基于谱系聚类法的用户击键特征识别  

Recognition of User's Keystroke Features Based on Hierarchical Clustering Method

作　　者：杨继龙[1] 陈曦[1] 杨伟康[1] 张琼文[1] 

机构地区：[1]电子科技大学计算机科学与工程学院,成都611731

出　　处：《电脑编程技巧与维护》2010年第4期35-36,共2页Computer Programming Skills & Maintenance

摘　　要：以用户击键特征为依据,提出了一种基于谱系聚类法的识别算法。该算法通过谱系聚类法对用户击键特征向量进行聚类分析,并形成各向量之间的谱系关系,从而由谱系关系来对击键特征向量进行识别。该算法的主要特点是使用欧氏距离进行分类,算法实现简单并且识别速度快。由于采用的聚类算法的简单性,其识别精度尚有待提高,因此该算法适用于击键识别的简单应用。a new user recognition approach using keystroke features based on Hierarchical Clustering Method is proposed. The algorithm classify user＇s keystroke feature vectors by Hierarchical Clustering Method, create the hierarchical relationship among vectors, and achieve the recognition of vectors. The main features of the algorithm is classifying by Euclidean distance, simple and fast. As a result of the simplicity of the clustering algorithm, its recognition accuracy has yet to be improved, so the algorithm is applied to a simple identity verification.

关 键 词：谱系聚类法 击键特征 身份认证 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]