Application of Self Organizing Maps to User Authentication Using Combination of Key Stroke Timings and Pen Calligraphy

Recently, security of the computer systems becomes an important problem. Almost all computers use the password mechanism for the user authentication. But password mechanism has many issues. In this paper, we propose a kind of biometrics authentication method using the combinations of key stroke timings and pen calligraphy. For this method, selection of the phrase is important. We analyzed the key stroke timings and pen calligraphy using Self Organizing Map (SOM) and proposed the improved SOM (Pareto learning SOM) for this purpose. We examined the effectiveness of this method with the authentication experiments using the map organized by SOM.. Key-Words: Biometrics, Authentication method, Neural network, Self Organizing Map