Research On Mining Accompanying Behavior Pattern Methods Based On Spatial-temporal Trajectory Data

The popularity of mobile locate-enable devices and Location Based Service(LBS)generates massive spatio-temporal data every day.The characteristics of spatio-temporal trajectory data serve as an important reference basis for characterizing moving objects,in which the behavior patterns of moving objects are closely related to the moving trajectories.Trajectory data mining has been applied in many fields to find the behaviour pattern while discovering traveling companions based on spatio-temporal trajectory data become become one of the most fundamental techniques in these areas.This paper focuses on the accompanying behavior pattern in significant mobile behaviour patterns and proposes a flexible framework named Group Seeker for discovering traveling companions in vast real-world trajectory data.Meanwhile,an algorithm named TCo D to discover traveling companions is designed to avoid the companion candidate omitting problem happening in previous researches based on the time snapshot slicing within a short period.On the other hand,the paper also presents a detailed analysis of the one man carrying multiple mobile phones(MCMP)pattern and designs related mining algorithms based on the Group Seeker framework for discovery and analysis.This pattern widely exists in real scenes,especially telecommunications fraud.In this paper,the specific algorithm part of the framework is divided into general accompanying pattern mining and MCMP pattern mining respectively.And the experiments on two real-world data sources and the simulation data source show the effectiveness and reliability of algorithms and also prove that this framework has a higher-efficiency performing to get results to discovery satisfying companions in a long-term period.