Taxi Experience Path Mining And The Construction Of Its Knowledge Graph

The rapid economic development has brought about the continuous growth of the number of urban motor vehicles,and the use of paths recommended by various navigation systems has become the main choice of people in daily travel.However,with the prominence of urban traffic congestion and people’s urgent demand for convenient and fast travel,traditional traffic route recommendation is facing challenges,and there is an urgent need to incorporate some new route recommendation strategies to supplement the existing route recommendation methods.Urban cab drivers have accumulated valuable experience in long-term travel activities,which can provide an important data source for new route recommendation strategies.Knowledge can be used to describe human cognition of things and summarize experience,and applying it to the transportation field can help people better grasp transportation knowledge.Therefore,this study collects the data of driving trajectories of cabs carrying passengers by using in-vehicle GPS system,analyzes the empirical characteristics of cab drivers in choosing the most efficient paths in the road network through data mining methods,and establishes the empirical path knowledge graph to apply the driving experience of cab drivers in traditional path recommendation.The main work of this study is as follows:1)To study the establishment of urban road network model based on proximity relationship.This study corrects the original road network data from OSM(Open Street Map)to solve the problems of lack of road length data,insufficient information of road network nodes,and the number of roads in the road network are many and cluttered.Road nodes are extracted and an urban road network model based on proximity relationship is established to describe the connectivity and directional relationship of nodes in the urban road network.2)To study the mining of cab drivers’ empirical paths.We extract the hotspot areas for picking up and dropping off passengers from a large amount of cab trajectory data,find out the trajectory data in each OD interval(starting point to end point interval)respectively,use Hidden Markov based road network matching method to get specific cab driving routes,extract driver’s empirical trajectories by trajectory similarity calculation method,and analyze the characteristics of cab driver’s empirical paths.3)Study the construction of empirical path knowledge graph.Combining the city road network model,cab drivers’ empirical paths and their features,road nodes are extracted as entities,directions and distances between road nodes are used as relations,and attributes of road nodes and features of empirical paths are used as attributes to build the empirical path knowledge map by the bottom-up method.