Research Issues
We study on Visual Analytics Tools for Network Data to detect illegal accesses by interactively analyzing IP packets data.
We develop detecting systems for collaborative behavior with multi sources, bots by distributions of data transmission intervals.
We develop an effective and efficient graph mining system, which searches a huge amount of internet access logs for malicious cyber attacks by using various graph pattern mining techniques.
We propose the method to detect cyber attacks that are difficult to find by human-eyes using a powerful tool of machine learning and information theory.
We aim at detection of coordinated attacks by estimating important network structure based on statistical models which describe relationships between hosts.
We design efficient and robust algorithms for detection of cyber attacks using data compressions.
Research Group Page