Analysis Of User Trajectory In Heterogeneous Multi Scenario Source

With the rapid development of GPS technology and the widespread use of mobile devices,it is easy to get the trajectory data of users.By these data,we can reconstruct the trajectories of the users and get the moving rules of the users.Introducing the heterogeneous scenario source to the user trajectory analysis is order to more accurately mine the users' moving rules which what users do at what time,what place and what kind of environment.Therefore,the main work of this paper includes the following three aspects:1.A new Hausdorff distance about spatial-temporal trajectory similarity based on time restriction is proposed.Aimed at the problem that existing similarity measurements is not very good for the spatio-temporal trajectory data,the algorithm directly make the time attribute and position attributes of trajectory to involve in trajectory similarity calculation and mine all similar sub-trajectory in two longer tracks using the sliding window to judge the similarity between two longer tracks.The validity and reliability of the method are verified by experiments.2.A user's specific time trajectory feature extraction algorithm based on filter and refinement strategy is proposed.Aimed at the problem that it is difficult to find the user's interest points in a long time(such as 3 months)with a specific period of time(such as:daily 8:00?10:00),the algorithm divides into two stages to extract the user's trajectory feature.In the filter step,the user's trajectories in the same period for several certain days are clustered based on density to obtain the user's stops;in the refinement step,the stops are clustered to obtain the user's interests.The validity and reliability of the method are verified by experiments.3.An algorithm of predicting the user's next location based on non maximum limit is proposed.Aimed at the problem that the existing forecast algorithms which predict the user's next location do not take into account current outdoor temperature and the distance which between the next place and the the user's current location,the algorithm bases on Naive Bayesian classification algorithm,combines with the user's current time,current location,current outdoor temperature and the distanace which between the next place and the the user's current location,uses the non maximum limit to predict the user's next location.The validity and reliability of the method are verified by experiments.