Context Aware Web Service Recommender Supported by User-Based Classification

Generally, the recommendation systems aim to present the items for the users based on their preferences that take from past behavior. However, Context-sensitive recommender systems have been tailored the recommendations according to a specific situation. In other words, it uses contextual information for making better recommendations. Context modeling is one approach of context-sensitive recommender systems that incorporate context directly into the recommendation process. In particular, the context in the web services recommender system is no less important from that in marketing. The proposed work includes two stages; the first one related to suggest services by using user-based classification that depends on majority voting. As for the second stage, related to suggest a services with contextual information using the method based on neighborhoods. In fact, the latter works at least on three dimensions namely; user, service, and context. The proposed work has been constrained with only services that be suggested from the first stage, hence, it would be narrowed the domain of services that forms the bulk. The proposed method has been applied on a real-world dataset; WS-DREAM, where the method records success percentage significantly, between 76% to 81% for response time and 50% to 87% for throughput. Worth to mention, the comparison has been conducted with the latest work such as; WSPred.