Doi: 10.1145/1435417.1435439
By Jeffrey Heer, Fernanda B. Viégas, and Martin Wattenberg
This article describes mechanisms for asynchronous collaboration in the context of information visualization, recasting visualizations as not just analytic tools, but social spaces. We contribute the design and implementation of sense.us, a Web site supporting asynchronous collaboration across a variety of visualization types. The site supports view sharing, discussion, graphical annotation, and social navigation and includes novel interaction elements. We report the results of user studies of the system, observing emergent patterns of social data analysis, including cycles of observation and hypothesis, and the complementary roles of social navigation and data-driven exploration.
Visual representations of information often lead to new insights by enabling viewers to see data in context, observe patterns, and make comparisons. In this way, visualizations leverage the human visual system to improve our ability to process large amounts of data. Card et al. 6 describe how visualization supports the process of sensemaking, in which information is collected, organized, and analyzed to form new knowledge and inform further action. They emphasize the ways visualization exploits an individual’s visual perception to facilitate cognition.
In practice, however, sensemaking is often also a social process. People may disagree on how to interpret the data and may contribute contextual knowledge that deepens understanding. As participants build consensus or make decisions they learn from their peers. Furthermore, some data sets are so large that thorough exploration by a single person is unlikely. This suggests that to fully support sensemaking, visualizations should also support social interaction. In this spirit, a recent report23 names the design of collaborative visualization tools as a grand challenge for visualization research.
These considerations are not just hypothetical. For example, the manager of a business group in our company described to us how quarterly reports are disseminated within his organization via e-mail. Heated discussion takes place around charts and graphs as the group debates the causes of sales trends and considers possible future actions. However, writing about particular trends or views is difficult, involving awkward references to attached spreadsheets from the e-mail text. Furthermore, the discussion is scattered and disconnected from the visualizations,
making it difficult for newcomers to catch up or others to review and summarize the discussion thus far. According to the manager of the group, the analysis process could benefit from a system for sharing, annotating, and discussing the visualized data.
Similar scenarios appear in other domains. Moreover, experiences with deployments of visualizations hint at ways that social phenomena already occur around visualizations. Wattenberg and Kriss27 describe the response to NameVoyager, an online visualization of historical baby name trends. Playful yet often surprisingly deep analysis appeared on numerous blogs as participants discussed their insights and hypotheses. Observing the use of a physical installation of the Vizster social network visualization, Heer18 noted that groups of users, spurred by storytelling of shared memories, spent more time exploring and asked deeper analysis questions than individuals. Similarly, Viégas et al. 24 found that users of the PostHistory e-mail archive visualization immediately wanted to share views with friends and family and engage in storytelling.
While suggestive, these observations provide only a circumstantial understanding of the social aspects of asynchronous analysis around visualizations. In the case of the NameVoyager and PostHistory, the findings were essentially accidental. Vizster was designed for playful interaction, but in a synchronous and less analytic context. It would therefore be valuable to replicate these findings to deepen our understanding of this type of interaction.
Furthermore, if social interaction is an important accompaniment to data visualization, it is natural to look for ways to support and encourage it. To address both these goals, we designed and implemented a Web site, sense.us, aimed at group exploration of demographic data. The site provides a suite of interactive visualizations and facilitates collaboration through view bookmarking, doubly linked discussions, graphical annotation, saved bookmark trails, and social navigation through comment listings and user profiles. We then conducted user studies to observe closely how people engage in social data analysis. The studies also allowed us to evaluate the new design elements in the site and suggest directions for future work.
A previous version of this paper was published in the Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, April 2007.
References:
Archives