Anita Devi

Saurabh Charaya

Mukesh Maan

Abstract

Software maintainability is proven to be one of the important factors to predict software Quality. Nowadays users are more focused on quality rather than quantity. In order to predict software maintainability at an early stage with better accuracy, soft computing techniques namely fuzzy logic is widely used and helpful in minimizing cost and maintenance time. In this paper, we proposed and implemented a fuzzy model using four parameters based on the ISO9126 Quality model viz: analyzability, changeability, stability, and testability to predict software maintainability level. Each input parameter is divided into Low, Moderate, and High whereas the output parameter is divided into Very_Low, Low, Moderate, High, and Very_High. Proposed model yields software maintainability level with better accuracy.

Keywords- Software Maintainability, Fuzzy Logic, Neural Network, Software Quality