Trajectory Pattern Mining And Route Planning

The rapid rise of social networks based on geographic information services has changed people’s travel habits.These social networks allow people share their travel experience through the checkout and other ways.These social networks have accumulated a lot of checkout data.These data usually contain two aspects of information.On the one hand the content of checkout provides the semantic information.On the other hand,the location information is provided by the coordinates of the data.It is a hot spot to study the trajectory data by analyzing people’s driving or travel experience based on the spatio-temporal data mining and providing the users with effective navigation.The earliest study of the moving trajectory,mostly focus on extract POI information from the GPS trajectory,and in accordance with the location of the history of similarity between the user to carry out the line route planning.Now on the trajectory pattern mining and route planning work,it can be divided into the following three categories: landmark identification,trajectory pattern mining,route planning.In this work,we using historical check-in data to solve route planning.A path recommendation algorithm that combines location information and semantic information is studied.First,we divide the space into grids.Then mapping the check-in data into the grids.After mining the textual of the check-in data,we get the semantic information of the grids.Using the semantic information of the grids,we clustering the grids into regions.And find out the user’s points of interest in each region.Secondly,we infer the connectivity of the user’s points of interest in each region and the connectivity between different regions by mining the location information of the data.We build the two layers data structure that is a map.Finally,according to the user’s query set of locations,we find all the similar users by historical trajectory.Then calculate the probability of the query sequence to determine the access order.Eventually,we get the route planning result.