Modeling Object-Oriented Software Maintenance Costs Using Adaptive Neuro-Fuzzy Inference Systems

International Journal of Computer Science and Technology | Vol. 10, No. 5, May 2008 | pp. 33–40

DOI: 10.46882/2008/IJCST/105014

Research Article

Title: Modeling Object-Oriented Software Maintenance Costs Using Adaptive Neuro-Fuzzy Inference Systems

Names of Authors: Hasan M. Tarawneh¹

Authors’ Affiliations: ¹Department of Software Engineering, Al-Balqa Applied University, Salt, Jordan

Abstract: Accurately projecting software maintenance efforts during the early stages of the software development life cycle remains notoriously difficult due to the non-linear interactions of structural codebase properties. This study develops a predictive model for object-oriented software maintenance effort utilizing an Adaptive Neuro-Fuzzy Inference System (ANFIS). The framework utilizes Chidamber and Kemerer (C&K) metrics—specifically Depth of Inheritance Tree (DIT), Coupling Between Object Classes (CBO), and Weighted Methods per Class (WMC)—as key inputs to map out architectural complexity. The ANFIS network was trained and validated using empirical data extracted from 24 historical open-source Java projects. The system’s predictive accuracy was evaluated using the Mean Absolute Relative Error (MARE). Experimental analysis showed that the ANFIS model achieved a low MARE of 11.4%, significantly outperforming traditional multiple linear regression models (which yielded a MARE of 26.8%) and standalone artificial neural networks (yielding a MARE of 18.2%). The model offers software project managers a precise tool for optimizing human resource allocations during post-deployment phases.

Keywords: Software maintenance, ANFIS, Chidamber and Kemerer metrics, software metrics, neuro-fuzzy systems, cost estimation

Manuscript Timeline: Received February 14, 2008; Revised March 28, 2008; Accepted April 18, 2008; Published May 10, 2008

Citation: Tarawneh, H. M. (2008). Modeling Object-Oriented Software Maintenance Costs Using Adaptive Neuro-Fuzzy Inference Systems. International Journal of Computer Science and Technology, 10(5), 33–40.

Subscribe to read the full article: https://www.internationalscholarsjournals.org/subscribe-to-read