タイトル(別表記) Proposing an Unsupervised Approach to Evaluate Essays Using IDF on Reference Data
フルテキストURL NLC2018-33.pdf
著者 大野 雅幸| 竹内 孔一| 泉仁 宏太| 小畑 友也| 田口 雅弘| 稲田 佳彦| 飯塚 誠也| 阿保 達彦| 上田 均|
抄録(別表記) In this paper, we describe an on-going study of developing an automatic essay-scoring system in Japanese. Essay scoring systems have already been developed and used mainly in English, while not many previous studies have been done on Japanese essay evaluations. Most of the methods and systems of automatic essay evaluation need not small number of previously human-graded essays for calibrating the parameter of regression functions or parameter of machine learning. The previous studies show the high performance for essay evaluation task, however, it must be not easy to assume large graded essays in, for example, actual tests or entrance examinations. Thus, we take a approach to evaluate Japanese essays without previously human-graded essays but with assuming reference data related to essay questions. The proposed method is a simple one, that is, evaluating the essays with co-occurrences with the reference data in their words or morphemes. In the method technical terms would be given high scores using neologd dictionary and idf values. Experimental results show that the proposed method works well in our developing Japanese mock trial writing tests. Key words automatic scoring of essays, human annotation, supporting
キーワード 自動採点 (automatic scoring of essays) アノテーション (human annotation) 採点支援 (supporting system of essay evaluation) idf neologd
発行日 2018-12
出版物タイトル 電子情報通信学会技術研究報告. NLC, 言語理解とコミュニケーション
118巻
355号
出版者 電子情報通信学会
開始ページ 103
終了ページ 108
ISSN 09135685
NCID AA11524632
資料タイプ 学術雑誌論文
言語 Japanese
OAI-PMH Set 岡山大学
著作権者 copyright@2018
論文のバージョン publisher
タイトル(別表記) A Study of Extracting Related Documents for Essay Evaluation Modules
フルテキストURL NLC2017-24.pdf
著者 泉仁 宏太| 竹内 孔一| 大野 雅幸| 田口 雅弘| 稲田 佳彦| 飯塚 誠也| 阿保 達彦| 上田 均|
抄録(別表記) We are developing an automatic Japanese essay-scoring system that is composed of 4 evaluation criteria, comprehensiveness, logical consistency, validity, spelling and grammar. In this paper, we discuss the most powerful approach to extract documents of Wikipedia that relates to the reference texts of the target essay theme for validity evaluation. The reason for using Wikipedia documents for evaluating validity of students’essays is that we assume that validity can be evaluated by the expanded discussions in Wikipedia documents that relates to the essay theme. Experimental results show that the skip-gram based word vector is the best approach to extract relating documents to reference texts among several keyword-based evaluation approaches.
キーワード 小論文の自動採点 (Automatic scoring of answers of essay-writing tests) 単語ベクトル (Word vector) Skip-gram Wikipedia
発行日 2017-09
出版物タイトル 電子情報通信学会技術研究報告. NLC, 言語理解とコミュニケーション
117巻
207号
出版者 電子情報通信学会
開始ページ 47
終了ページ 51
ISSN 09135685
NCID AA11524632
資料タイプ 学術雑誌論文
言語 Japanese
OAI-PMH Set 岡山大学
著作権者 copyright©2017 IEICE
論文のバージョン publisher