Intrusion Detection Model Based On Association Rules

With the emergence of the bank on network and electronic commerce, the Internet is more and more quickly entering the each aspect of social life. Fallowing are all kinds of network attacks. More and more people thought of the importance of network security.IDS is a technology which can protect our information. It can monitor our systems or network, and find intrusions. The research and development of IDS have become an important subject about network security. To overcome the insufficiency of the ID, we systemically study on the automatic modeling of intrusion detection, research algorithm, optimize algorithm and so on.The thesis begins its discussion by brief introducing the concept of IDS, the classification of the IDS, the idea of data mining, the notion of association rules. Make a brief outline about the Apriori algorithm in association rules, and analyze and research the shortage on applying it to the domain of ID. Then we study on how to analyze the data effectively and automatic modeling in this field after collecting the primitive data, and this is the key point of research. Based on the research above, a kind of ID module based on association rules that have relationship analysis ability is proposed. The method used the results in aspect of finding association rules. The algorithm that improved association rules is proposed, and this algorithm has been implemented the automatic establish modeling. Finally, we design, analyze, implement the core component of the ID that cooperates with association rules. And we has obtained a quite ideal experimental result.