Efficient Search For Activity Trajectories Based On Road Network

In recent years,with the popularity of mobile devices and the development of GPS technologies,research based on geographic location data emerges in endlessly,including similar trajectory search,activity trajectory search,trajectory recommendation with user preference and so on.Services based on these studies,such as travel route recommendation,have brought great convenience to people's life.In these studies,trajectory search is a core technique.The factors affecting trajectory search are spatial distance,activity information and road network information,etc.Existing trajectory search techniques take consideration of many factors,but they do not take consideration of spatial distance,activity information and road network information at the same time.To solve this problem,We study efficient trajectory search based on road network.On the basis of studying the data of trajectory location,we take the activity information and road network information into consideration.Among trajectory database,a trajectory will be recommended to user only when it meets the activity requirements of the query trajectory and the geographic distance is close and the road network information is similar,thus the recommended trajectory is more in line with the user's requirements.The algorithm will not only recommend the optimal trajectory,but also find k most similar trajectories.In order to support efficient query process,we map the road network information to a point in high-dimensional space,and then use RTree and RTree-3D technology to index the points in spatial domain and road network domain respectively.The algorithm we proposed takes full account of the space distance,activity information and road network information of the trajectory.It uses heuristic search strategy to find candidate trajectories in both space domain and road network domain and sets up upper and lower bounds to prune the search space effectively.When the candidate trajectory satisfies the activity information of the query point,we use the minimum activity point matching distance algorithm to calculate the distance between the query point and the candidate trajectory efficiently,and then get the comprehensive distance between the query trajectory and the candidate trajectory.At last,k similar trajectories are selected.Experiments on real data sets also demonstrate the effectiveness of our proposed method.Finally,we implement an efficient activity trajectory search system based on road network.