The Analysis And Application Of Urban Traffic Trip Data Based On Knowledge Graph

With the advent of the information age,every equipment,every car,and every person is creating a large amount of traffic information,such as the swipe card records of public transport travel,the track data of vehicle travel,an so on,but these data are insufficiently integrated,lack compact and effective organizational structure and have not yet formed a complete traffic knowledge system,so it is difficult to conduct indepth data mining and intelligent analysis applications.Therefore,it is necessary to use artificial intelligence technology to effectively integrate multi-source traffic data,mine the potential value of the data,provide convenient services for urban traffic travelers and assist traffic managers to formulate control strategies,so as to improve the travel quality of urban residents.The knowledge graph stores and expresses data in a graphic structure,which makes data retrieval speed is faster and can achieve a second-level response,and it can handle complex and diverse association analysis,which is more conducive to knowledge query.Moreover,the knowledge graph has the ability to learn,which can mine the hidden knowledge through knowledge reasoning.Therefore,this paper analyzes the big data of urban traffic travel based on the knowledge graph,and further applies it to the traffic travel scenarios.The main work of this paper is as follows:Firstly,the traffic vertical domain knowledge graph is constructed by the topdown method,which divides the entities and relationships in the traffic domain,and realize knowledge reuse and sharing by using the pattern layer of the knowledge graph.And the data layer of the urban traffic travel knowledge graph is stored in the graph database Neo4 j.The knowledge graph constructed in this paper includes the rail transit travel knowledge graph and urban road travel knowledge graph.Secondly,the knowledge reasoning model based on knowledge representation learning is used to complete the missing data of the head entity or tail entity in the triplet,and to excavate the potential relationship between the two traffic entities to enrich the knowledge contained in the knowledge graph.For example,this paper realizes the completion of the OD points of public transportation commuters and the exploration of the relationship between the urban road traffic situation and points of interest.This paper verifies the validity of the reasoning model by linking prediction task.Finally,based on the knowledge graph,the efficient retrieval of traffic knowledge is realized,and an intelligent question answering system for traffic travel scenarios is developed,in which the natural language questions are semantically analyzed based on the deep learning method,and the query results obtained from the graph database are displayed in a visual form.It provides traffic travel query services based on the knowledge graph,including shortest path query based on graph algorithm and companion query based on travel chain similarity.