Title: Extracting a knowledge from source code comprehesion using data mining methods

Authors: Ashutosh Mishra; Vinayak Srivastava

Addresses: Computer Engineering Department, Institute of Technology, BHU, Varanasi-221 005, Uttar Pradesh, India ' Computer Engineering Department, Institute of Technology, BHU, Varanasi-221 005, Uttar Pradesh, India

Abstract: In software maintenance, source code comprehension is a very vital task. The comprehension of the source code is performed by different tools for various purposes. Data mining is one of the important and versatile methods in this context. Data mining methods and tools have been widely used in software engineering in general and software maintenance in particular. We present in this paper a methodology to extract knowledge using data mining methods which would be very much useful for software maintenance. The data mining methods for clustering, classification and association rules have been deployed for source code comprehension. Our approach is holistic in nature that covers many aspects required for software maintenance whereas approaches by other researchers cover a partial aspect in this context. We have made a qualitative comparison of our approach with others and have derived the conclusion on the basis.

Keywords: source code comprehension; software maintenance; data mining; clustering; cohesion; coupling; knowledge retrieval; knowledge extraction; software engineering; classification; association rules mining.

DOI: 10.1504/IJKEDM.2012.051240

International Journal of Knowledge Engineering and Data Mining, 2012 Vol.2 No.2/3, pp.174 - 199

Received: 08 May 2021
Accepted: 12 May 2021

Published online: 28 Dec 2012 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article