Welcome to the Intrusion Detection
System Homepage. This project is a collaboration with NCSU
and FIT
If you have any questions regarding this project or if you want
to get involved, please contact Shlomo Hershkop
shlomo@cs.columbia.edu
Updates:
- Weekly Meetings for Spring 2003:
For the entire group takes place every Wednesday
other week at 4pm. We will be sending out an email over the list. if you do
not recieve this email , please contact shlomo.
- An overview technical report summarizing our research is
available in the publication section of the website. It is called
"Real Time Data Mining-based Intrusion Detection". There is another
paper on "Adaptive Model Generation" which summarizes the systems
pary of our research.
Description:
This project is a data-mining based approach to detecting
intruders in computer systems. The project approaches the
intrusion detection problem from a data-mining perspective. Large
quantities of data are collected from the system and analyzed to
build models of normal behavior and intrusion behavior. These
models are evaluated on data collected in real time to detect
intruders.
There are 12 sub projects which together compose the Intrusion
Detection System Project:
- HOBIDS - Host Based Intrusion Detection System.
- HAUNT - Network Based Intrusion Detection System.
- DIDS - Distributed Intrusion Detection System.
- DW-AMG - Data Warehousing and Adaptive Model Generation.
- MEF - Malicious Email Filter
- FWRAP - File System Wrappers
- ASIDS - Advanced Sensors for IDS
- TAG - The Attack Group
- IDSMODELS - Intrusion Detection Models Generation
- IDSWATCH - Intrusion Detection Visualization
- DuDE - Denial-of-Service Detection and Response System
- Response - Automated Intrusion Detection and
Response Rule-Based System
Before IDS, our group used to work on the
JAM project.
