Ens Barrett, Goodwin Sarah, Prouzeau Arnaud, Anderson Fraser, Wang Florence Y, Gratzl Samuel, Lucarelli Zac, Moyle Brendan, Smiley Jim, Dwyer Tim
IEEE Trans Vis Comput Graph. 2021 Feb;27(2):1193-1203. doi: 10.1109/TVCG.2020.3030334. Epub 2021 Jan 28.
Collaborative visual analytics leverages social interaction to support data exploration and sensemaking. These processes are typically imagined as formalised, extended activities, between groups of dedicated experts, requiring expertise with sophisticated data analysis tools. However, there are many professional domains that benefit from support for short 'bursts' of data exploration between a subset of stakeholders with a diverse breadth of knowledge. Such 'casual collaborative' scenarios will require engaging features to draw users' attention, with intuitive, 'walk-up and use' interfaces. This paper presents Uplift, a novel prototype system to support 'casual collaborative visual analytics' for a campus microgrid, co-designed with local stakeholders. An elicitation workshop with key members of the building management team revealed relevant knowledge is distributed among multiple experts in their team, each using bespoke analysis tools. Uplift combines an engaging 3D model on a central tabletop display with intuitive tangible interaction, as well as augmented-reality, mid-air data visualisation, in order to support casual collaborative visual analytics for this complex domain. Evaluations with expert stakeholders from the building management and energy domains were conducted during and following our prototype development and indicate that Uplift is successful as an engaging backdrop for casual collaboration. Experts see high potential in such a system to bring together diverse knowledge holders and reveal complex interactions between structural, operational, and financial aspects of their domain. Such systems have further potential in other domains that require collaborative discussion or demonstration of models, forecasts, or cost-benefit analyses to high-level stakeholders.
协作式可视化分析利用社会交互来支持数据探索和理解。这些过程通常被设想为专业专家团队之间正式的、扩展的活动，需要具备复杂数据分析工具的专业知识。然而，有许多专业领域受益于对具有广泛知识的利益相关者子集之间短时间“突发”数据探索的支持。这种“临时协作”场景将需要引人入胜的功能来吸引用户的注意力，以及直观的“即拿即用”界面。本文介绍了Uplift，这是一个与当地利益相关者共同设计的新型原型系统，用于支持校园微电网的“临时协作式可视化分析”。与建筑管理团队的关键成员进行的一次启发式研讨会表明，相关知识分布在他们团队的多个专家手中，每个人都使用定制的分析工具。Uplift在中央桌面显示器上结合了一个引人入胜的3D模型、直观的有形交互以及增强现实、空中数据可视化，以支持对这个复杂领域的临时协作式可视化分析。在我们的原型开发期间及之后，对来自建筑管理和能源领域的专家利益相关者进行了评估，结果表明Uplift作为一个引人入胜的临时协作背景是成功的。专家们认为这样的系统有很大潜力将不同的知识持有者聚集在一起，并揭示其领域在结构、运营和财务方面之间的复杂相互作用。这样的系统在其他需要向高级利益相关者进行协作讨论或展示模型、预测或成本效益分析的领域还有进一步的潜力。
