Design And Implimentation Of The Personalized Web Service Recommendation Sub-System In Web Service Search Engine

SOA is becoming more and more popular for its flexibility. As the main implementation of SOA, Web services are now playing an important role in application development. Because web services can be integrated into various services easily, application developments are becoming cheaper and more efficient. Now developers are eager to find a good way for web service discovery. However, the current UDDI for web service indexing can’t meet their needs. In this paper, a personalized web service recommendation sub-system based on the web service searching engine is proposed. The recommendation sub-system uses collaborative filtering as its main recommending method to help developers find their needs as supplements of the search results.Web service recommendation sub-system is one of the most important modules in web service search engine. Four customized recommendation algorithms have been pointed out in four different situations from which users can get personalized recommendations. First of all, the algorithm of improved newton law of cooling is introduced in hot services recommending situation. Second, the paper improves the user-based collaborative filtering algorithm and the item-based collaborative filtering with browsing history based collaborative filtering and tag based semantic analysis algorithm in order to alleviate the impact of the cold start problem. Third, a tag based service recommendation algorithm is pointed out for personalized service recommendation of tag. At last, QoS filtering and ranking algorithm is introduced to ensure the quality of recommended results.First, the paper talks about commercial recommendations and web service recommendations to introduce the related works in the field. Then technologies in the field are listed. After that, the demands of the recommendation system are analyzed in detail and thus the design and implementation of the system is carried out. Afterwards, unit tests and system tests are conducted for verification. At last, the paper summarizes the works and points out some possible improvements.