Spatio-temporal Analysis Of Arson And Identification Of Gang-related Arson Based On Data Mining

Arson is a special form of crime.It has the characteristics of complicated criminal motives,and it is difficult to investigate and collect evidence.In addition to causing casualties and property damage,arson may also have a huge social impact and cause psychological panic in the public.At present,data mining technology is increasingly becoming a powerful tool to prevent and combat crime.In the study of arson prevention strategies,starting from the temporal and spatial patterns to mining the pattern of arson will help the deployment of police and firefighting resources,and reduce redundant work and response time.Rapid identification of gangrelated arson will provide effective support for the detection and prevention of arson cases.In order to analyze the temporal and spatial characteristics of arson and explore the its spatio-temporal clustering effects,this paper uses the Knox method based on the to test whether the occurrence of arson incidents has significant near repeat effect.Experimental comparisons were made based on arson data from Texas and Ford Worth.Results show that arson usually show increased likelihood of repeat occurrence in places close to the initial incident in space and time.But data in different geographic areas also showed differences in spatio-temporal patterns.In order to analyze the spatial clustering pattern of arson incidents and solve the problem that classic clustering algorithms cannot be directly applied to mixed data,this paper proposes an improved clustering method ω-HACMD based on the spatiotemporal and nontemporal attributes of arson.This method improves the Gower distance on the basis of the entropy weight method,and integrates it into the agglomerative hierarchical clustering algorithm,thus taking into account the different value distributions of category attributes when clustering.Experimental results show that the proposed method outperform popular mixed data clustering models,such as k-prototype and Gower + k-medoids method.In order to effectively identify gang arson cases,this paper propose a hybrid method that combines ensemble learning and intelligent optimization algorithms to solve this problem.First,we develop the recursive feature elimination(RFE)-based feature selection method to remove redundant features.Second,for the data imbalance problem,we determine the optimal processing algorithm from 18 candidate algorithms.Third,after trying a combination of multiple base classifiers,we obtain the optimal base classifier combination.Fourth,when integrating the prediction results of the base classifier,we propose a weighted ensemble strategy.Finally,we use the differential evolution(DE)algorithm to optimize the parameters of the base classifier and the weight of the combination,which further enhances the identification ability of the model.To verify the actual performance of the proposed method,we conducted experiments on the US National Fire Incident Reporting System(NFIRS)database.The results show that the proposed method is significantly superior to other popular machine learning methods,which proves that this method can provide a more reliable decision basis in the detection of arson cases.This paper verifies that the use of data mining technology can discover the hidden patterns of arson from the perspectives of pre-crime prevention and post-crime investigation.On the one hand,through spatio-temporal analysis and geographic visualization technology,the gathering rules of arson and arson modus operandi in time and space can be obtained,thereby providing decision support for the deployment of police and fire fighting resources.On the other hand,an effective gang-related arson recognition model is established by using related characteristic data,which provides a reliable decision-making basis for the detection of the case.