Downloads: 115 | Views: 207 | Weekly Hits: ⮙1 | Monthly Hits: ⮙1
Research Paper | Computer Science & Engineering | India | Volume 3 Issue 6, June 2014 | Rating: 7 / 10
Towards the Roadmap of Defect Predictors and its Diagnostics
Shubha Sharma, Anubhooti Papola
Abstract: Data mining and efficient knowledge discovery in database have been attracting a significant amount of research; industry attention in mostly all the fields. The paper discusses about Bug prediction: an essential phase in the process of bug triaging and fixing process. Bug prediction solutions have been proposed. The demonstration of their effectiveness however faces number of issues including bug triaging; inappropriate evaluation measures: likelihood; precision and recall values. The major drawback of bug prediction technique is unsatisfactory prediction resulting in least accuracy and conspicuous bug report. With the new advancement; machine learning classifiers have proved and emerged as a trusted way to predict the introduction of bugs in changes made to sources. The demerits of existing classifier based prediction techniques are insufficient data and slow performance. The objective is to reach optimal classification performance with nave bayes accompanied with svm (support vector machine) classifiers and substantial feature extraction; further analyzing its performance on the basis of its evaluation measures-likelihood; precision and recall value; f- measure. The evaluation result show that proposed approach can achieve better search results than existing search programs. the work will help in precise and relevant prediction of bug reports and will help in diagnostics of mislabeling of bug reports.
Keywords: bug reports, prediction, search quality, maintenance
Edition: Volume 3 Issue 6, June 2014
Pages: 894 - 899