Micro process analysis of maintenance effort: an open source software case study using metrics based on program slicing. Issue 9 (4th October 2012)

• Record Type: Journal Article
• Authors: Kula, Raula Gaikovina; Fushida, Kyohei; Yoshida, Norihiro; Iida, Hajimu
• Is Part Of: Journal of software. Volume 25:Issue 9(2013)
• Page Start: 935
• Page End: 955
• Publication Date: 2012-10-04
• Subjects: software processes -- mining software repositories -- program slicing -- software metrics; Software engineering -- Periodicals; Computer software -- Development -- Periodicals; Software maintenance -- Periodicals; 005.1
• DOI: 10.1002/smr.1572
• Languages: English
• ISSNs: 2047-7473

Abstract

SUMMARY: For any software project, most experts regard the maintenance phase as the most effort and cost intensive of all phases in the software development life cycle. This is due to the high maintenance effort, time, and resources needed to effectively address issues during software maintenance (maintenance activities). Mismanagement of these efforts can lead to the degradation of software maintainability. Understanding the assessment of the related software processes can help sustain or improve maintainability during these maintenance activities. Recent studies have shown that current software process assessments are expensive, generic, and complex, especially for smaller organizations. In this paper, we investigate an alternative software process assessment approach performed by analyzing fine‐grained processes (micro processes) of maintenance activities. This approach assesses maintenance efforts based on micro processes in relation to their impact on source code. The approach derives maintenance effort from the complexity and duration of micro processes and uses proposed metrics based on program slicing to measure change impact. In this paper, we investigate an alternative software process assessment approach by analysing fine‐grained processes (micro processes) of maintenance activities. At statistically significant levels, results suggest that the level of the maintenance efforts correlates with its impact on source code. Copyright © 2012 John Wiley & Sons, Ltd. Abstract : For any software project, most experts regard the maintenance phase as the most effort and cost‐intensive phase in the software development life cycle. In this paper, we investigate an alternative SPI approach by assessing maintenance efforts. Using program slicing‐based metrics, we classify the maintenance effort of micro processes in relation to its impact on source code. Using three open source software case studies, we determined that maintenance efforts correlate with its impact on source code at statistically significant levels.