Bug Localization Using Multi-objective Approach and Information Retrieval
Detecting software bugs is considered to be an active area of research as bugs detected after the delivery of software is considered to be very expensive to deal with. Whenever a new software bug is detected, software developer faces extreme difficulty in
- PDF / 529,831 Bytes
- 15 Pages / 439.37 x 666.142 pts Page_size
- 29 Downloads / 188 Views
Abstract Detecting software bugs is considered to be an active area of research as bugs detected after the delivery of software is considered to be very expensive to deal with. Whenever a new software bug is detected, software developer faces extreme difficulty in detecting the exact location of the bug in the product. Identifying bugs efficiently through genetic algorithms is the active area of research nowadays. Based on the bug reports dataset, a self-operating genetic algorithm, Strength Pareto Evolutionary Algorithm (SPEA II) identifies and ranks the application files in the reference code as per their likelihood of carrying bugs. In the presented exposition, a textmining strategy, term frequency–inverse document frequency (TFIDF) is applied for the proper ranking of application files. The ranking is based on the history-based similarity and lexical similarity within the generated report of bugs and the documentation of the application program interface (API). The striking feature of the paper is that a multi-objective approach is used to improve upon the conventional techniques such that contradictory demands of increasing the similarity index and A. Sood (B) · S. Sharma · A. Khanna · A. Tiwari · D. Gupta Maharaja Agrasen Institute of Technology, New Delhi, India e-mail: [email protected] S. Sharma e-mail: [email protected] A. Khanna e-mail: [email protected] A. Tiwari e-mail: [email protected] D. Gupta e-mail: [email protected] V. Madan Delhi Technological University, New Delhi, India e-mail: [email protected] S. Doss Botho University, Botswana, Botswana e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2021 D. Gupta et al. (eds.), International Conference on Innovative Computing and Communications, Advances in Intelligent Systems and Computing 1165, https://doi.org/10.1007/978-981-15-5113-0_58
709
710
A. Sood et al.
minimizing the number of suggested classes are met simultaneously. Thus, the proposed new approach employs SPEA II as the multi-objective genetic algorithm to meet the conflicting demands and uses both history-based as well as lexical similarity for information retrieval and ranking. Keywords Fault localization · Multi-objective optimization · Bugs · Learning to rank · Information retrieval
1 Introduction Debugging and correction of software defects consume plenty of time. Also, often defects are reported in the bug reports by various stakeholders that include end-users, testers, and developers. These reports describe the circumstances of the deviation in the expected behavior of the software. Furthermore, they contain extra information regarding the listed bugs and faults. The task of the developer is to locate the files that have the probability of accommodating the bug, narrow down the files that are more probable of being faulty by investigating the bug and finally to fix them efficiently. This is the main challenge because often developers end up examining a lot of application files so as to be able to map the rele
Data Loading...
