Spatial And Temporal Visual Analysis Of Indirect Sparse Sampling Data

Spatio-temporal data visualization is an important branch in the field of information visualization.with the widespread popularity of sensor devices and mobile communication devices,there are more and more sparse sampling track data,such as road monitoring data,company employee punch card data,and mobile phone connection base station data.This type of data is collected differently from traditional GPS data.It does not require objects to carry GPS positioning devices.There is a problem that the sampling breadth is far better than the traditional trajectory data,but the sampling accuracy of a single object is insufficient.In actual scenarios,sometimes indirect information is needed to infer the trajectory of a moving object,such as malicious SMS data scattered by a pseudo base station.The data has certain deviations in time and space information.The trajectory data is sampled indirectly.To this end,how to explore and analyze the spatiotemporal characteristics of indirect sparsely sampled data is especially critical.This paper mainly solves the trajectory fitting work of indirect sparse sampling data and the exploration of space-time visual analysis of this kind of data.(1)A trajectory fitting method for malicious short message data.The method mainly solves the problem of trajectory data grouping and trajectory data reconstruction.The trajectory data grouping method combines the text content of the malicious short message,calculates the text similarity and the space-time similarity of the short message to divide the trajectory group,thereby determining the malicious short message sent from the same pseudo base station.The trajectory data reconstruction method mainly uses the following three steps: clustering and extracting the moving mode,resampling to generate the trajectory contour,combining the road network binding trajectory,and finally reconstructing the trajectory that conforms to the actual pseudo base station activity.(2)Time and space visual analysis of malicious SMS data.Using malicious SMS data to infer the behavior of pseudo base stations,this paper starts from three perspectives of time,space and semantics.Visual analysis of the time dimension,using various visualization methods such as rasterized view mode,stack view,calendar view,etc.to display pseudo base station activity at different time granularities.Spatial dimension visual analysis,using map-based heat map,scatter plot and spatial visualization view of different functional areas to show the pseudo base station law,combined with the trajectory fitting method to explore the behavior pattern of mobile pseudo base station.In the semantic dimension visual analysis,word cloud maps,pie charts,and river maps are used to reveal different types of pseudo base station behavior.