A SURVEY ON HIGHLY ACCURATE PREDICTION ALGORITHM FOR UNKNOWN WEB SERVICES WITH FUZZY CLUSTERING

Abstract

Quality of Service (QoS) assurance is an important factor of service recommendation. The web services which are never been used before by users have some indefinite QoS values for that service, and hence the accurate prediction of indefinite QoS values is important for the successful consumption of Web service-dependent applications. Collaborative filtering is the technique which is broadly accepted in the prediction of indefinite QoS values as it is significant for predicting missing values. Though, collaborative filtering derived from the processing of subjective data. The QoS data of Web services are generally objective, implies that present collaborative filtering-based approaches are not applicable every time for indefinite QoS values. Fuzzy clustering method is the new approach of achieving the QoS prediction with calculating the user’s similarity increases the prediction accuracy and this is verified by evaluating experiments with other processes. The quality of web services is additionally supposed as a multi-dimensional object, and every dimension is one feature of the web services non-functional properties. This approach is used in web services ranking by a score function and multi-dimensional QoS properties are mapped into a single dimensional value.