信息可视化自动生成技术的研究

信息可视化是辨识复杂数据模式的重要手段，为了降低技术门槛，人们开发了一系列标准化的可视化组件，极大推广了可视化技术的应用。但可视化创作效率仍然比较低下，无法满足数据新闻、商务智能等领域中实时性大规模图表创作方面的需求。为解决该问题，本项目围绕“自动可视化设计与生成”这一可视化领域的国际前沿课题展开研究，拟设计并开发一个可视化叙事内容自动生成系统, 用于根据用户输入的结构化数据自动设计并生成具有叙事情感表达能力的可视化图表。通过本项目申请人拟回答，如何根据多维度数据自动生成具有情感表达能力的可视化图表，如何自动生成具有逻辑的可视化叙事内容等科学问题，探索可视化情感设计范式，完善针对多维度数据的可视化自动生成技术，并构建自动可视化叙事内容生成系统。项目分为成功案例调研及情感可视化设计空间构建，自动可视化生成关键技术攻关，及系统集成三个阶段，预期产出原型系统1个，高质量国际论文15篇，专利2篇。

Information visualization is an efficient way to help people identify complex data patterns. In order to lower the technical barrier, a number of standardized visualization components have been developed, which greatly promoted the application of visualization technology. However, the efficiency of visualization creation process is still low, and it cannot meet the needs of real-time large-scale chart-creation requirements in the fields of data journalism and business intelligence. In order to address the problem, this project focuses on a frontier topic in the field of information visualization: the "automatic visualization chart design and generation". It is planned to design and develop a visual narrative content generation system, which is used to automatically design narrative visualizations based on the input data. Through this project, we intend to answer scientific questions such as “how to automatically generate visual charts with emotional expression capabilities based on multi-dimensional data” and “how to automatically generate logical visual narrative content”. We will explore the emotional visual design paradigms, and develop techniques for generating narrative visualization charts from a multidimensional data. The project consists of three phases: (1) successful case investigation and emotional visualization design space construction;(2) study on the automatic visualization generation, and system integrations. Through the project, we expected to develop 1 prototype system, publish 15 high-quality papers, and file 2 patents.