Group Perspectives for Voyant: A Speculative Design Approach to Prototyping

1. Introduction

Visual Matters is a SSHRC funded project, originally led by Stéfan Sinclair, that is imagining ways we can collaborate in the analysis and visualization of textual data on data walls and other large shared surfaces like data tables. We are documenting interface challenges, imagining designs and prototyping through experiments with a particular tool, Voyant-Tools (voyant-tools.org). We are using Voyant-Tools as a context for design thinking to come up with new ways to visualize humanities text analysis data. To come up with ideas, we used a speculative design process to help develop value-driven interfaces through which to explore different ways of knowing. Speculative design is a future oriented design process not aimed at efficiency and problem solving, but instead at imagining speculative futures and their objects in a way that provokes critical discussions about alternative possibilities (Dunne / Raby 2013). In doing this, we are challenging the view that visualizations are an objective way of representing data (Munzer 2014) by exposing and representing the subjectivities that come about during the process of data collection, analysis and representation (Drucker 2014).

Large non-projector based touch sensitive displays are relatively new and as with other new products their usage introduces some challenges (Pernice 2015) as well as affordances which might not be effectively utilized. One of these affordances is the space in front of the data wall which accommodates the presence of many people who can simultaneously interact through the touch sensitive display and which can be made sentient for other types of interactions.

2. Method

For this research which is an inquiry into new ways of thinking and doing through visualisation on data walls, we started by in-person testing of the use of a data wall and data table that run on a Windows 10 operating system to see what works or does not work with a generic system. By trying Voyant on a touch sensitive data wall we could learn from the affordance of being able to display and interact with visualizations in greater detail, but we also faced many interface challenges among which were physical exhaustion from the interaction on such a large surface. Further, some of Voyant’s user interface features don’t work on touch surfaces, and we found it difficult to use the shared virtual keyboard. Finally, it was difficult to implement boundaries between different groups of users simultaneously using the data wall. In the full presentation we will outline the types of interface features that don’t carry over from the PC interfaces to a data wall.

To come up with ideas of future usage of the data wall, we organised online speculative design workshops that started in the summer of 2020. Our objective was not simply to generate ideas of how to redesign the user interface but also to think of how to make use of the space around the data wall; how and with what to interact with the data wall; what type of interaction should the data wall promote among the people around it; what are contexts in which data walls could be used. We use a variety of methods to promote speculative thinking.

To promote ethical ideas we used values to guide the generation of speculative ideas, as a way to emphasize the metaphysical aspects of design ideas, rather than functional problem solving ideas. Some of the values we identified as mattering to humanities visualization were multiple perspectives (for example, how to represent what is left out from the analysis, how to remind users of the subjectivity of the data and visualisation); collaboration (how could we do collective text analysis, how do we communicate our analytical process); fun (how can we make using Voyant fun, what are funny visualisations, can we use Easter eggs); comfort (how can Voyant give feedback to users to facilitate usage, how do we make Voyant accessible to a variety of users, what’s the most relaxed way to use Voyant on a data wall). In the presentation we will show the values-driven ideas developed.

3. Outcomes

The next step of the research was to create working prototypes from the ideas generated and implement them in Voyant. To help us to choose the prototypes to develop, we imagined a use case scenario in which a lecturer and students were collectively analysing a text on a data wall. Over the course of the project, we created three prototypes: the Book Map which identifies mentions of locations in a text and traces the locations on a map with an animated path; Mobile Voyant which permits the control of Voyant through a mobile device by multiple users; and the Embedder which makes it possible to open webpages in a Voyant panel next to the other analytical panels. The prototypes bring new possibilities to collective analysis of text by groups of people. The two main contributions of the prototypes are that they address the issue of subjectivity of text analysis by bringing multiple perspectives into the analysis and also permit multiple sources of information to be displayed at the same time, and this can be seen as an attempt to bring in other data left out in the analysis. We will conclude the presentation with a demonstration of the prototypes.