Web Services Performance Prediction with Confusion Matrix and K-Fold Cross Validation to Provide Prior Service Quality Characteristics

Abstract

The Information Technology sector has experienced substantial growth, especially in the realm of application development, over the last few decades. This ongoing evolution highlights the complexity of contemporary IT applications, which no longer rely on a singular component. Developers now integrate a multitude of components from diverse sources, including users and vendors. Thoroughly examining the quality of external software components before integration into an application is imperative for ensuring optimal service performance. Numerous methodologies and approaches are available for appraising the quality of software components, including the Software as a Service Model (SaaS), Quality of Experience (QoE), and alternative models for identifying quality. These models employ traditional techniques to assess factors like availability, integrity, accessibility, security, performance, and reliability, collectively contributing to the measurement of Quality of Service (QoS). The suggested method enables thorough scrutiny of software components through a 360-degree evaluation, utilizing the Confusion Matrix for predicting performance. This evaluation method ranks web services based on throughput and response time, providing tangible values for decision-making by service users. The classification mechanism aids in categorizing standards within a benchmark web service dataset. By utilizing this performance measuring method, one can determine service quality through the confusion matrix, aiding in the identification of the best web services and contributing to the optimization of application performance.