Research And Implementation Of Data Mining System For The Elderly Trajectory Data

With the development of information technology,the speed of data production is getting faster and faster and higher requirements for data processing theory have been brought forward.Therefore,related technology is developing rapidly.At the same time,with the rapid development of positioning technology,a large number of moving objects get an effective location tracking,resulting in massive trajectory data.The massive data spawned a huge demand for data analysis.Under this background,trajectory data mining technology has been developing rapidly.However,because of the complex and changeable characteristics of trajectory data,there is no uniform theory and technology suitable for all scenarios.According to different target types and application scenarios,different trajectory data mining techniques can only be employed in different scenes.Meanwhile,China already entered the aging society,more and more elderly solitary people are facing with the problem of dementia.Therefore,Therefore,based on the above background,this paper studies the method of data mining for elderly trajectory.The primary work is concluded as follows:GPS,Beidou navigation,and LBS base station positioning technologies are frequently used method to collect trajectory data.These methods may cause the problems of positioning errors,positioning position drift,and even missing positioning points.Therefore,it is very necessary to preprocess for trajectory data.By using the open source trajectory data set of Microsoft,this paper employs Calman filter algorithm and studies the pedestrian maneuver model.Meanwhile,the model is adopt to smooth the pedestrian trajectory.By analyzing the trajectory data set,this work determines the pedestrian mobility model and its various parameters,and uses the data set to simulate the proposed algorithm.It is found that the mobility model can match the motion rule of pedestrians,and the filtering effect is obvious.He traditional solution to solve the problem of losing the old man is generally to simply specify a range in the map,such as the circle or rectangle to delineate the range of the old man's activity.If the old man exceeds the range,the alarm can be triggered.This method can not adjust the range of activity according to the law of the old man's activity.Based on the anomaly trajectory detection method,this paper proposes a grid partition algorithm to detecting the abnormal track of the old man.The algorithm can automatically adjust the range of activity according to the activities of the elderly and realize the "growth" of the range of activity.At the same time,the algorithm is also a real time processing algorithm for large data.It can be easily implemented by using large data real-time processing framework Storm.This paper also presents the hot region recognition algorithm based on grid partitioning.The algorithm also has real-time quality,and it can be easily implemented using real-time processing framework Storm.In addition,the algorithm has the advantages of simple structure andlow algorithm complexity.Finally,the algorithm is verified by using the open source trajectory data set of Microsoft Asia Research Institute,which can meet the demand of the people identifying hot spots.At the same time,due to the useing of the grid,we can calculate the degree in pedestrians of interest to each grid by the traveler,and can directly use the collaborative filtering algorithm to find potential regions of interest for different pedestrians.Finally,in order to verify the above algorithm and solve the problem of missing elderly people,this paper designs and implements the track data mining prototype system.The system uses a real-time large data processing framework Storm,and uses Spring Boot to build a data visualization website.Then,the results are visualized through the website,and the abnormal situation of the old man's trajectory is pushed to the mobile APP so that the family can respond to it in time.