Normalizing Security Events with a Hierarchical Knowledge Base - Information Security Theory and Practice Access content directly
Conference Papers Year : 2015

Normalizing Security Events with a Hierarchical Knowledge Base

David Jaeger
  • Function : Author
  • PersonId : 999010
Amir Azodi
  • Function : Author
  • PersonId : 999011
Feng Cheng
  • Function : Author
  • PersonId : 999012

Abstract

An important technique for attack detection in complex company networks is the analysis of log data from various network components. As networks are growing, the number of produced log events increases dramatically, sometimes even to multiple billion events per day. The analysis of such big data highly relies on a full normalization of the log data in realtime. Until now, the important issue of full normalization of a large number of log events is only insufficiently handled by many software solutions and not well covered in existing research work. In this paper, we propose and evaluate multiple approaches for handling the normalization of a large number of typical logs better and more efficient. The main idea is to organize the normalization in multiple levels by using a hierarchical knowledge base (KB) of normalization rules. In the end, we achieve a performance gain of about 1000x with our presented approaches, in comparison to a naive approach typically used in existing normalization solutions. Considering this improvement, big log data can now be handled much faster and can be used to find and mitigate attacks in realtime.
Fichier principal
Vignette du fichier
978-3-319-24018-3_15_Chapter.pdf (1.14 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01442546 , version 1 (20-01-2017)

Licence

Attribution

Identifiers

Cite

David Jaeger, Amir Azodi, Feng Cheng, Christoph Meinel. Normalizing Security Events with a Hierarchical Knowledge Base. 9th Workshop on Information Security Theory and Practice (WISTP), Aug 2015, Heraklion, Crete, Greece. pp.237-248, ⟨10.1007/978-3-319-24018-3_15⟩. ⟨hal-01442546⟩
88 View
1627 Download

Altmetric

Share

Gmail Facebook X LinkedIn More