Defect Prediction in Software Entities Classified in Terms of Level Dependencies

Abstract

Unit testing is the core fundamental to ensure code is in accordance with the design specifications. The coding and unit testing standard reflects the stability of project (not to mention the testing effort).Code stability is greatly influenced by the efforts of unit testing, which can be automated to reduce the human efforts. In spite of several tools identified for unit testing, tools need to be able to identify the level dependencies or depth of program entity usage in software fragments. This factor greatly influences unit testing complexity. Higher the level of dependency, the greater the complexity of unit testing the code. Here based on level dependencies we predict defects in any expression. A predicting defect-prone software component is an economically important activity and so has received a good deal of attention. However, making sense of the many, and sometimes seemingly inconsistent, a result is difficult. The main objectives of this paper are unbiased and comprehensive comparison between competing prediction systems. This paper mainly focuses on two learning algorithms OneR, Naive Bayes. By using those two algorithms we calculate the error rate. We can predict defects based on those error rates.