Enhancing the Accuracy of Case-Based Estimation Model through Early Prediction of Error Patterns

Abstract

The paper tries to explore the importance of software fault prediction and to minimize them thoroughly with the advanced knowledge of the error-prone modules, so as to enhance the software quality. For estimating a new project effort, case-based reasoning is used to predict software quality of the system by examining a software module and predicting whether it is faulty or non faulty. In this research we have proposed a model with the help of past data which is used for prediction. Two different similarity measures namely, Euclidean and Manhattan are used for retrieving the matching case from the knowledge base. These measures are used to calculate the distance of the new record set or case from each record set stored in the knowledge base. The matching case(s) are those that have the minimum distance from the new record set. This can be extended to variety of system like web based applications, real time system etc. In this paper we have used the terms errors and faults, and no explicit distinction made between errors and faults. In order to obtain results we have used MATLAB 7.10.0 version as an analyzing tool.