COSPA: Identifying Key Classes in Object-Oriented Software Using Preference Aggregation

Abstract

Software comprehension helps us to maintain software better. The developers often use documentation and mutual communication to understand software. However, the incompletion of documentation usually causes some problems such that we cannot acquire the detailed information of software. So researchers are now using key classes as starting points for software understanding and propose several effective methods for key class identification. Such methods are mostly implemented using supervised and unsupervised learning methods; however, they may be subject to class imbalance, insufficient amount of data, and lack of accuracy. In this paper, we propose a novel approach, called COSPA (identifying key Classes in Object-Oriented Software using Preference Aggregation), to identify key classes in software. Our approach is inspired by the social choice theory and uses Kemeny-Young method to find a preference that minimizes the total difference between different pairs of sequences returned by three key class identification methods (i.e., a-index, InDeg, and OSE). Further, we rank the classes in descending order according to that preference set, and a small number of top-ranked classes are considered as key candidates. To better evaluate the performance of our approach, we apply it to six Java software systems and compare it with nine state-ofthe-art approaches. In the experiments, we use Friedman to evaluate all methods in the whole set of research subjects; results show that, though no single method works always best in all subject systems, our method indeed performs better than others.INDEX TERMS Complex network, key classes, divergence, software comprehension.

[Retracted] A Novel Approach to Automate Complex Software Modularization Using a Fact Extraction System

“…ey came to the conclusion that identifiers provide good design results that the file has a good expert comparison, and those comments have a good expert comparison, but that bad design results and routine calls perform poorly. e authors in [9][10][11][12] proposed a method for improving the accuracy of autonomous software architecture reconstruction.…”

Semantic Code Graph—An Information Model to Facilitate Software Comprehension

“…For object-oriented programming languages, software networks can consist of multiple levels of networks, such as file-level, package-level, class-level, or function-level collaboration networks [29]. The Class Collaboration Network (CCN) is particularly interesting to analyze [10], [32], [37], [38] because classes are the main building blocks in OOP. Arora and Goel [24], [27] proposed JavaRelationshipGraphs (JRG) -a software network model specialized for the Java language.…”

“…Software networks are extensively used to facilitate software comprehension and maintenance through software architecture recovery [13], [25], finding key entities [10], [38], [62], [63], code modularization [60], [64], [65], finding code similarities [66], code visualization [67]- [69], and tracking code evolution [70].…”

“…Regarding research on models for code dependency structure, there exists a body of work demonstrating the utility of software network analysis on various theoretical models [10], [11], [13], [22]- [33]. However, even in the case of widely researched models, such as Call Graph (CG) [11], [12], [34]- [36] or Class Collaboration Network (CCN) [10], [32], [37], [38], the underlying implementations, developed by the researchers to extract and analyze these software models, are rarely published. Consequently, it is very difficult to reproduce the results or pursue further research in the code dependency comprehension field, especially leveraging new or commercial projects.…”

Graph Buddy — an interactive code dependency browsing and visualization tool

“…The most important central code entities are typically good starting points for reading the source code in the process of software comprehension [5]. The scg-cli automatically finds the most important nodes in the SCG graph based on the set of nine different metrics [6], including LOC (line of codes associated with a code entity), out-degree and in-degree, eigenvector centrality, Katz centrality, PageRank, betweenness centrality and Harmonic Centrality.…”