Clustering web services would greatly boost the ability of web service search engine to retrieve relevant ones. An important restriction of traditional studies on web service clustering is that researchers focused on utilizing web services’ WSDL (Web Service Description Language) documents only. The singleness of data source limits the accuracy of clustering. Recently, web service search engines such as Seekda! allow users to manually annotate web services using so called tags, which describe the function of the web service or provide additional contextual and semantical information. In this paper, we propose a novel approach called
, in which both WSDL documents and tags are utilized for web service clustering. Furthermore, we present and evaluate two tag recommendation strategies to improve the performance of
. The comprehensive experiments based on a dataset consists of 15,968 real web services demonstrate the effectiveness of
and tag recommendation strategies.