摘要
基于启发式规则的隐私保护关联规则挖掘算法中均通过删除项或增加项改变规则的支持度,对非敏感规则的支持度影响很大。针对上述不足,提出一种将删除项和增加项2种操作相结合的方法,在执行删除项操作后寻找合适的事务,对该事务执行增加项操作。实验结果表明,利用该算法清洗数据库所产生的规则丢失率和相异度均有所下降。
All the heuristic approaches are realized by deleting an item or inserting an item, of which the negative effect on non-restricted rules is too much. Focusing on the shortcoming, this paper presents an algorithm combining those two operations, which insert an item into proper transaction after deleting that item. Experimental results show that privacy preserving data mining algorithm based on item-moving has lower miss rate and dissimilarity.
出处
《计算机工程》
CAS
CSCD
北大核心
2009年第12期59-60,63,共3页
Computer Engineering
关键词
关联规则
隐私保护
数据挖掘
association rule
privacy preserving
data mining
