Tuesday, August 30, 2011

Reaction: Value of Info Viz and Automating Design of Graphical Presentations

The Value of Information Visualization
There is no doubt that anyone has failed to recognize the value information visualization. However, as the authors mention, proving this by quantifiable metrics is also equally important. I found this very interesting and different as personally I always felt benefits of InfoViz over plain data were always obvious. As discussed in class, and also mentioned in the paper, "amplifying cognition" is the main objective of visualization. The paper also mentions some classic examples of which most were discussed in class, like Minard's map of the march of Napolean's Army into Russia and John Snow's map of London which led to the conclusion that cholera is a water-bourne disease.
So, to quantifiably measure how effective or the value of the viz, demands a great understanding of human cognition and perception. I also liked how the difference between confirmatory, exploratory and automated methods was brought about in the paper. Gestalt theory I could easily identify with, being familiar with it from before. I found it very interesting to note how quantifying the value of information visualization takes into consideration so many factors like cognition, perception, principles of psychology.
Automating the Design of Graphical Presentations of Relational Information
User interface should make use of human visual abilities to present info to viewers in an effective manner. Presentation tool is a great idea. But what I feel is a very big challenge is that handling so many types of input data and be able to classify them into graphic design criteria. The approach that the authors have taken has its roots in the analogy that each presentation is a sentence, which implies that every presentation of data will have semantic definitions like the formal language. There are parts of this paper which I found hard to understand. Say, it was easy to understand that there are conventions which tell us that a particular arrangement of graphical objects represent some information. But how to formalize these conventions as semantic definitions was a bit difficult to comprehend. Having seen so many examples of visualization and tools presented in class, this paper made me realize the kind of thinking and theory that goes behind building it.