基于Web新闻的案（事）件时空信息解析方法
Interpretation of spatio-temporal information of crime cases based on Chinese Web news

：探讨基于词库与规则相结合的案（事）件新闻文本时空信息解析方法. 通过构建时间词库和表达规则、派出所和地名等词库以及“触发词-案事件类型”二元分类器，实现对案（事）件新闻中的案发时间、案发地点、案（事）件类型和出警派出所信息的抽取，并引用设计规范化规则，实现时空信息的规范化输出. 实验分别选取本文解析盗窃案件数据和2014年1月至2015年3月福州市四个中心行政城区的公安盗窃案件数据进行比较，利用核密度估计算法研究犯罪集聚区，得到的集聚结果基本一致，发现福州市盗窃犯罪集聚发生于茶亭派出所、东街派出所和瀛洲派出所等辖区.
An approach which based on the combination with lexicon and rules to parse spatiotemporal information of crime cases from the web news was investigated. Designed time word lexicon，time expression rules，police station lexicon，toponymy lexicon and “trigger-crime type” binary classification to extract the information of time，address，police station and crime event type from the web news. Through the design of standardized rules，achieved the structured data of the spatiotemporal and attribute information of the crime cases. It took Fuzhou City as an example，based on the two sets of the data of theft event from four main districts of Fuzhou City from January 2014 to March 2015，one is from the Public Security Department of Fuzhou，the other is the extraction result by this paper. Using the kernel density estimation method to study the crime hotspots，the result are basically the same: Fuzhou theft crime agglomeration occurred in Chating，East Street and Yingzhou Police District