Mining Positive and Negative Association Rules with Weighted Items
Many interesting rules are lost if the minimum support is high,but combinatorial explosion is likely to occur if the minimum support is small.To solve the problem,the traditional association rules are extended by allowing a weight to be associated with each item in a transaction to reflect interest or intensity.The positive association rules can be found in the past weighted association rules,whereas the negative weighted association rules are as important as the positive.Misleading rules maybe occur when simultaneously studying the positive and negative association rules.In this paper,an algorithm for mining the positive and negative weighted association rules (PNWAR) is proposed to solve the corresponding question.The correlation method is applied to the algorithm for mining the weighted association rules.Not only is the difference of the items in the database solved,but also the negative association rules are mined and the contrary rules are eliminated using the algorithm.An experiment is performed and the results of it show that the algorithm is efficient.
Association Rule Algorithm Apriori PNWAR Weight Frequent Itemset Correlation
He Jiang Yuanyuan Zhao Xiangjun Dong Shiju Shang
School of Information Science and Technology,Shandong Institute of Light Industry,Jinan,250353,China
国际会议
大连
英文
437-441
2008-07-27（万方平台首次上网日期，不代表论文的发表时间）