基于FP-tree和约束概念格的关联规则挖掘算法及应用研究  

Mining algorithm of association rule based on FP-tree and constrained concept lattice and application research

作　　者：付冬梅[1] 王志强[1] 

机构地区：[1]北京科技大学自动化学院,北京100083

出　　处：《计算机应用研究》2014年第4期1013-1015,1019,共4页Application Research of Computers

基　　金：国家自然科学基金重点资助项目(51131005);国家科技基础性工作专项基金资助项目(2012FY113000)

摘　　要：现有的基于频繁模式树FP-tree和概念格的规则挖掘算法在构造概念格时存在重复遍历FP-tree问题,在挖掘后件约束的规则时算法构造的概念格包含冗余结点。针对这两个问题,提出了通过遍历FP-tree生成候选概念格节点的策略,并根据候选概念格节点进一步构造规则约束条件下无冗余概念格。通过实际项目中大气腐蚀数据进行算法的应用,结果表明该算法比现有算法具有更高的挖掘效率且腐蚀规则结果对材料腐蚀现状研究具有重要指导价值。In existing literatures, the rule mining algorithms based on FP-tree and concept lattice repeat scan FP-tree when generating the concept lattice, and the concept lattice contains redundant nodes when mining rules with constraint condition. To solve the proldems, this paper proposed a new algorithm, which generated the candidate concept lattice nodes by scanning FP-tree and then created the constrained concept lattice based on the rule constraint. It tested the algorithm by atmospheric corrosion data of actual project. The result shows that the proposed algorithm has much superior performance in efficiency to the original algorithm and has the important consultant functions to the research of material corrosion status.

关 键 词：规则挖掘 频繁模式树 约束概念格 材料腐蚀 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]