Research on User Demand-Driven Service Matching Methodology

Services have seen a spike, and thus, how to accurately identify the needs of the users and match them with the appropriate services becomes the key to the research of service matching. Service matching involves a lot of work, of which the existing methods to complete the matching mainly depend on semantic similarity while the information of the service function is rarely taken into consideration. To solve the problem above, a user demand-driven method will be introduced. Based on the original service model, the user demand model is established, which aims at the actual needs of users. To cope with incomplete and inaccurate expressions of users' demands, the UserCF is applied to matching the similar needs of users so that the tag set is generated, making the user service tag information complete and modified; the LMIA is recommended to match the classified demands of users level by level with the purpose of weighted fusion. In the analysis of living examples, it is suggested that the method is more feasible than workable than the existing methods of service matching, and meanwhile, an 5.3% higher precision are obtained.