中文摘要：

在互联网日益发展的今天，网络诈骗频繁发生，犯罪嫌疑人的身份极难识别，给网络诈骗案件的侦破带来了困难。该项目以即时聊天工具QQ为例，对网络诈骗嫌疑人身份识别及追踪给出一个完整的解决方案。文章通过键盘击键行为特征识别嫌疑人身份，击键特征能够高精准地区分不同的人，而且键盘普及率很高，这给系统运行提供了硬件环境。通过监听用户键盘击键提取击键特征，对用户进行身份识别，通过IP地理位置信息定位犯罪嫌疑人，从而协助执法人员侦查网络诈骗案件。

英文摘要：

In the increasing development of the Interact today, Interact frauds happen everywhere and anytime, to identify the suspect is so difficult that detect network bilk case brings difficulty. This project takes the instant chat tool QQ for example and gives a complete solution for the Internet fraud detection. Through the behavior features of suspects, the project can differentiate different people at the high precision, and keyboard popularity rate is very high, which provides hardware environment for the system operation. In the system, this project extracts keystroke feature by monitoring users＇ keystrokes, and introduces the local authentication and server authentication to identify users. After positioning for the suspects through the IP, the system assist the police to investigate network fraud cases.