Intrusion detection using adaptive time-dependent finite automata

In intrusion detection system, signature discovery is an important issue, since the performance of an intrusion detection system heavily depends on the accuracy and abundance of signatures. In most cases, we have to find these signatures manually. This is a time-consuming and error-prone work. Some researchers apply data mining to the intrusion detection system. However, they are almost for anomal IDS detection. In this paper, we use a causal knowledge based on inference technique to discover useful signature for intrusion, and to raise the detection performance. The paper presents how Hsiao's sequential approach and finite automata are used in the causal knowledge acquisition and to support the causal knowledge reasoning process.