Development of a keylogger malware incorporating machine learning techniques

Abstract

Due to the rapid globalization due to rapid advancement in technology, we have seen a drastic rise in the number of online users over the past ten years. Although technology has brought about many conveniences and benefits to our lives, we must also acknowledge the rise in the number of cyber-crimes. Recently, many innocent people have fallen prey to crimes involving the malicious usage of keyloggers, where hackers install keyloggers into victims’ devices to track their device activities to steal confidential information and passwords. Moreover, as technology advances at a rapid rate, cyber criminals are also evolving and developing more advanced methods to prevent themselves from getting caught by authorities. As such, it is becoming increasingly important to be aware on how to prevent ourselves from falling prey to such cyber-attacks by understanding how these malicious keyloggers work. The purpose of this project is to explore how machine learning can enhance the utility of a keylogger in order to understand how to not fall prey to such crimes. Through the usage of machine learning and Abstract Topic Modelling, the keylogger is able to decipher what general topics the key logs are about, which can allow hackers to piece different pieces of document together to retrieve the original document.