Route Planning System Based On Experienced Route

With the widespread use of mobile smart devices,the Location Based Services(LBS)have developed rapidly,and among which route planning is a very essential part.Existing route planning system is motivated by some criteria,such as distance,time,cost and so on,which is hard to utilize an accurate model to combine all the criteria,so that the returned result can't always satisfy the all requirements of the users.Some efforts are made to improve the route planning result by using the trajectory data,which contains users' spatial-temporal information.Different from the traditional route planning system,this paper proposes a new route planning methodology by mining the user travel pattern from the trajectory dataset and generating the Experienced Routes.By processing and filtering the original sampling data,we first conduct an experienced trajectory database,then use turning probability and transfer probability to measure the travel experience and compute the experienced route according to trajectories from the database,finally we recommend the result to the user.Therefore,we design and realize a route planning system based on experienced routes.Our contribution can be summarized as follows:(1)We propose a new method to obtain the experienced trajectory from the primitive sample data,which can be adopted in a distributed computing platform for big data.(2)We propose turning probability and transfer probability to measure the travel experience and thus compute the experienced route by using the travel pattern.To reduce the spatial complexity,we further propose a processing framework based on partition.(3)We finally realize a route planning system based on experienced route.Experiment results on the real dataset shows the practicability of the system compared with the mature commercial route planning applications.