摘要
针对传统的口令认证方式在防止密码共享以及密码被盗方面的脆弱性,提出将用户名———口令认证与击键特性认证相结合的认证方式,并给出一种基于模糊c均值聚类进行击键特性认证的新方法.通过利用模糊c均值对用户的击键特性进行训练,生成用户击键特性样本;识别时将当前要求认证用户的击键特性与之进行匹配,并根据系统设定的阈值α来确定当前用户是否为合法用户.实验结果表明,此种击键特性认证方法具有较高的用户识别性能.
Due to the weakness of traditional password systems on deterring password sharing and stolen, this paper describes a new method of fuzzy c-means clustering to enhance user authentication based on the keystroke authentication studied by researchers. Fuzzy c-means is used to train the keystroke dynamics of users when entering passwords on a keyboard. After being trained, the system identifies whether a user is a legal user by comparing the current user's keystroke dynamics with his keystroke profile values within a (certain) precision threshold α. The experiment results indicate this approach has a good discerning ability.
出处
《延边大学学报（自然科学版）》
CAS
2005年第2期130-134,共5页
Journal of Yanbian University（Natural Science Edition）
关键词
模糊C均值聚类
击键特征
身份认证
fuzzy c-means clustering
keystroke
identity authentication
