An Empirical Study to investigate the Effectiveness of Different Variants of SMOTE for Improving Web Service Anti-Patterns Prediction.

In today’s world, IT professionals must ensure that all enterprise applications are running smoothly and are communicating with each other. Service-Oriented Architecture(SOA) provides the organization with a framework that makes the management of information technology systems affordable and manageable. Service-Based Systems(SBS) need to adapt themselves over time to fit in the new client prerequisites. These outcomes in the weakening of the software systems quality and plan and may cause the emergence of poor solutions called Anti-patterns. An anti-pattern is a repeated application of code or design that leads to a bad outcome. The research uncovered that the presence of anti-patterns thwarts the software systems advancement and maintenance. The early prediction of these anti-pattern using extracted features from source code helps to reduce the software system’s maintenance and enhance the quality of the software. This present work’s ideology is to investigate the viability of different data sampling technique variants empirically and the machine learning technique, Naive Bayes, in the anti-patterns prediction in web services.