计算机工程与应用 ›› 2011, Vol. 47 ›› Issue (32): 23-26.

• 博士论坛 • 上一篇    下一篇

可伸缩的网格资源查找可视化方法研究

仲倩黎,陆忠华,迟学斌,张宏海   

  1. 中国科学院 计算机网络信息中心 超级计算实验室,北京 100190
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2011-11-11 发布日期:2011-11-11

Study on scalable visualization for grid resource searching

ZHONG Qianli,LU Zhonghua,CHI Xuebin,ZHANG Honghai   

  1. Supercomputing Center,Computer Network Information Center,Chinese Academy of Sciences,Beijing 100190,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2011-11-11 Published:2011-11-11

摘要: 受笛卡尔坐标系的制约和资源数量的影响,大量资源的多变量显示与分析成为网格资源查找的难点之一。提出了一种基于平行坐标可视化技术的查找新方法,将资源多维状态数据映射到二维空间。提出了新的可视化度量:资源覆盖度、资源重叠度和资源相似度。对资源的可视化效果及聚类质量的评价及优化问题进行了研究。新方法克服了传统查找方法可伸缩性差的缺点。基于新度量的可视化可以动态调整坐标轴的数据分布,从视图结构角度对聚类质量进行定量分析。

关键词: 网格, 资源查找, 平行坐标, 可视化

Abstract: Restricted by Cartesian coordinate system and affected by the amount of resource,the multivariable visualization and analysis of a large number of resources become difficult points during grid resource searching.A new visualization method for resource searching based on parallel coordinate is presented to map multi-dimensional data to two-dimensional space.Three new visualization measures are given,coverage of resource,overlap of resource and similarity of resource.The assessment and optimization of visualization result and clustering quality are studied.New method overcomes the drawback of unscalability of traditional methods to handle growing amounts of resource.The visualization based on new measurements dynamically adjusts the data distribution on coordinate axis and quantificationally assesses the clustering assessment.

Key words: grid, resource search, parallel coordinate, visualization