Abstract. In software as a service (SaaS) model, application market provides various software services for users to access. However, large amount of software services are difficult to choice because of few attributes and incomplete description to illustrate their functionality. Besides, the fetch results from application market may not match user preference and waste user much time to get the desired software service. In this work, we propose an approach to improve software service searching effectiveness in an application market. Several advanced techniques are enforced. Information retrieval technology analyzes the description of a software service to get its key concepts. The association rule mining technology discovers the hidden association between various software service key concepts. The relationships of software service key concepts and discovered association rules are built a semantic network to connect relevant key concepts of software services. After configuring the software service attributes for quality of service consideration, the multi-criteria decision analysis is used to get the ranking order of the candidate software services. The software services key concepts, discovered association rules, semantic network, and multi-criteria decision analysis approach are built a recommendation system. User gets the reasonable software service based the ranking order of candidates from the recommendation system. We hope the proposed approach facilitates user to get the software service effectively in a popular application market.