数据可视化的标准数据集和顶级会议

在许多计算机科研领域，都有专门的标准数据，用以衡量科研工作是否具有创新性。例如，在图形图像处理领域用了几十年的美女回眸图，中文分词领域的Chinese Treebank（CTB5）数据集等等。但是在数据可视化领域中是否有类似的数据集以及评价标准呢？

由于数据可视化概念的宽泛性，涉及范围很广，因此测试数据集的类型就更广。例如气象数据可视化与一般商业公司销售数据可视化 从数据到案例结果都有巨大的不同。除去科学可视化领域的一些专门数据，对于普通的数据可视分析，目前获得普遍认可和使用的还是历届VAST chanllenge的数据集。

国际可视化年会IEEE VIS是可视化领域最顶级的学术会议，而IEEE VIS中的IEEE VAST Challenge是可视化与可视分析领域最重要、规模最大的竞赛。VAST Challenge举办的目的，一方面是为了提供易于获取的数据集。往年的VAST Challenge的数据集，以及当年参赛者提交的作品，都可以在马里兰大学的可视分析基准库中获取（http://hcil2.cs.umd.edu/newvarepository/benchmarks.php）。可视分析领域的研究者可以利用VAST Challenge中提供的公开的，逼真的数据集，以及数据集涉及的具体场景和分析任务的描述，开展研究工作。另一方面，也是为了帮助可视分析领域的研究者通过处理逼真的人造数据集，更好地理解现实的可视分析问题中的设计难点，发现可视分析中新的研究话题。在我看过的多篇论文中，都引用了VAST Challenge的数据做参考。之前我也接触过VAST Challenge的数据，下面就我所熟悉的数据做个说明：

Suggested topics for papers include, but are not limited to:

• Visual representations and interaction techniques including the principles for depicting information, new visual paradigms, statistical graphics, geospatial visualizations, the science of interaction, and approaches for generating visual analytic visualization and interactions.
• Data management and knowledge representation including scalable data representations for high volume and stream data, statistical and semantic signatures, and synthesis of information from diverse data sources.
• Mathematical foundations and algorithms for data transformations to allow interactive visual analysis.
• Analytical reasoning including the human analytic discourse, knowledge discovery methods, perception and cognition, and collaborative visual analytics
• Presentation, production, and dissemination methods including methods and tools for capturing the analytics process, methods for elicitation of stakeholder constraints, priorities & processes for incorporation in analysis, and storytelling for specific and varying audiences.
• Applications of visual analysis techniques, including but not limited to applications in science, engineering, humanities, business, public safety, commerce, and logistics as far as they contribute to visual analytics are of particular interest.
• Evaluation methods, visual ethical analysis such as privacy, security, & regulatory compliance, interoperability, and technology practice & experience.
• Discourse visualization and visual representations of the reasoning process.
• Algorithms and technologies which are fundamental for visual analytics, including user and device adaptivity, web interfaces and mobile or other novel devices.