Root cause analysis of software bugs using machine learning techniques

Abstract

Root cause analysis (RCA) is a systematic process for identifying “root causes” of problems or events and an approach for responding to them. The factor that caused a problem or defect should be permanently eliminated through process improvement. In the context of Software development process it may be used to refer to a specific module or a category of bug which in turn can be useful for tackling the problem at its root. In this paper we propose a machine learning approach for finding root cause of a newly filed software bugs which in turn would help in the faster and cleaner resolution of software bugs. This proposed approach is evaluated for feasibility study on an open source system eclipse. [7], [6]