Research For Personalized Web Service Selection Mechanism Based On Multi-layer QoS Ontology

This research of this paper comes from the topic"The Architecture System of Enterprise Application Software for the Survey & Design Based on SOA Technology", which is the third sub-project of the project"Research and Application for Key Technology of the Survey & Design Enterprise Informatization". This Project is one of the Key Projects in the National Science & Technology Pillar Program during the Eleventh Five-Year Plan Period.Nowadays, there are more and more Web services with similar functionality in the network. After process of service discovery, people usually get a group of similar services. They might search again to find the most suitable one. As a result, service selection based on nonfunctional properties, which is also called Quality of Service (QoS), has been proposed. Existing researches focus on selecting services for different users using the same QoS preference. However, the traditional mechanism may not provide users with the most suitable service due to their different preference to the same QoS properties. So, the process of service selection should meet the QoS personalized requirement of different users.Describing Web service with semantic information is becoming possible with the rapid development of semantic web and ontology. Thus, this paper construct the QoS ontology model using sematic Web service description to describe personalized QoS preference in a better way. A semantic Web service personalized selection method based on multi-layer QoS ontology was proposed. A Web service personalized selection mechanism also was designed to select Web service in personalized way using various kinds of QoS properties and to provide semantic support for QoS properties description.This paper mainly accomplished the following tasks:①It analyzed the service-oriented architecture and Web service technologies. The related technologies of semantic web, ontology and semantic Web service were also analyzed. The information of the Quality of Service (QoS) of Web Service had been introduced.②It researched the related contents of the traditional QoS ontology model and Web service selection method. After analyzing advantages and disadvantages of those works, a multi-layer QoS ontology (ML-QoS Ontology), was constructed by ontology editing tools. ③In view of the weak points of traditional methods in converting the personalized weights of user's QoS properties, the order relation description and related mathematical theory were used to transform the personalized weights of the user's QoS properties.④A semantic Web service personalized selection algorithm(SWSPS) based on ML-QoS ontology was proposed and it outperformed traditional algorithms in the recall ratio and precision ratio.⑤A personalized Web service selection mechanism based on the SWSPS algorithm was proposed. The user satisfaction degree was obtained by the feedback response module.⑥The simulation and experiments showed the efficiency and performance of the proposed semantic Web service personalized selection method.