数据挖掘技术在刑事案件信息分析中的应用

发布时间：2018-01-07 15:33

本文关键词：数据挖掘技术在刑事案件信息分析中的应用　出处：《西安电子科技大学》2011年硕士论文　论文类型：学位论文

【摘要】：随着“金盾工程”的建设和应用，各级公安机关业务系统积累了大量的刑事案件原始数据信息。如何采用信息技术，发现更多数据中隐藏的犯罪规律，以期提高公安机关打击侦破犯罪的水平，成为一个极具现实意义的课题。本文基于数据挖掘技术,结合SQL Server2000数据库,通过数据清洗、概化及连续属性离散化把大量不适用挖掘的数据转化为适用挖掘的数据，首先解决了数据预处理问题；然后基于关联规则,重点研究了Aprior算法的改进，通过对剪枝和连接步的优化，有效提高了算法运行效率，分析兵团某师近三年的刑事案件数据记录，，得出频繁项集，导出关联规则，分析了案件类别与犯罪嫌疑人年龄、职业、户籍、文化程度等因素之间的关联关系；运用ID3算法，基于信息增益，构建了以案件类别为核心的犯罪分析决策树，并对模型进行了解释评估；最后设计开发出一个小型仿真刑事案件信息分析系统，验证了算法实现，为数据挖掘关联规则技术在地区公安信息分析中的应用提供了新思路，并拓展了该技术的使用范围。
[Abstract]:With the construction and application of the "Golden Shield Project", the operational system of public security organs at all levels has accumulated a large amount of original data information on criminal cases. How to use information technology to discover the hidden laws of crime in more data. In order to improve the level of public security organs to crack down on crime, become a very practical significance of the subject. This paper is based on data mining technology, combined with SQL Server2000 database, through data cleaning. Generalizability and continuous attribute discretization transform a large amount of unsuitable mining data into suitable mining data. Firstly, the problem of data preprocessing is solved. Then based on the association rules, the paper focuses on the improvement of Aprior algorithm. Through the optimization of pruning and linking steps, it effectively improves the efficiency of the algorithm, and analyzes the criminal case data records of a certain division of the Corps in the past three years. The frequent itemsets are obtained, the association rules are derived, and the correlation relationship between the category of cases and the age, occupation, household registration and education level of the criminal suspects is analyzed. Using ID3 algorithm, based on information gain, a crime analysis decision tree with case category as the core is constructed, and the model is interpreted and evaluated. Finally, a small simulation criminal case information analysis system is designed and developed, which verifies the implementation of the algorithm, and provides a new idea for the application of data mining association rule technology in regional public security information analysis. The application scope of the technology is expanded.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：D918;TP311.13

【参考文献】