Visual Editing and Comparison of Multivariate Graphs Using Multiple Interactive Displays (GEMS 2.0)

数据无处不在。如今，越来越多的智力工作是基于测量、收集、建模、合成或模拟的数据。这个建议关注的是一类特别重要的数据：图表。图形在许多应用领域都是普遍相关的，比如通信、交通、生物学和气候学。在这些领域中，对方法和工具的需求越来越大，这些方法和工具不仅支持处理图的结构（即它的节点和边），而且还支持处理可能与图相关的众多数据属性。全面理解这种多变量图对于使用它们的人来说是一个重大挑战。处理数据通常意味着分析它们。交互式可视化分析已被证明是深入了解数据的有效手段。然而，除了理解图的结构和属性的基本信息之外，还有更多的东西要使用图。在本提案中，我们计划研究如何将经典的探索性图分析推进到包括多个图的比较分析，以及在必要时直接编辑图的节点和边及其相关属性的可能性。这种多变量图的组合可视化探索、比较和编辑是一项复杂的工作，因为它涉及数据的各种关联的可视化表示和与它们交互的各种方式。然而，典型的数据工作场所仍然由鼠标加键盘交互的经典桌面界面组成，这被证明对复杂的图形分析是无效的。为了支持在多变量图上流畅无缝地工作，需要更好、更自然的用户界面。因此，在本提案中，我们将研究自然形式的交互，如触摸、以身体为中心和基于凝视的交互，以及由大型和移动交互显示器组成的多显示环境。总之，这个项目的重点是开发新的和基本的概念、方法和工具，用于在多显示环境中比较和编辑具有新颖交互模式的多个多元图形。

Data is everywhere. Nowadays, an increasing amount of intellectual work is based on measured, collected, modeled, synthesized, or simulated data. This proposal focuses on a particularly important class of data: graphs. Graphs are universally relevant in many application domains, such as communication, transportation, biology, and climatology. In these domains, there is an increasing demand for methods and tools that do not only support working with a graph's structure, that is, its nodes and edges, but also with numerous data attributes that may be associated with the graph. Understanding such multivariate graphs comprehensively is a significant challenge for the people working with them.Working with data usually means analyzing them. Interactive visual analysis has proved to be an effective mean to gain insight into data. Yet, there is more to working with graphs than understanding basic information of their structure and attributes. In this proposal, we plan to study how classic exploratory graph analysis can be advanced to include comparative analysis of multiple graphs and the possibility to directly edit a graph's nodes and edges as well as its associated attributes when necessary.Such a combined visual exploration, comparison, and editing of multivariate graphs is a complex endeavor, because it involves various linked visual representations of the data and diverse ways of interacting with them. The typical data workplace, however, still consists of classic desktop interfaces with mouse plus keyboard interaction, which proved to be ineffective for complex graph analysis. To support smooth and seamless work on multivariate graphs, better and more natural user interfaces are needed. Therefore, in this proposal we will investigate natural forms of interaction, such as touch, body-centric, and gaze-based interaction in concert with multi-display environments consisting of large and mobile interactive displays.In summary, this project focuses on developing novel and fundamental concepts, methods, and tools for comparing and editing multiple multivariate graphs with novel interaction modalities in multi-display environments.