摘要
根据用户的击键行为特征,提出了一种基于差别子空间的识别算法,该算法仅依据用户前几次成功登录的击键特征计算出能够代表用户击键的共性特征向量,进而利用当前用户击键特征向量与共性特征向量的欧几里德距离作为判别依据来判定用户的身份。该算法主要进行内积运算,实现简单且识别速度快,实验结果表明该算法误报率较低,鲁棒性较强。
A new user recognition approach using keystroke features based on difference subspace is proposed. The proposed approach calculates the common feature vector to depict a user＇s keystroke pattern according to keystroke features of the user＇s several recent successful authentications, and then utilizes Euclid distances between current keystroke pattern vectors and common feature vectors to identify users. The performance of the proposed approach is evaluated and experimental results show the computational simplicity, high efficiency and low false recognition rates of the proposed approach compared with some existing methods, it is also shown that the approach is robust.
出处
《计算机工程》
CAS
CSCD
北大核心
2007年第11期204-205,221,共3页
Computer Engineering
关键词
差别子空间
击键特征
共性特征向量
Difference subspace
Keystroke feature
Common feature vector
