Keystroke Logging: Integrating Natural Language Processing Technique to Analyze Log Data

Abstract

Cyberwarfare is observed very frequently as always some or the other country is targeting to ruin its enemy country by hacking confidential data from vital computer systems. This has led to dangerous international conflicts. Hence, to avoid illicit entry of other than military person or a government official several tools are being used today as spyware. Keyloggers are one of the prominent tools which are used in today’s world to obtain secret or confidential data of a legitimate and contradictory a malicious user too. These keyloggers are advantageous and taken up positively for monitoring employee productivity, for law enforcement and the search for evidence of the crime. While it’s negative illegitimate use includes data theft and passwords. The keylogger is today witnessed as a malicious attack and is looked upon as a security threat. But every coin has two sides. Keylogger actually helps in avoiding several security breaches and also aids in detecting several crimes across the net world followed by other fellow countries. This fact has motivated to write this paper and as a consequence, an experimental analysis too was carried out in order to conclude that keyloggers’ log file helps identify the person by analyzing proper pattern of the words entered in the file. This paper focuses majorly on the aspect of natural language processing, where a log file obtained thru keylogger software is thoroughly processed via the algorithm as described in the paper. The results yielded a fair understanding of the results obtained as one can easily identify the words used and on the basis of that can also know the type of person on the other end with his ideas, malicious one or of a legal kind.