A Large-Scale Study Of Fault Feature Extraction From Github Repository Using Data Science Techniques

Abstract

Nowadays data play a vital role in all the fields. Machine learning (ML) is a Data Science technique in which the past dataset can be used for future prediction. Fault is the result of the occurrence of an unexpected value in the place of expected value. Fault occurs in all fields, but the severity makes the difference in the industry. The severity of the fault is measured by its priority in the software industry. The unrectified fault may cause software failure. In this paper, thethree major development application-oriented software like website, mobile, and gaming are considered.Lots of programming languages areused for creation of these software. These applications are developed by all categories of projects like short, medium, and long in the software industry. This research works proved the necessity of fault taxonomy dataset by calculating rework range in this software creation. The fault feature data for this research work is collected from the GitHub public repository, which contains the language commit details in all categories of software industry.This work also helps in feature extraction by web scraping method to identify the rework in software development.