A Clustering-Based Filtering Method for Similar Source Code Fragment Search

• SHIMARI Kazumasa

  Osaka University

• ISHIO Takashi

  Nara Institute of Science and Technology

• INOUE Katsuro

  Osaka University

Bibliographic Information

Other Title

• 類似ソースコード片検索結果に対するクラスタリング技術を用いたフィルタリング手法

Abstract

We propose a clustering-based filtering method for a similar source code fragment search. This method groups code fragments by clustering and excludes codes with a low probability of correct answers from the output. We confirm the usefulness of the proposed method using an existing dataset of similar code fragments including bugs.

Journal

• 電子情報通信学会論文誌D 情報・システム

  電子情報通信学会論文誌D 情報・システム J103-D (10), 751-753, 2020-10-01

  The Institute of Electronics, Information and Communication Engineers

Keywords

• code search
• clustering
• filtering
• normalized compression distance

Details

• CRID: 1390567172584194688
• NII Article ID: 120006953692
• DOI: 10.14923/transinfj.2020jdl8009
• ISSN: 18810225
• ISSN: 18804535
• HANDLE: 10061/14212
• Web Site: https://naist.repo.nii.ac.jp/records/4262
• Text Lang: ja
• Data Source:
  • JaLC
  • IRDB
  • CiNii Articles
  • KAKEN
• Abstract License Flag: Disallowed