BUG SEVERITY PREDICTION

Abstract

Assessment of the degree of deformities is basic so as to designate assets for testing and to plan inquire about exercises viably. We use content characterization methods in this paper to foresee and assess the seriousness of deformities. The outcomes depend on the deformity portrayal of Nasa venture issue necessities. We utilized the philosophy of Support Vector Machine to evaluate the recurrence of the issue pro-acclamations. In this examination study, a basic word lexicon approach is recommended to decide the earnestness of the bug as outrageous or nongenuine. It is discovered that the example of exactness and accuracy is generally a similar utilizing various methodologies of highlight choice and characterization. In any case, for each of the four parts, Chi square test and KNN classifier give most extreme exactness and precision effectiveness. The proposed work will assist Triage with identifying seriousness based bugs and designate these bugs to explicit engineers. The paper introduces another and robotized technique called Severity Problem Assessment that helps the test engineer in appointing levels of seriousness to reports of imperfections. Severity depends on conventional content mining and AI strategies for existing assortments of records of deformities. The paper gives a contextual analysis on the utilization of Severity and ITS(issue tracking system) of Nasa. The discoveries of the contextual investigation show that Severity is a decent indicator of the seriousness of the issue, while it is easy to utilize and powerful.