solr
SolrQueryCompletionProxy
QueryCompletionProxy

Bitte aktivieren Sie JavaScript in Ihrem Browser, damit Sie unseren Katalog nutzen können.

Applying Hadoop for log analysis toward distributed IDS

  • drucken Drucken
  • E-Mail Versenden
  • lokal speichern Speichern
  • Permalink
    /TouchPoint/statistic.do
    statisticcontext=result&action=permalink
  • Lesezeichendienste Lesezeichendienste
    /TouchPoint/statistic.do
    statisticcontext=result&action=bookmark

Aufsätze & mehr (1/1)

Speichern in:

Applying Hadoop for log analysis toward distributed IDS

Therdphapiyanak, Jakrarin , Piromsopa, Krerk
In: Proceedings of the 7th International Conference on ubiquitous information management and communication, 2013, p.1-6
ACM
ISBN 1450319580 , 9781450319584
Volltext-Verfügbarkeit online prüfen
Volltext
Aufsatz

  • Exemplare
    /TouchPoint/statistic.do
    statisticcontext=fullhit&action=holding_tab
  • Das möchte ich haben
    /TouchPoint/statistic.do
    statisticcontext=fullhit&action=availability_tab
  • mehr zum Titel
    /TouchPoint/statistic.do
    statisticcontext=fullhit&action=availability_tab
  • Rezensionen
    /TouchPoint/statistic.do
    statisticcontext=fullhit&action=allreviews_tab
Autor:Therdphapiyanak, Jakrarin
Autor:Piromsopa, Krerk
Titel:Applying Hadoop for log analysis toward distributed IDS
In:Proceedings of the 7th International Conference on ubiquitous information management and communication, 2013, p.1-6
Verlag:ACM
ISBN:1450319580
ISBN:9781450319584
Zusammenfassung:In this paper, we apply Hadoop for large-scale log analysis. Our main objective is to efficiently detect an abnormal traffic from high volume data. Due to the high volume of data traffics, the size of traffic logs is usually exceed the capacity of a standalone IDS. Thus, it is practically impossible to perform useful analysis with these data. In most cases, an analysis is usually done when an attack occurred for digital forensics.We proposed applying K-Means algorithm to cluster high volume log data. The resulted clusters are useful in classifying minority as possible intruders. In addition, we proposed IP address summarization method to capture the characteristic of each cluster.Our implementation allows high volume data traffics to be analyzed with a distributed analysis system using K-Means algorithm and data mining. The eventual result is to reduce a chance of being attacked. The prominent points of our implementation are anomaly detection with large file sizes and the distributed processing.However, this paper is just a preliminary study. There exist several opportunities for optimization. Nonetheless, our implementation can point out anomaly. The K-Means Algorithm can provide a new knowledge useful for enhancing security of the system.
Schlagwörter:distributed log analysis / log analysis / log analysis using Hadoop / map-reduce
Volltext (Primo):http://dx.doi.org/10.1145/2448556.2448559