基于信息可视化技术的数据检索是一个新的研究领域。在地理空间元数据检索过程中,多维可视化技术将有助于显示检索过程,改进检索结果,提供数据间隐含信息来提高检索精度。本文基于地理空间元数据的特点,对三种可用的多维可视化技术:平行坐标、星状图和散点图矩阵的形式和特点进行了分析。同时研究了地理空间元数据检索的关键技术:可视化技术、交互浏览、数据组织和检索算法。最后探讨了可视化检索的评估理论和方法。
It is an new research field for data retrieval based on information visualization technique. In the process of geospatial metadata retrieval, multivariate visualization will display the retrieval process, offer implicit information among data to improve the retrieval result and accuracy. First, this paper introduces three usable visualization technologies: parallel coordinates, star plots and scatter plot matrix, which based on the characteristics of data retrieval. Second, It pay more attention to the key issues: visualization technique, interactive browse, data organization and retrieval algorithm. Finally, we analyse the evaluating usability of visual retrieval for improving the design of retrieval process.