Research And Implementation Of Service Aggregation Method Driven By User Demand

In the context of large-scale services,users are confronted with a large number of services,and their needs are characterized by diversity and complexity.Users expect to get services with multiple functions that can meet their diverse needs in a more convenient and fast way.For example,a route planning service can not only provide route information,but also determine whether the route is blocked or not.Therefore,the concept of service aggregation was born.Service aggregation is defined as the process of aggregating atomic services with similar characteristics into a functionally complex service.However,most of the existing service aggregation methods use a bottom-up approach,aggregating similar services to form services that satisfy multi-level functions according to the service functions.In this process,due to the lack of consideration of user needs,the aggregated results often fail to meet the individual needs of users.In addition,from the perspective of the application of service aggregation,on the one hand,the number of aggregated services is large and the types are extensive,on the other hand,the aggregated services are still in a discrete state and there is a lack of correlation between services,which does not provide good support for further service discovery and service recommendation and affects the application effect of service aggregation.In response to the above two problems,this paper designs a set of user demand-driven service aggregation method,which is divided into two stages to complete service aggregation : demand service matching and matching result aggregation to form service community.The main work are as follows.Firstly,this paper establishes a service base model based on the attribute information about the service,and on this basis constructs a user requirement model by combining user requirements.In terms of to demand expression,a demand service matching method based on similar users is proposed to the problem that users usually have difficulty in describing their demands completely and accurately.The user-based collaborative filtering algorithm is used to achieve similarity matching and preference prediction between users and users,and the demand labels of similar users are used to supplement and correct the demand labels of current users to obtain the recommended set of user demand labels.Based on this,the iterative algorithm LMIA for layer-by-layer matching of demands and services is proposed to complete layer-by-layer matching of services using users ' basic demands,desired demands,and excited demands.In this paper,we conducted simulation experiments using the crawled service data.In the experiments,we compare the other two algorithms at the same time,and the experimental results illustrate the improvement in the method in terms of both the check-all rate and the check-accuracy rate,especially in terms of the check-accuracy rate by 5.3%.Secondly,the service community is formed by aggregating services based on the results obtained from service matching combined with users ' demand information.In this process,the matched result set and user's expected demand is used to build a service network and establish the association relationship between services and services.Based on this,a weighted GN algorithm based on user needs,User-WGN,is proposed.The algorithm optimizes the calculation of side meshes and the determination of termination conditions of the traditional GN algorithm,and introduces the desired needs of users into the algorithm to enhance the fit between users ' needs and the community.Afterwards,the constructed communities are verified with the existing communities,and the communities with high similarity are merged to avoid duplication of communities and promote the discovery and recommendation of services.The feasibility of this method is illustrated by example analysis compared with existing community building algorithms.Finally,based on the proposed theory and algorithms,this paper designs and implements a user-driven service aggregation system.The requirement analysis and process design of the system are explained in the implementation phase of the whole system.Various pages of the core process are shown in the paper to illustrate the feasibility and effectiveness of the approach proposed to this paper.