Xia Lin
Drexel University, USA
July 3, 2007, in the suburb of the beautiful city Zurich, I attended and witnessed the information summit on the campus of ETH Zurich Science City. More than 100 selected scientists, artists, architects and other participants engaged in a one-day intensive discussion and debate on the future of information visualization. This is an eye-opening event for all visualization researchers and will have a significant impact on next few years’ research in a variety of topics related to information visualization. The summit was divided into 9 groups:
Each group was to come up with a specifc research goal that is achievable by 2010. The results and reports of each group can be found from the above links. Here are some of my own impressions of the Summit.
A main group discussion I attended is the Software Infrastructures Group. Led by the group leaders, Dr. Katy Börner, Dr. Bruce Herr, and Dr. Jean-Daniel Fekete, the group focused its discussion on how to promote interoperability among more than 100 visualization APIs and more than 200 visualization algorithms. As if to demonstrate the point, each participant presented one of their own systems such as Cytoscope for information visualization in biology, EvoBerry for email visualization, tYNA for biological network visualization, PAJEK for social network visualization, etc. All these systems are promising; all of them address a significant problem of information visualization; all have different visualization styles; none share the same data structures or APIs, and none can work interoperatively with each other.
The solution proposed by the group is to go back to the data structure level to unify systems through better defined semantic metadata, APIs, and modularized implementations of visualization algorithms. The one-sentence goal they reach is: “by 2010 we will have a sound architectural model for Information Visualization and open source implementations to quickly build visualization applications suited to a wide variety of users and domains.”
This is a noble goal for information visualization at current stage. It emphasizes the open source software architecture that allows “plug-and-play” for different algorithms and different types of data sets. It is very challenge to achieve this goal, however. Dr. Katy Börner pointed out, a few years ago, similar discussion took place. While the participants all favored the solution, nothing happened after the meeting. To avoid falling into the same trap, the idea of a collaborative consortium was proposed. Four or five research groups represented by the participants have volunteered to participate in this collaborative effort to develop software specification and implement open source visualization tools that can be used by a variety of users and domains.
I truly hope this consortium will take off and the new implementation will reach the user by 2010. I also hope that members of consortium will pay attention to visualization applications in the real world, not just the visualization software.
Wondering around to a different group, the Information Aesthetics Group, I got a complete different vision of information visualization. This is a truly multidisciplinary group. The self-introduction around the table revealed that almost everyone has some art and design backgrounds – some are architects, some are visual art and media designers; some are interaction or user experience designers; some have industrial design background; some are, simply, artists. They are all researchers of information visualization. A main focus of their discussion is how to establish information aesthetics as an interdisciplinary research field. Their goal, “in 2010, we will have engaged in a broad discourse between art, design and information visualization scientists about the significance and purpose of aesthetics, including the aspects of usability, criticality, perception, experience, emotion, engagement, theory, criticism and the body.”
Aesthetics is a human need. From my own research on information visualization, I understand the importance and significance of aesthetics to information visualization. Clearly, if the goal of information visualization is to amplify user’s cognition, researchers much build aesthetics into their visualization. What catches user’s attention is aesthetics, not visualization algorithms. Aesthetics create impact. It helps user understand and develop insights from the results that algorithms generate. There are too many examples of visualization projects with excellent algorithms but poorly designed visual displays. A significant step to advance information visualization therefore involves promoting and marking aesthetics to visualization researchers and users. A list of questions discussed in the group is a good starting place:
Need a definition of “information aesthetics”? I really like their description: “Aesthetics is a fundamentally subjective concept that comprises the beauty in both nature and art, with the aim to enliven and invigorate both body and mind. It is capable of positively stimulating the human senses, including, but not necessarily limited to, the visual sensory system.”
An interesting group that I wish I could spend more time with is the group on Basic Narratives in Visualization. What a good idea it is “to develop a theoretical framework and best practice for identifying and applying basic narratives in visualization.”
Visualization is an art to tell a story from data using visual images and interaction functions. In our studies on author citation mapping (http://project.cis.drexel.edu/authormap/ ), we learned that every image our system created tells a long story to people who understand it. We have users who literally spent hours to tell us the story they read from the author map the system generated. The challenge for us is how to let every viewer understand the story. The data is fundamental and the source of the story. The visualization algorithms are the one that create the story schema. The aesthetical design can make the story “real” and finally, the narrative practice will help users develop a good understanding of the story.
It is interesting to observe that, initially, participants were all talking about visualization in different domain languages. As the discussion went on, participants within the groups shared more and more common languages. I wish there were more across-group dialogs to cultivate common languages and understanding among visualization researchers. The researchers with different backgrounds came to this Summit from different directions. The software infrastructure group, for example, takes the path from data to information visualization; the information Aesthetics group, on the other hand, brings the user’s perspective to information visualization. Do they meet successfully at the Summit? Time will tell.
See you in 2010!