Accession Number:

AD1066792

Title:

Nave Bayes Log File Reduction and Analysis

Descriptive Note:

Technical Report,07 Aug 2017,10 Jul 2018

Corporate Author:

US Army Research Laboratory Aberdeen Proving Ground United States

Report Date:

2019-01-01

Pagination or Media Count:

47.0

Abstract:

The application of Bayes theorem in computer science dates back to the 1960s and continues to be heavily used in Nave Bayes classifiers in machine learning. In this report, we propose the use of a Nave Bayes-based classifier for automated analysis and data reduction of text-based log files generated by various computer systems and the services they provide. The intended application of this technique is to automate the reduction of voluminous log files to a more manageable size and, with reasonable accuracy, retain log lines containing potential indicators of malicious cybersecurity activity or other infrequent interesting activity that should be examined further through other means.

Subject Categories:

  • Computer Programming and Software
  • Statistics and Probability
  • Computer Systems Management and Standards

Distribution Statement:

APPROVED FOR PUBLIC RELEASE