Paper Title
C4.5 Classifier for Bug Triage with Feature Selection using Kruskal Algorithm
Abstract
It is mandatory to find out the bug in the software application while developing the particular software product.
Software companies are spending a huge amount of money nowadays for handling software bugs. A compulsory phase of
solving bugs is Bug Triage, which appropriately allocates developer to a new bug. Text categorization methods are put into
use to perform automatic bug triage which reduces the time and cost required in manual task. Data reduction problem have
arises due to large availability of data. Various solutions have been provided to diminish the large size of data and enhance
the efficiency of bug data. In existing system, prediction order of applying data reduction techniques has been done utilizing
Naive Bayes whereas feature selection is done using CH method. In existing system the accuracy of the result generated is
low and the time taken by Naïve bayes is more. Kruskal algorithm is used for feature selection to solve the problem of result
accuracy in existing system by generating a minimum spanning tree. Whereas the time taken by Naive Bayes is reduced by
utilizing C4.5 algorithm which is based on decision tree generation. The proposed system efficiently decreases the large data
and helps in enhancing the quality of bug triage.
Index Terms—Bug, Bug triage, Kruskal Method, Feature selection, C4.5.