Research Article
An Approach for Multiple Choice Question Answering System
8 downloads
@INPROCEEDINGS{10.1007/978-3-030-92942-8_7, author={Dinh-Huy Vo and Anh-Khoa Do-Vo and Tram-Anh Nguyen-Thi and Huu-Thanh Duong}, title={An Approach for Multiple Choice Question Answering System}, proceedings={Nature of Computation and Communication. 7th EAI International Conference, ICTCC 2021, Virtual Event, October 28--29, 2021, Proceedings}, proceedings_a={ICTCC}, year={2022}, month={1}, keywords={MCQ BERT MLM Multiple choice Language model Unigram Bigram Masked Language Model}, doi={10.1007/978-3-030-92942-8_7} }- Dinh-Huy Vo
Anh-Khoa Do-Vo
Tram-Anh Nguyen-Thi
Huu-Thanh Duong
Year: 2022
An Approach for Multiple Choice Question Answering System
ICTCC
Springer
DOI: 10.1007/978-3-030-92942-8_7
Abstract
Multiple choice question (MCQ) system, which is a form of question answering system, includes a question, a set of choices and the correct answers from these choices. The rapidly growth of linguistics models improves natural language understanding and motivates the automatic systems in natural language processing. This paper proposes an approach based on the Language Model for MCQ system as incomplete questions in English tests. This not only reduces the burden of human experts to teach their students, but also is useful for self-studying the students. We perform many experiments to evaluate the effectiveness and achieve 85.45% accuracy for our proposal.
Copyright © 2021–2025 ICST
