隨著網絡入侵技術的不斷發展, 入侵的行為表現出不確定性、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據挖掘的入侵檢測系統的設計和具體實現，也就是用于數據預處理和分類、聚類挖掘的數據挖掘技術。在數據預處理中，我們使用基于屬性抽取的方法去除干擾屬性。最后，我們對系統進行了測試，通過測試結果我們發現挖掘的效率和正確率，而系統確實能夠有效的檢測到已知未知攻擊。
關鍵字: 入侵檢測；數據挖掘；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