@article{oai:ipsj.ixsq.nii.ac.jp:00146540, author = {岩本, 舞 and 中村, 真人 and 小島, 俊輔 and 中嶋, 卓雄 and Mai, Iwamoto and Masato, Nakamura and Shunsuke, Oshima and Takuo, Nakashima}, issue = {4}, journal = {情報処理学会論文誌教育とコンピュータ(TCE)}, month = {Dec}, note = {プログラミングの学習では,学生が様々な課題に取り組むことが重要である.しかしながら,教員が課題を採点するのに時間がかかるため,多くの課題を課することは難しい.そこで,本研究では,学生に多くの課題を解いてもらうために,オンラインジャッジシステムを開発した.またこのようなシステムでは,他人のソースコードを流用し提出する行為が問題となる.そのため,オンラインジャッジに不正コピーを検出する機能を実装した.以前の筆者らの研究では,トークン列長を判定基準とする手法を提案したが,単純な長さをしきい値としており,誤検知が多かった.そこで,本稿ではトークン列の複雑度と完全トークン列を用いた検出手法を提案する.実験の結果,学生が提出した3種類の課題プログラムにおいて,適合率Pが大幅に上昇した.また,本システムを使用した学生にアンケートを実施したところ,89%の学生が,プログラムを書けるようになったと回答した., Learning computer programming requires students to deal with different variations of assignments. Instructors, on the other hand, have difficulty giving students sufficient volume of assignments due to time constraints assessing students' work, which is highly time-consuming. This study presents an online judge system to check the students' assignments so that students will have more opportunities to tackle different assignments. A system of this kind, however, can be problematic in that students may copy other source codes and submit it as their own work. We implemented a function to detect illicit copies into the system. In previous researches of ours, such methods based on the detection standard of the token length have been proposed, with many misdetections occurred. This paper proposes the detection method using the token sequence complexity and the complete token sequence. As the results of experiments, our new method was successful in improving the accuracy rate on three different kinds of assignments submitted by students. Moreover, we sent out questionnaires to students about the system, to be reported from 89% of them out of the entire respondents that they became capable of programming after using the new system.}, pages = {38--47}, title = {不正コピー検出手法を備えたオンラインジャッジシステムの開発}, volume = {1}, year = {2015} }