Big data analysis platform for software reliability assessment

2017 Fiscal Year Final Research Report

• Project/Area Number: 15K01192
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Social systems engineering/Safety system
• Research Institution: Hiroshima University
• Principal Investigator: Tadashi Dohi 広島大学, 工学研究科, 教授 (00243600)
• Project Period (FY): 2015-04-01 – 2018-03-31
• Keywords: ソフトウェア信頼性 / ソフトウェアメトリクス / バグ予測 / リポジトリ / データマイニング / 信頼性評価 / ビッグデータ / 確率モデリング

Outline of Final Research Achievements

In the modern software development it is common to record the meta data observed in the development process, such as design information, process information, source code, etc., in the database called repository, and to utilize them for development and maintenance activities through the development project. It is known that as the development size increases the capacity of repository should be increasing. Then the data mining technologies for the repository data are becoming important. In this project, we developed a new fundamental technology to assess software reliability by utilizing the repository data. Especially, we analyzed several open source project data and developed a methodology and tool to quantify the software reliability measures.

Free Research Field

信頼性工学