文章基本信息

标题：An Architecture for Integrated Online Analytical Mining
本地全文：下载
作者：Usman, Muhammad ; Asghar, Sohail
期刊名称：Journal of Emerging Technologies in Web Intelligence
印刷版ISSN：1798-0461
出版年度：2011
卷号：3
期号：2
页码：74-99
DOI：10.4304/jetwi.3.2.74-99
语种：English
出版社：Academy Publisher
摘要：Several works in the past proved the likelihood and interest of integrating Online Analytical Processing (OLAP) with data mining and as a result a new promising direction of Online Analytical Mining (OLAM) has emerged. In this paper, a variety of OLAM architectures in the literature were reviewed and the limitations in the previously reported work have been identified. Literature review reveals the fact that none of the previously reported OLAM architectures have integrated enhanced OLAP with data mining. This work is an attempt to enhance OLAP performance in terms of cube construction time and to improve visualization by providing interactive visual exploration of the data cube. The main aim of this paper is to propose an integrated OLAM architecture that not only overcomes the existing limitations but also extends the architecture by adding an automation layer for the schema generation. In the proposed work, hierarchical clustering has been used as the data mining technique and three types of warehouse schemas namely were automatically generated. A prototype of automatic schema generation has been developed and schema generation algorithms have been provided. In addition to this, we implemented and deployed the proposed architecture. Validation has been done by performing experiments on real life data set. Experimental results prove that the proposed architecture improves the cube construction time, empowers interactive data visualization, automate schema generation, and enable targeted analysis at the front-end. With this seamless integration a higher degree of improvement is achieved which makes significant advancement in the modern OLAM architectures.
关键词：Automatic schema;cube;decision support system;data mining;data visualization;hierarchical clustering;OLAP