期刊名称：International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
印刷版ISSN：2278-1323
出版年度：2014
卷号：3
期号：10
页码：3433-3436
出版社：Shri Pannalal Research Institute of Technolgy
摘要：The Frequent pattern mining is one of the important tasks used in data mining domain and frequent data mining approaches are widely applied onto static database as well as data stream but the data have been accumulated more quickly in recent years and the corresponding databases have also become huger, and thus, general frequent pattern mining methods have been faced with limitations that do not appropriately respond to the massive data, so it is necessary to conduct more efficient and immediate mining tasks by scanning databases only once. In this paper, we analyze the different sliding window approach which can perform mining operations focusing on recently accumulated parts over data streams and mine all of the frequent patterns in the data stream environment and efficiently compressing generated patterns to solve that problem. In addition, we focus on different algorithm which not only use support conditions but also weight constraints expressing items and mining weighted maximal frequent patterns over sliding window model-based data streams, and also we survey the different approaches which always extract mining results regarding the latest data over data streams, and can gain the resulting patterns more quickly through the maximal frequent pattern technique and weight conditions.
