期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2015
卷号：4
期号：7
页码：6131
DOI：10.15680/IJIRSET.2015.0407158
出版社：S&S Publications
摘要：Association rule mining is one of the well known data mining approach which extracts previouslyunknown relationships among attributes. Thus the discovered know-how can bestow competitive edge to enterprises inthe real world in making strategic decisions. Having said this, it is very expensive to extract frequent item sets andgenerate association rules using statistical measures such as support and confidence every time when the underlyingdatabase is updated. To get rid of the drudgery of reinventing the wheel for every modification in order to reflectpresent database many researchers contributed by proposing algorithms and data structures. When algorithm is costeffective and what is the underlying data structure with index support used are the questions arise. In this paper wereview the present state of the art of index support for frequent item set mining for dynamically updating associationrules. The insights found in the literature can be of use in making well informed decisions for maintaining associationrules and updating them automatically when database is subjected to changes such as insert, update, and delete.
关键词：Frequent item set mining; association rule mining; post mining of association rules; rule maintenance
