Approaches for Categorization of Reusable Software Components

Abstract

Reuse repositories manager manages the reusable software components in different categories and needs to find the category of reusable software components. In this paper, we have used different pure and hybrid approaches to find the domain relevancy of the component to a particular domain. Probabilistic Latent Semantic Analysis (PLSA) approach, LSA, Singular Value Decomposition (SVD) technique, LSA Semi-Discrete Matrix Decomposition (SDD) technique and Naive Bayes Approach purely as well as hybrid, are evaluated to determine the Domain Relevancy of software components. It exploits the fact that Feature Vector codes can be seen as documents containing terms -the identifiers present in the components- and so text modeling methods that capture co-occurrence information in low-dimensional spaces can be used. The FV code representation of clusters or domains is used to find the domain-relevancy of the software components. PLSA has provided better results than LSA retrieval techniques in terms of Precision and Recall but its time complexity is too high. SVD Transformation with Naïve Bayes scheme has outperformed all other approaches and shows better results than the existing approach (LSA) being used by some open source code repositories e.g. Sourceforge. The DR-value determined is close to the manual analysis, used to be performed by the programmers/repository managers. Hence, the tool can also be utilized for the automatic categorization of software components and this kind of automation may improve the productivity and quality of software development

Using Latent Dirichlet Allocation for automatic categorization of software

“…Similarly, different IR techniques such as LSI, pLSI, and Naïve Bayes approaches have also been applied to categorize reusable software components [15]. Software components have also been retrieved using IR techniques for software reuse.…”

On using machine learning to automatically classify software applications into domain categories

“…Although, we could not compare our approach to the Decision Tree based model proposed in (Kawaguchi et al 2003) because we extract terms and API calls from applications, we use the same Decision Tree algorithm with our attributes selection strategy. Sandhu et al (Sandhu et al 2007) propose an unsupervised approach using Naïve Bayes classification and a hybrid model of Naïve Bayes with LSA; the model categorize 63 components belonging to six categories or domains that were defined manually. Therefore we cannot compare our model to the one proposed in (Sandhu et al 2007) because ours is supervised and we use 4,031 applications (745 from Sharejar and 3,286 from SourceForge) that are categorized in a predefined list of categories.…”

“…Sandhu et al (Sandhu et al 2007) propose an unsupervised approach using Naïve Bayes classification and a hybrid model of Naïve Bayes with LSA; the model categorize 63 components belonging to six categories or domains that were defined manually. Therefore we cannot compare our model to the one proposed in (Sandhu et al 2007) because ours is supervised and we use 4,031 applications (745 from Sharejar and 3,286 from SourceForge) that are categorized in a predefined list of categories.…”

“…Support Vector Machines, Decision Trees and Naïve Bayes have been used previously for software categorization in (Ugurel et al 2002), (Kawaguchi et al 2003), and (Sandhu et al 2007), respectively. RIPPER and IBK have not been used for software categorization; however they have been widely used in text categorization (Sebastiani 2002).…”

“…Thus, in a categorization problem, the inputs for NB are the attributes and the output is the probability distribution of the project on the categories. NB was used for software categorization by Sandhu et al (Sandhu et al 2007).…”