Clustering and Recommendation for Semantic Web Service in Time Series

• Yu Lei,
• Wang Zhili,
• Meng Luoming,
• Qiu Xuesong,

Vol. 8, No. 8, August 28, 2014
10.3837/tiis.2014.08.010, Download Paper (Free):

• Recommendation
• clustering
• Composite service
• semantic

Abstract

Promoted by cloud technology and new websites, plenty and variety of Web services are emerging in the Internet. Meanwhile some Web services become outdated even obsolete due to new versions, and a normal phenomenon is that some services work well only with other services of older versions. These laggard or improper services are lowering the performance of the composite service they involved in. In addition, using current technology to identify proper semantic services for a composite service is time-consuming and inaccurate. Thus, we proposed a clustering method and a recommendation method to deal with these problems. Clustering technology is used to classify semantic services according to their topics, functionality and other aspects from plenty of services. Recommendation technology is used to predict the possible preference of a composite service, and recommend possible component services to the composite service according to the history information of invocations and similar composite services. The experiments show that our clustering method with the help of Ontology and TF/IDF technology is more accurate than others, and our recommendation method has less average error than others in the series of missing rate.

Statistics

Cite this article