Development of a Next-Generation Visualization Systems for Information Mining

在从自然科学到人文社会科学的广泛学科中，三维计算机图形可视化在有效理解目标物体的复杂结构和行为方面起着不可或缺的作用。模块化可视化环境（MVEs）基于数据流范式，支持可视化和非过程化编程，被视为当代可视化系统的实际标准。然而，对于一般用户来说，选择和连接预先存储的模块来创建有效的应用程序网络是一项困难的任务，因为他们不是可视化专家，也不知道如何将他们的数据集映射到最佳的图形形式。在本研究中，开发了一种新的MVE，称为GADGET，它利用可视化专家的知识和专业知识来帮助用户根据自己的可视化目标设计模块网络。此外，通过开发基于MWMD的工作台，支持复杂多媒体数据库查询的设计和校园局域网流量管理专用信息可视化系统的设计，为构建下一代信息挖掘可视化系统提供了丰富的知识储备。

Three-dimensional computer graphics visualization plays an indispensable role in effective understanding of complex structure and behavior of target objects in the wide spectrum of disciplines ranging from natural sciences, to humanities and social science. Modular visualization environments (MVEs), which are based on the dataflow paradigm allowing for visual and non-procedural programming, are viewed as the de fact standard of contemporary visualization system. However, it is a difficult task for the general users to select and connect pre-stored modules to create effective application networks, since they are not visualization experts, and do not know how to map their datasets to best graphical forms. In this research, a new MVE, called GADGET, has been developed, which utilizes the set of visualization experts' knowledge and expertise to assist the user to design module networks according to his/her visualization goals. In addition, through the development of an MVE-based workbench, called MWMD, for supporting the design of complex multimedia database queries and a specific information visualization system for campus LAN traffic management, a rich store of knowledge has been obtained to figure out the next- generation visualization system for information mining.