An interview with…Chaomei Chen

Today I’m really glad to post the interview with Dr. Chen ( from the College of Information Science and Technology, Drexel University, Philadelphia ) that wrote some interesting books about infovis.

1 - When, how and why have you decided to research on infovis?

My earlier research on graphical overview maps in hypertext systems took me to infovis. Much of the study of such maps was motivated by two common usability problems with hypertext, namely the cognitive overload problem and the lost in hyperspace problem. In a 1996 meta-analysis of hypertext systems, we statistically identified the universal significant effect of graphical maps. This finding along with my interests in associative networks led to the development of a generic conceptual framework - Generalized Similarity Analysis (GSA). I presented GSA to the 1997 ACM Hypertext conference. GSA is simple, intuitive, and generic for modeling and visualizing a wide variety of information structures, ranging from the structure of multiple interrelated websites, implicit structures in digital libraries and scientific literature, to users’ click-through patterns or their footprints in a virtual world. Some of the work we do today is still influenced by GSA; for example, the increasingly popular analysis of co-authorship networks and co-citation networks is indeed a special case in the unifying framework of GSA. GSA provides a good starting point for modeling and visualizing associative networks. I was attracted to infovis by pioneering works such as SemNet, graph drawing algorithms, and ConeTrees. My first book on information visualization and virtual environments was completed in the winter of 1998 and published in 1999. So as it appeared, it took two or three years to reach the field of infovis from the field of hypertext.

2 - Which is, to you, the most interesting project you have worked on and why?

To me, the attraction of information visualization boils down to its potential to bring you some deeply intriguing and thought-provoking images. Yes, the Aha! moments. It may not happen all the time, but once it comes, it makes all your efforts worthwhile. Since it is hard to put such revelations on a scalar scale, instead of naming a single project, let me give you a couple of examples. I can still remember vividly how I was wondering what I was about to see while my first animated 3-dimensional model of a growing citation landscape was being computed. I can also remember the time when the bridges connecting three thematic continents of terrorism research were chosen algorithmically. I am sure we haven’t seen the most exciting ones yet.

3 - On what themes are you working now?

I continue to work on visualization tools that can help scientists and the general public to keep abreast of the advance of a scientific knowledge domain. In particular, CiteSpace, a freely available Java application, continues to evolve. A new challenging theme we are working on is to help people better understand conflicting opinions. Another relevant theme we are working on is to track the diffusion of information over geographical maps. The theme on conflicting opinions is an extension of some of the topics I have worked in the areas of visualizing scientific paradigms and conceptual revolutions. A common task is to understand the basis of an argument, the evidence for a counter-argument, or a cultural, social, and historical context that can illumine the nature of a controversy. So the new theme goes beyond the scope of scholarly communication and scientific inquiries. We are working on visual analysis of book reviews in order to answer questions such as what differentiating features tell apart positive reviews and negative reviews.

4 - If you should suggest an education programme to a student interested in InfoVis, which University Course do you suggest him/her?

InfoVis is interdisciplinary in nature. Two types of knowledge are essential: principles and technologies that drive the crafts of visual thinking and communication, and innovative problem-solving beyond a single field. I would not attempt to point to one specific program and omit many other good ones; instead, it would a good idea to look for a program that is an integral part of a multidisciplinary research environment.

5 - How do you think infovis researches will evolve in your country for next years?

There are several areas where a tremendous growth can be expected: visual analytics, text mining, data mining and knowledge discovery, and social sciences.