Visualization Technology And Application Based On Clustering

With the popularity of computer applications,the amount of data generated by information system is increasing.There is much important information hidden behind the surge in data.In order to understand the huge quantity of data,the visualization is widely used.In order to make the process of data analysis more intuitive and interactive,the data visualization can combine the advantages of human visual with subjective cognition.As a result,it can dig out more valuable and easy-to-understand information.Howeve,if we only use visualization,there are some problems that the quantity of data is too large,or the date is too complex to process.So it is necessary to use the clustering algorithms to simplify and integrate data.The emphasis of this research is clustering-based visualization technology.In this paper,we analyze the space-time activity track of the pseudo base station and the space-time distribution of the short message content by the spam message data provided from mobile security software.The specific research content is as follows:First,summarize the data visualization and clustering technology to lay a foundation for the research.In the example part,use the order of time as the pseudo base station space-time thermodynamic map to describe the time and location information of the pseudo base station.Next,draw the moving trajectory of the pseudo base station: cluster the data by the clustering method based on ASDCT to identify the multiple pseudo base station trajectories behind the same number at the same time of the day;Then converge the trajectory which has similar space-time laws as the same pseudo base station by the clustering method based on structural similarity to locate the pseudo base station more accurately.Finally,summarize the space-time laws of different types of spam messages sent by the pseudo base station according to the classification of spam messages sent by the pseudo base station.The visualization technology with clustering solves the problem that the traditional pseudo base station strike method can not connect the time to the space dynamically.It can predict the activities of the pseudo base station more accurately and provide a more energeticl basis for the surveillance of pseudo base station.