模式匹配是基于攻擊特征的信息過濾系統中的網絡數據包分析技術，匹配算法的性
能直接影響到整個系統的效率，是當前信息過濾監測系統的一個主要瓶頸，因此以速度較快的BM算法為基礎，提出了一種改進的字符串匹配算法，充分考慮模式匹配失敗的信息，使其在每一次跳躍中跳過盡可能大的距離。通過實驗證明了改進的算法減少了匹配的次數，具有更高的效率。
關鍵詞: 信息過濾；BM 算法;改進算法;模式匹配
Abstract: The pattern matching is one kind of network data package of parsing technique in the information filtration system which is based on the attack characteristic,matching algorithms performance immediately influence to the efficiency of the overall system,which is a main bottleneck in the current information filtration and observation system.so taking BM algorithm whose speed is quite quick as the foundation an improved algorithm is proposed,using the pattern matching defeat's information, causes to jump over the great distance as far as possible in each caper. Had proven through the experiment the improvement algorithm reduced the match number of times and had a higher efficiency.
Key words: information filtration;BM algorithm; improved algorithm; pattern match