An improved methodology on information distillation by mining program source code

Abstract

This paper presents a methodology for knowledge acquisition from source code. We use data mining to support semi-automated software maintenance and comprehension and provide practical insights into systems specifics, assuming one has limited prior familiarity with these systems. We propose a methodology and an associated model for extracting information from object oriented code by applying clustering and association rules mining. K-means clustering produces system overviews and deductions, which support further employment of an improved version of MMS Apriori that identifies hidden relationships between classes, methods and member data. The methodology is evaluated on an industrial case study, results are discussed and conclusions are drawn.