An automated tool for collection of code attributes for cross project defect prediction
Ruchika Malhotra, Bhavya Bansal, Chitranshi Jain, Ekta Punia · 2017
This paper presents a tool that automates the process of data collection for defect or change analysis. Prediction of defects in early phases has become crucial to reduce the efforts and costs incurred due to defects. This tool extracts the information from Git Version Control System of open source projects. Two consecutive versions of one single project have been used by the tool to obtain results. The tool generates a matrix containing code churn (added lines, deleted lines, modified lines, total LoC), complexity, pre-release bugs and post-release bugs of each file of source code. The obtained software metrics can be used to measure the development process of a software and therefore in analysis and prediction purposes.