Fuzzy reasoning approach for predicting web services QoS/QoE with ANFIS

Abstract

Nowadays, the web service (WS) usage in information systems (IS) includes determining a feasible WS that fulfils a set of non-functional requirements of Quality of Services (QoS) and user’s needs of Quality of Experience (QoE). While most existing studies evaluate WS from one perspective, i.e., users, and are based on data-driven approach, which employs a numerical dataset to learn a reasoning model, they overlook that users express their needs in a non-numerical form. To address these issues, we propose a new fuzzy reasoning approach for predicting WS QoS/QoE with the adaptive neuro-fuzzy inference system (ANFIS) that encompasses multiple viewpoints and perspectives, and is also suitable for linguistic terms. To verify the efficiency, we implemented the proposed approach, conducted two experiments and compared them. The results show a good performance of the proposed approach for predicting WS QoS/QoE, and, consequently, it can be considered a suitable tool for predicting.