Crime Pattern Analysis Based On One-key Alarming System Of Public Transportation

Criminal activities are highly related to our daily life,especially for those happened on public transportations.Taking the taxi alarm data in Ecuador as the object of this research,we aim to find the relationship among taxi alarm data,other types of crime data from 911 data platform and geographic data of local roads.In order to find the optimal resolution of the regular grids needed for canonical correlation analyses(CCA),two kinds of models are built to simulate the spatial patterns of raw data,which are linear aggregation models and areal aggregation models.What’s more,a certain kind of weight matrix with double thresholds is used in calculating spatial correlation indexes.By comparing results from different regular grids with different resolutions,spatial patterns are found and used to find out the optimal resolution needed for CCA.While doing CCA,several kinds of data are taken out as they have much lesser relationships with taxi alarm data than other ones.Though CCA,two combinations of most related data sets are found.Then with the condensed data sets,grey canonical correlation analyses are made to adjust the results of CCA,which have also taken timing effects into consideration.At last,crime patterns of those happen around taxies are taken into consideration to speculate the reasons to the results we acquired.In this study,models of regular grids are built to find out the optimal resolution for further analysis,then CCA and grey CCA are used along with crime patterns on taxies to reveal the crime pattern.Local police officers can use this study as a guidance to the true nature of crime activities around taxies,and it can also help preventing such crimes.