A Machine Learning–Driven Personalized Web Service Recommendation and Prediction System for Intelligent User-Centric Applications

Abstract

The dynamic and time-aware Quality of Service attributes are forecasted for the recommendation of efficient web services. The forecasting of QoS attributes are attained primarily by statistical techniques and optimization algorithms. These approaches lack in retrieving the optimized results and the absence of exploitation nature in optimization algorithms results in ineffective performance. To overcome the drawbacks, the optimization and machine learning technique is introduced in this research work. The genetic algorithm (GA) with high exploration ability and grey wolf optimization (GWO) algorithm with high exploitation ability is introduced to acquire the optimized QoS attributes. The performance of the GA-GWO is enriched by minimizing the over fitting issue. The machine learning based gradient boosting algorithm is incorporated for solving the over fitting problem.The proposed GA-GWO is highly efficient in acquiring the QoS attributes, which are prominent in managing and modelling the web services. The recommendation of personalized web services and attaining QoS is investigated by the performance of the proposed GA-GWO algorithm and it outperforms the other QoS attribute retrieving approaches.