By grouping Web services that share similar functionalities, Web service clustering can greatly enhance Web service discovery and selection. Most existing clustering techniques are designed to handle long text documents. However, the descriptions of most publicly available Web services are in the form of short text, which impairs the quality of service clustering due to the sparseness of useful information. Towards this issue, we propose a new service clustering approach based on transfer learning from auxiliary long text data obtained from Wikipedia. To handle the inconsistencies in semantics and topics between service descriptions and auxiliary data, we introduce a novel topic model – Tag aided dual Author topical model (TD-ATM), which jointly learns two sets of topics on the two data sets and automatically couples the topic parameters to avoid the potential inconsistencies between these two data sets. Experimental results show the proposed approach outperforms several existing Web service clustering approaches.