摘要： 以Hadoop为代表的可扩展大规模数据库难以进行多维可视化分析。为此，设计基于B/S架构的可视化分析框架Bizard。数据模型通过封装底层数据接口以支持业界多维数据访问协议XMLA，从而在展现层易于接入支持XMLA的传统分析工具，同时采用视图物化技术提高分析性能，利用互联网技术丰富用户分析体验。实验结果表明，该框架能在高达千万条记录级的数据上进行多维可视化分析。
关键词:
数据仓库,
可视化分析,
Hadoop软件,
大规模数据,
XMLA协议
Abstract: To address the multi-dimensional visualization analysis problem of large scale data storages with high scalability represented by Hadoop, this paper designs and implements a visualization analysis framework, called Bizard. Bizard’s data model encapsulates the underlying data access interface and provides XMLA protocol for the presenting layer, which makes it convenient to use conventional analytical tools. Meanwhile, Bizard uses materialized view technologies to improve query performance and the RIA technology to enrich the user analysis experience. Experimental results show that the framework can process multi-dimensional visualization analysis on large data sets with the number of rows up to tens of millions.
Key words:
data warehouse,
visualized analysis,
Hadoop software,
large scale data,
XMLA protocol
中图分类号:
游进国, 杨卓荦, 胡建华, 奚建清. 一种支持大规模数据的多维可视化分析框架[J]. 计算机工程, 2011, 37(19): 26-27,31.
LIU Jin-Guo, YANG Zhuo-Luo, HU Jian-Hua, XI Jian-Qing. Multi-dimensional Visualized Analysis Framework for Supporting Large Scale Data[J]. Computer Engineering, 2011, 37(19): 26-27,31.