DocumentCode :
3772423
Title :
Hadoop-Based System Design for Website Intrusion Detection and Analysis
Author :
Xiaoming Zhang;Guang Wang
Author_Institution :
Dept. of Comput., Beijing Inst. of Petrochem. Technol., Beijing, China
fYear :
2015
Firstpage :
1171
Lastpage :
1174
Abstract :
It is very important and practical to make data analysis for intrusion detection based on large scale data. For the current system problem in simulation and off-line analysis, a set of system is proposed as intrusion detection and analysis for truly website. The system is integrated with two subsystems of intrusion detection and large data analysis. Through network construction and software design, the system achieved functions of intrusion detection, data storage, data cleaning and data mining. It can make large data analysis under Hadoop system and Mahout Framework. Firstly, the well-known KDD data is studied to get the relationship of training entries and testing accuracy. The 10% of the training samples presents enough data in classification algorithm. Under the monitoring system, experimental results show that the system can quickly establish the random forest model. It can achieve good detection effects.
Keywords :
"Intrusion detection","Training","Monitoring","Cloud computing","Algorithm design and analysis","Data mining","Testing"
Publisher :
ieee
Conference_Titel :
Smart City/SocialCom/SustainCom (SmartCity), 2015 IEEE International Conference on
Type :
conf
DOI :
10.1109/SmartCity.2015.231
Filename :
7463886
Link To Document :
بازگشت