Automatic Tagging of Functional-Goals for Goal-Driven Semantic Service Discovery

With the growing number of Web services available on the Web, finding services that match user's query becomes more difficult. To deal with the problem, numerous approaches employed semantic techniques in terms of annotation (or tagging) and service discovery algorithms. However, the problem of semantic service discovery based on the identification of fine-grained goals of users and services is still challenging due to the lack of semantic information in Web services. To enable the goal-driven semantic service discovery, we propose an automatic functional-goals tagging approach which employs a set of natural language processing (NLP) procedures based on the descriptions of Web services. In addition, to check the effectiveness of the functional-goals tagged, we designed a goal-driven semantic service discovery algorithm and compared it with other approaches: keyword-based, ontology-based, and topic-based service discovery. As results, our proposed goal-driven semantic service discovery achieved 75.7% of precision and 59.6% of recall (finally, F1=62%) that outperform other discovery approaches.