Abstract: Searching for frequent item sets is the most fundamental and important problem in mining association rules.An algorithm FP-SPMA for frequent item sets mining of sharing prefix based on FP-Tree is proposed.Constructing FP-Tree to compress business database，reduce candidate item sets rapidly by sharing prefix and pre-delete branches.Without recursion condition pattern tree，the algorithm performance has obvious raising.

Key words: frequent item sets, FP-tree, sharing prefix, FP-SPMA

摘要： 在数据挖掘中发现关联规则是一个基本问题，而发现频繁项集是关联规则挖掘中最基本、最重要的问题。提出了基于FP-Tree的共享前缀频繁项集挖掘算法－FP-SPMA算法。构造FP-Tree来压缩事务数据库，通过共享前缀和前瞻剪枝快速减小候选项集，无需递归构造条件模式树，算法性能有明显的提高。

关键词: 频繁项集, 高频繁模式树（FP-Tree）, 共享前缀, 基于FP-tree的共享前缀频繁项集挖掘算法（FP-SPMA）