A modified framework to detect keyloggers using machine learning algorithm

摘要

Keyloggers are very dangerous programs that does the monitoring of all the activities carried in our PC. Some of the activities include capturing screenshots of all the activities performed in PC screen, recording the activity performed by browser and generating the different keystrokes of the various activities performed in our PC. These activities are difficult to be traced by any detectable softwares. Hence there is an urgent need to detect the presence of keyloggers in our system and nullify all the existing keyloggers present in PC. They do all the spying and steal all the sensitive, confidential and important information. This information could be used for harmful purposes and endanger the life of the person associated with it. This is really a grave threat to the society. We have tried to find a solution to this grave problem. We proposed a new detection technique that will help in detecting all the keyloggers present in our PC using Machine learning algorithm. The different keyloggers which are available or being installed are detected using Support Vector Machine learning algorithm. After various analysis the result has been generated and it is counter verified with some of the already available anti-keylogger tools.