尚未实现吗?审查现有可视化建议系统的概念理论和实验支持 (Are We There Yet? A Review on Existing Perceptual Theory and Experiment Support for Visualization Recommendation Systems)

A growing body of research focuses on helping users explore complex datasets faster by automatically suggesting visualization designs of possible interest. However, existing visualization recommendation systems only enumerate, rank, and recommend a small group of visualization designs. Our goal is to understand whether there is enough theoretical and experimental knowledge in current literature to inform visualization recommendation systems to assess the entire visualization design space. Thus, in this paper, we present a literature review comparing and ranking the quality of visualization designs in visual perception and human performance. We structure our review by first defining the visualization design space where visualizations must be compared to recommend effective visualization designs. We then perform the review by using a comprehensive schema to record the theoretical and experimental results of visualization comparison, which can also be used to guide the future construction of visualization recommendation systems. To analyze the literature coverage, we develop an interactive tool that can help explore current literature coverage of visualization comparison and identify gaps efficiently and effectively. Based on our findings, we highlight new opportunities and challenges for the community in working towards a comprehensive visualization ranking for informing visualization recommendation systems.

翻译：越来越多的研究侧重于帮助用户通过自动建议可能感兴趣的可视化设计,更快地探索复杂的数据集。然而,现有的可视化建议系统仅罗列、排行和建议一组可视化设计。我们的目标是了解当前文献中是否有足够的理论和实验知识,为可视化建议系统提供信息,以评估整个可视化设计空间。因此,我们在本文件中提出文献审查,比较视觉化设计的质量和人类性能,并对其进行排序。我们通过首先确定可视化设计空间,将可视化设计与建议有效的可视化设计进行比较。然后,我们利用一个全面的系统来进行审查,以记录可视化比较的理论和实验结果,这也可用于指导未来可视化建议系统的构建。为了分析文献覆盖面,我们开发了一个互动工具,可以帮助探索当前可视化比较的文献覆盖面,并高效和有效地查明差距。根据我们的研究结果,我们强调社区在努力为可视化建议系统提供全面的可视化排名方面所面临的新机遇和挑战。