Wednesday, August 31, 2011

Reaction: The Value of Information Visualization

The paper discusses the issues regarding the value of Information Visualization in an elegant manner. The discussion starts with exactly the actual defining the "visualization" and why identifying the value of InfoVis is so difficult.

The definition for "visualization" from Card, Mackinlay, and Shneiderman - "the use of computer-supported, interactive visual representations of data to amplify cognition", is probably the neatest of many definitions, and gives the exact holistic meaning of the term.

I agree with the authors that, probably the reason for why the value of InfoVis is not given much recognition, as InfoVis is always one of the reasons for the discovery of the pattern apart from identifying and collecting the data etc., but not the only reason for the discovery.

It is quite interesting to see that "Data Mining" can potentially be a competing domain for InfoVis, as this domain aims to finding patterns using automated algorithms and machine learning/artificial intelligence. I think authors have done a good job in convincing their audience that by just making use of Data Mining, it is really not possible to understand and get valuable insights of the data, as the data mining models might not be suitable for every kind of data. And moreover, inorder to make a decision about which data mining model to employ, the research scientist should first able to visualize the information.

Overall, it was quite enlightening to read this paper. The authors have come quite strong in putting forward the importance of Information Visualization.