Friday, November 14, 2014

Visualization and Decision-Making Using Structural Information

Author: Boris Kovalerchuk


Kovalerchuck's research aimed to highlight how and when certain types of visualization techniques should be used.  From an intelligence perspective, visualizations should not often not used to their full effectiveness as they are often presented to present a great deal of information to the audiance.  Take for example many of the info-graphics seen in newspapers.  These info-graphics often are quite colorful, look professional, and give the reader a great deal of information.  Kovalerchuk states that these should not be the types of visuals that analysts use with their decision makers.  Kovalerchuck identifies two main purposes of data visualization techniques for intelligence professionals; discovered relations/pattern (DRP) visuals and decision making model (DMM) visuals.

1) DRP Visuals
DRP visuals help the analysts in his or her analysis of the situation.  These are often referred to as exploratory visuals.

2) DMM Visuals
DMM visuals assist decision makers in making decisions.  These visuals are often more simplified than DRP visuals and should create a clear image of what the issue is and lead to ideas on how to address the issue in question.

A DRP visual will guide the analysts to create the DMM model.  The key finding of Kovalerchuk's research into data visualization techniques is that decision makers are comprehend and make better decisions from visuals they are most familiar with.  Examine the following image,

This simple image shows the relation of human deaths (black squares) next to water pumps (black circles) in relation to city blocks.  The deaths all occurred in close proximity to a water pump.  This is an overly simplified geospatial analysis that most analysts are familiar with.  Decision makers may not be familiar with this technique and have to think more about what the visual means.

This is a bar chart image showing the death toll within 250 yards of certain water pumps.  A geospatial graphic could have been designed to show the exact same information.  Decision makers are used to seeing this type of chart.  Kovalerchuk found no disadvantages to using this type of chart over more complicated geospatial charts to present findings.

I agree with many of the findings of this research paper.  I agree that, we as analysts should seek to offer visuals that help decision makers best make decisions.  It also makes a great deal of sense to me that showing visuals that decision makers are familiar with, such as bar and pie charts, would be as effective as showing them more complicated charts that analysts are used to.

My main issue with this research is that the author does not give any information into how he came to these conclusions.  It appears that he mostly analyzed literature on the topic.  Some of his statements though make it appear as if he did conduct human research on how visuals effect decision making.  I would be very interested to see a study on how decision makers comprehend advanced visuals usually slated for analysts (geospatial analysis) vs traditional data visuals (turning the content of a geospatial analysis into a bar or line chart).


Kovalerchuk, B. (2001).  Visualization and Decision-Making Using Structural Information.  Proceedings of International Conference of Imaging Science, Systems, and Technologies.


  1. This comment has been removed by the author.

  2. I agree that the purpose of visualizing information in an intelligence context is to present the information in a way that facilitates greater understanding and faster comparisons to disparate data sets without increasing the complexity of what the product presents.

    However, I disagree with Kovalerchuk's finding indicating that info-graphics should not be in an analyst's toolbox. In my mind, I draw many parallels between exceptionally crafted info-graphics and exceptionally crafted executive dashboards, which often consist of more than aggregations of convergent information through structured schematics like bar charts and histograms.

    Instead, I think we should push the envelope and call for greater visual literacy among the general populace and within professions.

    Using visualization to tell a story about data, like with info-graphics, is a technique used over 10,000 years ago in the Paleolithic era through cave drawings. Furthermore, I contend that while graph charts excel at reducing complexity and conveying overall patterns, they are not as conducive to communicating process or characteristics of individual data points in every situation.

  3. Was there any statistical analysis to backup Kovalerchuk's claims? Or is this a paper merely identifying his preference?

    1. John,

      No there was not. This research came from an analysis of visualization research and his own experiences.

  4. Ricardo,

    I believe that Kovalerchuk was trying to establish that info-graphics are most often used display a lot of information in colorful and visually appealing way. The industry that uses info-graphics the most is news and journalism. They are often very vibrant and colorful. With most info-graphics I have seen, they really only make is so I can learn a lot of information instead of having to read it in text form. Kovalerchuk believes intelligence visualizations should be created to provide DM's with the ability to make decisions based on the graphics, not just to present a lot of information.

  5. Harrison,
    Do you think infographics are a fad, or do you think they are here to stay? In addition, do you think traditional data visuals such as pie charts, bar charts, etc. are better in the IC?