Distributed Uncertain Data Mining for Frequent Patterns Satisfying Anti-monotonic Constraints
Distributed Uncertain Data Mining for Frequent Patterns Satisfying Anti-monotonic Constraints
复制标题
DOI：
10.1109/waina.2014.11
复制
发表时间：
2014-05
期刊：
影响因子：
--
通讯作者：
C. Leung;Richard Kyle MacKinnon;Fan Jiang
中科院分区：
文献类型：
--
作者：
C. Leung;Richard Kyle MacKinnon;Fan Jiang
在生物、医学和生命科学的实际应用中，大量的不确定数据会在分布式环境中产生。频繁模式挖掘是一项重要的数据挖掘任务，它有助于从这些分布式数据库中发现频繁共存的项、对象或事件。然而，用户可能只对从这些数据库中挖掘出的所有频繁模式中的一小部分感兴趣。在本文中，我们提出了一个智能计算系统，（i）允许用户表达他们的兴趣，通过使用用户指定的约束和（ii）有效地利用用户指定的约束的反单调特性，并有效地发现频繁模式，满足这些约束从分布式数据库包含不确定的数据。
High volumes of uncertain data can be generated in distributed environments in many real-life biological, medical and life science applications. As an important data mining task, frequent pattern mining helps discover frequently co-occurring items, objects, or events from these distributed databases. However, users may be interested in only some small portions of all the frequent patterns that can be mined from these databases. In this paper, we propose an intelligent computing system that (i) allows users to express their interests via the use of user-specified constraints and (ii)effectively exploits anti-monotonic properties of user-specified constraints and efficiently discovers frequent patterns satisfying these constraints from the distributed databases containing uncertain data.