Trip Purpose Identification of Urban Residents Based on Taxi GPS Data and POI Data

A series of traffic problems are prominent with the rapid development of society and economy and people’s living standards. It is extremely important to carry out reasonable traffic planning, in which the trip purpose of urban residents is an important reference factor. Base on taxi Global Positioning System(GPS) data of Qingdao, the data is preprocessed, and citizens’ trip purpose is identified by the Random Forest algorithm and other methods in machine learning. Base on the Point of Interest (POI) data of Qingdao, We can also identify the trip purpose by using k-d tree in Nearest Neighbor Search. Results show that the machine learning methods is feasible and achievable to infer the traveling purpose of citizens, and the accuracy of the Random Forest method is the highest, and combined with results of k-d tree to make a comprehensive judgment on the trip purpose of the citizens, the approach which combines GPS and POI data yields the best performance.