基于數(shù)據(jù)挖掘的入侵檢測系統(tǒng)設計和實現(xiàn)

隨著網(wǎng)絡入侵技術的不斷發(fā)展, 入侵的行為表現(xiàn)出不確定性、復雜性、多樣性等特點，
入侵檢測面臨許多有待解決的關鍵問題。本文詳細介紹了基于數(shù)據(jù)挖掘的入侵檢測系統(tǒng)的設計和具體實現(xiàn)，也就是用于數(shù)據(jù)預處理和分類、聚類挖掘的數(shù)據(jù)挖掘技術。在數(shù)據(jù)預處理中，我們使用基于屬性抽取的方法去除干擾屬性。最后，我們對系統(tǒng)進行了測試，通過測試結(jié)果我們發(fā)現(xiàn)挖掘的效率和正確率，而系統(tǒng)確實能夠有效的檢測到已知未知攻擊。
關鍵字: 入侵檢測；數(shù)據(jù)挖掘；信息安全
Abstract: As the development of the network intrusion technology, the action of intrusion
represents variable, complicated, and uncertainty characteristic. Therefore, it faces so many
problems to resolve for intrusion detection. We introduce the design and implementation of the system which adopt the Data Mine technology of data pretreatment, classify and cluster patterns mine. We use the method based on features selection to get rid of noises. Finally, we give a test to the system. Through the test, we find that our intrusion detection system is efficient to detect known and unknown attack.
Keywords: Intrusion Detection; Data Mining; Information Security