Paper Details

In the managing software bugs software groups spend some amount of cost. An inevitable step of solvingbugs is Trojan horse triage, which goals to correctly assign a developer to a brand new worm. To reduce the time valuein guide work, textual content type strategies are implemented to conduct automatic computer virus triage. On this task,we cope with the trouble of statistics discount for malicious program triage, i.e., the way to lessen the dimensions andimprove the exceptional of worm statistics. In the proposed work integrate instance selection with feature choice tosimultaneously reduce statistics scale at the worm measurement and the phrase dimension. To determine the order ofapplying example selection and characteristic choice, we extract attributes from historic worm information sets andbuild a predictive model for a new bug records set. We empirically investigate the performance of statistics discount oncompletely 600,000 computer virus reviews of big open supply initiatives, particularly Eclipse and Mozilla. Theconsequences display that our statistics reduction can correctly lessen the statistics scale and enhance the accuracy ofmalicious program triage. This challenge presents an approach to leveraging techniques on data processing to shapedecreased and awesome bug records in software program improvement and upkeep

Shubham Shankar Gadge, Rajratn Keshav Gaikwad, Yogesh Madhukar Jadhav, Prof. Geetika Narang, “Bug Triage with Data Reduction Techniques”, International Journal of Advance Engineering and Research Development (IJAERD), Vol. 3, Issue 1, pp. 194-198, January 2016.