摘要： 针对交易数据库中数据项重要性不同的现象，引入加权支持度和最小支持期望的概念，提出一种基于关联图的加权关联规则模型，并在该模型基础上，设计了改进的加权关联规则挖掘算法。该算法扫描数据库仅一次，采用关联图存储频繁2项集信息，通过构建基于图的剪枝策略，减少验证频繁项集的计算量，有效提高加权频繁项集的生成效率。

关键词:

mso-ascii-font-family: 'Times New Roman', mso-hansi-font-family: 'Times New Roman'">加权关联规则, 最小支持期望, 关联图layout-grid-mode: line, mso-bidi-font-size: 10.0pt, mso-fareast-font-family: 楷体_GB2312">

Abstract: By introducing the concept of weighted support and minimum support expect, a new model of weighted association rule mining is presented in order to solve the problem that items have not the same importance in datasets. Based on the model, a new improved algorithm for mining weighted association rules based on association graph is proposed. The algorithm only scans the database once, stores the frequent 2 itemsets with association graph, and builds an effective pruning strategy to reduce the computation. It improves the efficiency of weighted frequent itemsets generation.

Key words: weighted association rule, minimum support expectation, association graph

中图分类号: 

• TP311.52