Research On Semantic-based Hierarchical Matching Model Of Transportation Information Services

With the recent breakthrough in Cloud Computing and Artificial Intelligence technology,the development of Intelligent Transportation System(ITS)has also been deeply affected.Traffic information service is regarded as an important part of ITS.The content is diversified,including traveler behavior analysis,dynamic path navigation and personalized recommendation.Confronted with massive and heterogeneous traffic information service data,how to query the accurate and real-time user services is one of the hotspots in the field of intelligent transportation.Traditional keyword-based service matching methods are no longer applicable to a large number of service sets.The precision or recall rate based on the semantic service matching method has been greatly improved.However,semantic matching research is not mature,such as the lack of user personalized information and service context information.Especially for the service application field,the particularity of transportation is distinguished with any other industries.This paper focuses on the semantic traffic information service matching model based on the above problems.The main contents of the research include:(1)Introducing the basic structure of Web services and semantic Web services,analyzing the particularity of traffic information service compared with Web services,and comparing the advantages and disadvantages of three classical semantic Web service matching models.(2)By learning the semantic Web service matching method,a hierarchical matching model of traffic information service with three-layer structure is proposed.It includes that the service text data preprocesses and the clustering algorithm is used to automatically classify the service set.Combining the geometric distance model and information theory,the mixed mode of the model implements the single concept pair matching in the service function,and uses the bipartite graph algorithm to realize the optimal matching between the concept groups of the service function.In the traffic information service matching,the model adds the QoS attribute because the user’s individualized demand and the user’s feedback on the service are taken into account.The requests of user service return the set of services that meet the threshold after a layer-by-layer match.(3)According to the idea of three-layer service matching,designing the hierarchical matching model structure diagram of the traffic information service and the service description framework with semantic information,and realizing the function modules of the model.(4)The experimental part tests the validity of the model through the industry standard data set OWL-TC4.The results show that the service clustering process can significantly reduce the response time of service matching,and the traffic information service level matching model has better performance in terms of precision and recall rate.