Presentation | 2021-01-27 Automatic Short Answer Scoring using Thesaurus-Based Data Augmentation Hiroyuki Kato, Tsunenori Ishioka, Tsunenori Mine, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | In the field of natural language processing, the invention of large-scale general-purpose language models such as BERT has brought improvements in processing accuracy for various types of tasks, but the accuracy has not yet reached a practical level, also in automatic short answer scoring, and further improvements are desired. In this paper, we propose a method to improve the accuracy of automatic short answer scoring using a thesaurus to replace words in the answers. In an experiment using data from social studies and Japanese language mock exams for high school students, we found that the accuracy improved in the range with a small number of data for social studies and in the range with a large number of data for Japanese language. In addition, we found that in the case of highly accurate learning, the model understands the data after substitution and the original data in a different manner. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | NLP / AutoScoring / DataAugmentation / Thesaurus / BERT |
Paper # | AI2020-15 |
Date of Issue | 2021-01-20 (AI) |
Conference Information | |
Committee | AI |
---|---|
Conference Date | 2021/1/27(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Online |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | Naoki Fukuta(Shizuoka Univ.) |
Vice Chair | Yuichi Sei(Univ. of Electro-Comm.) / Yuko Sakurai(AIST) |
Secretary | Yuichi Sei(Nagoya Inst. of Tech.) / Yuko Sakurai(Tokyo Univ. of Agriculture and Technology) |
Assistant |
Paper Information | |
Registration To | Technical Committee on Artificial Intelligence and Knowledge-Based Processing |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Automatic Short Answer Scoring using Thesaurus-Based Data Augmentation |
Sub Title (in English) | |
Keyword(1) | NLP |
Keyword(2) | AutoScoring |
Keyword(3) | DataAugmentation |
Keyword(4) | Thesaurus |
Keyword(5) | BERT |
1st Author's Name | Hiroyuki Kato |
1st Author's Affiliation | Kyushu University(Kyushu Univ.) |
2nd Author's Name | Tsunenori Ishioka |
2nd Author's Affiliation | National Center for University Entrance Examinations(DNC) |
3rd Author's Name | Tsunenori Mine |
3rd Author's Affiliation | Kyushu University(Kyushu Univ) |
Date | 2021-01-27 |
Paper # | AI2020-15 |
Volume (vol) | vol.120 |
Number (no) | AI-344 |
Page | pp.pp.7-12(AI), |
#Pages | 6 |
Date of Issue | 2021-01-20 (AI) |