الفهرس | Only 14 pages are availabe for public view |
Abstract There are many physical biometrics such as iris patterns and fingerprints. There are also interactive gestures like how a person type on a keyboard, moves a mouse, holds a phone, or even taps a touch screen. Keystroke dynamics or typing dynamics is an automatic method that confirms the identity of an individual based on the manner and the way of the user typing on a keyboard. There are two types of keystroke systems, Fixed-text system, and free-text system and each of them has it is own importance. In this research paper, we are investigating the possibility of classifying individuals using features extracted from their keystroke dynamics with two different datasets: (1) fixed-text dataset with different difficulty levels and (2) free-text dataset with no restrictions what a user types on the keyboard. We worked on two online datasets, The MOBIKEY Keystroke Dynamics Password Dataset as a fixed-text dataset, and The Politehnica University Timisoara keystroke dataset as a free-text dataset. It is a very new dataset which has no published work on it up till now. We made a lot of data preprocessing on it as it was almost a row data. Investigation was done using several classification techniques: RandomForest (RF), Support Vector Machines (SVM), BayesNet (BN), and K-Nearest Neighbors (KNN). The highest accuracy achieved with the fixed-text dataset was 98.8% using RF for classification while the highest achieved accuracy with the free-text dataset was 87.58 % using RF classifier.We also made our own dataset and worked on two lines sequentially First We Worked on an online data “The MOBIKEY Keystroke Dynamics Password Database “and compared our results with it. The second line was making our own data for free text on keystroke authentication and work on it. It is hard for researchers to work on this topic because there is very hard to find free data set available online on this topic. So, we decided to make our own data set and work on it. We decided also to let it free and available online for other researchers. Our dataset consists of 20 person 12 women and 8 men with different ages. We investigated using traditional techniques: Equilidian Distance, Manhatten Distance, Mahalanobis Distance, Manhattan with Standard Deviation Distance. We compared our result with The MOBIKEY Keystroke Dynamics Password Database. The highest accuracy achieved with our dataset was EER=.34 with manhatten distance technique. |