Conference Paper
- Abstract
- This paper presents the results of a performance analysis of the question generation task based on a Korean pretrained language model for assessing elementary school students’ reading comprehension using literary texts. The Korean pretrained language model is based on SKT-KoBart and employs transfer learning using literary texts of narrative and expository structures, along with the KorQuAd dataset for question-answering tasks. Through the extraction of factual and inferential assessment items, BLEU scores are measured based on the type of literary text, and the performance analysis results of the downstream task of question generation are presented.
- KSP Keywords
- Language model, Literary text, Performance analysis, Question generation, Reading comprehension, Transfer learning, elementary school students, question answering