Research on a Multi-Level Combination of Software Defect Localization Method

Abstract

With the increasing scale and complexity of software, the probability of containing defects in the software is also increasing. Software defect localization is the key to reduce and remove software defects. The timeliness and effectiveness of defects localization directly affects the usability of the software. There are some problems in the traditional method of defects localization, such as the low accuracy of correlation defects localization and the inaccuracy of single defects distribution. In this paper, by comparing and analyzing the advantages and disadvantages of different types of localization methods, we improve the existing predicative localization method and track localization method based on a spectrum, and a multi-level software defect localization method(MCDL) is proposed by combining them organically, which effectively makes up for each other's shortcomings. This paper also proposes a new single vulnerability suspicion rate (SVSR) algorithm. Experimental results show that the proposed method is faster and more accurate than traditional methods.