摘要
传统的关联规则挖掘不能发现具有潜在价值的关联规则,如在挖掘交易数据库时,一些包含新商品的关联规则往往由于其信任度低而被删除,但是外部环境的动态性使得这些规则在某些特定时期对用户有很大的价值性。为了解决这个问题,保留具有潜在价值的关联规则,文中提出了基于外部环境的关联规则数据挖掘方法。在挖掘过程中,重新定义了信任度,并提出经济价值度的概念,根据信任度和经济价值度,可以有效地实现关联规则冗余性大小的排序,保留具有潜在价值的关联规则,适应用户的需求。实验表明,该方法可以有效地保留具有潜在价值的关联规则。
The association rules that have potential value can＇t be found by traditional association, rules mining. For example, when in the transaction database mining, some association rules that contain new goods information have always been removed, because of its low confidence. But these rules may have potential value. Users can＇t get these deleted rules But if the external environment changed, these rules may be useful to users. In order to solve this problem and reserve the rutes Contain potential Value, put forward a new method named asso- ciation rules mining based on the external environment. Redefined the confidence in the proposed method,and put forward the degree of economic value. According to this method can reserve the rules contain potential value and fulfill the ranking of association rules, then meet the demand of user. The experimental results show that the method reserves the rules contain potential value more efficiently..
出处
《计算机技术与发展》
2013年第1期115-118,共4页
Computer Technology and Development
基金
辽宁省自然科学基金项目(20102153)
关键词
关联规则
信任度
经济价值度
潜在价值
association rule
confidence
the degree, of economic value
potential value
