Web service selection based on ranking of qos using associative classification

TL;DR

This paper proposes a framework for selecting and ranking web services by combining QoS classification with semantic matching, improving service selection accuracy based on user preferences.

Contribution

It introduces a novel service selection framework that integrates associative classification for QoS level classification with semantic similarity for ranking.

Findings

Framework effectively classifies services into QoS levels.

Semantic matching improves ranking accuracy.

Experimental results confirm satisfaction of user requirements.

Abstract

With the explosive growth of the number of services published over the Internet, it is difficult to select satisfactory web services among the candidate web services which provide similar functionalities. Quality of Service (QoS) is considered as the most important non-functional criterion for service selection. But this criterion is no longer considered as the only criterion to rank web services, satisfying user's preferences. The similarity measure (outputs-inputs similarity) between concepts based on ontology in an interconnected network of semantic Web services involved in a composition can be used as a distinguishing criterion to estimate the semantic quality of selected services for the composite service. Coupling the semantic similarity as the functional aspect and quality of services allows us to further constrain and select services for the valid composite services. In this paper, we present an overall service selection and ranking framework which firstly classify candidate web services to different QoS levels respect to user's QoS requirements and preferences with an Associative Classification algorithm and then rank the most qualified candidate services based on their functional quality through semantic matching. The experimental results show that proposed framework can satisfy service requesters' non-functional requirements.