台中住宅與汽機車竊盜案件之空間分析

Abstract

中文摘要 依警政署的統計資料顯示，「竊盜案件」在各年度治安直接相關案類總數中所占的比率高達五成左右。本文從犯罪預防實務觀點針對台中市的竊盜案件進行研究，以核密度推估法、空間自相關分析尋找竊盜案件的熱點與群聚變遷；並基於犯罪理論中的「日常生活理論」、「社會解組論」建立相關係數矩陣與空間計量模型，探討竊盜案件的相關因素。 根據2016年至2019年可視化犯罪地圖顯示，竊盜案件逐漸減少。其中，2019年竊盜熱點聚集於台中市中心附近；而豐原區與潭子區交界處，不但群聚現象十分顯著外，其聚集強度也僅次於主要熱區而已。最後研究結果發現，便利商店附近最容易發生竊盜案件，而監視器的設置確實可以有效抑制住宅竊盜，然而汽機車竊盜卻沒有顯著的效果。 本文藉由犯罪地圖繪製、熱點分析方法，找出具顯著性的犯罪地理群聚及變遷現象，進一步利用空間誤差模與空間延遲模型作分析，以期該結果能協助警政單位制訂犯罪治安之改善策略。

Abstract According to statistics from the National Police Agency, the proportion of " Theft case " in criminal cases is as high as 50 percentage each year. This study explores the theft cases in Taichung City from the perspective of crime prevention practices. And then using the kernel density estimation and spatial autocorrelation analysis to compare the distribution of the robbery in different periods and draw the crime map. In addition, we based on the "Daily Life Theory" and "Social Disorganization Theory" to establish the spatial lag model (SLM) and the spatial error model (SEM) for further research. In the study of the spatial analysis, the visualized crime map from 2016 to 2019 proves that theft hot spot is toward to the center of Taichung. While secand hot spot is bondery of Fengyuan District and Tanzi District , there is an obvious phenomenon of spatial congregation.Finally, the results of the study found that the convenience store prone to thefts, and the setting of monitors can indeed effectively suppress residential theft, but has no significant effect on the theft of automobile-and-motorcycle This article uses crime mapping and hotspot analysis methods to find significant crime geographic clusters and changes, and further uses spatial error model (SEM) and spatial lag model (SLM) for analysis, hoping that the results can assist police units in formulating crime security improvement strategies.