Leveraging Big Data Analytics for Efficient Bug Localization in Large-Scale Software Projects

Abstract

As software application gets more advanced, the ways of addressing bugs in the codes have also improved in this research paper, the authors focus on how more complex parts of software can be developed while also combining text mining with algorithms to get better bug management systems. This research paper introduces an effective novel technique which combines two branches of computer science to manage bugs, thus providing great warranty while still being efficient it doesn’t degrade the overall performance of the system. Furthermore, by studying the previous techniques and models, we offer an approach based on the hybridisation of the Random Forest method and text mining which agrees with the integration of natural language processing techniques in the software repositories. A few authors have put together a mathematical model to explain integration of language with the parameters provided by a few others such that the change in the parameters can be measured using precision and recall together with f metric model. Results of the study concluded that the focus that the model had on precision was misplaced since the model lost focus on detecting bug instances which meant that certain areas of focus still required improvement. A great contribution to the gap was offered by the analysis of the data generated by the simulation showing difference in code changes which lead to differences in bugs appearing that more features of the model can be added and the model will still perform better. This study suggests that cycle-centric models for text mining hold great potential as they do not limit the approach to only one to be used hence the complexity of bugs is appreciated.

Keywords Bug Localization; Random Forest; Text Mining; Software Engineering; Machine Learning; Natural Language Processing; Feature Engineering; Data Analytics; Software Repositories; Model Evaluation