Week 4: Cognitive Design Principles for Automated Generation of Visualizations- A summary

This paper defines the cognitive principles associated with creating effective visualizations. It further applies these principles to analyze two problem domains that we encounter in our daily lives, creating route maps to find a destination and crafting assembly instructions to construct an object of daily use. Finally, the paper introduces us to concrete Computer algorithms that were designed on the lines of the cognitive principles and that automates the task of laying out route maps and generating assembly instructions.

Visualization helps us in learning a concept and is sometime more powerful than written words in comprehending ideas. In order to create meaningful visualizations, one needs to be trained in the graphic design tools and also need to have a clear understanding of the domain for which the visualization is designed.  Principle of Congruence requires visualizations to follow the mental picture and the Principle of Apprehension states that it should be easily understood by the readers. In that regard, the author talks about route maps and discusses how they are an important example of visualization that is studied in detail. Well compiled maps conform to the cognitive principles by

-          Representing the routes in sequence of paths and the turning points

-          Simplifying the geographical aspects of the path

Based on the above principles, the authors developed a Computer algorithm called LineDrive that automates the task of creating route maps and are used by popular internet map services like www.mapblast.com. The algorithm is designed based on the design ideas that were refined over a period of time and it applies them to perform the labor intensive task to draw the route map based on destination information.

An instruction to assemble an object is another set of visualization that we encounter in our daily life. We have to understand these instructions and apply them to meet our objectives. It is easy to see the differences between a well written and a poorly compiled instruction. Well designed instructions have the following features in common:

-          Sketches to illustrate the structures and spatial relations between assembled components

-          Multiple sketches breaking up the tasks into a sequence of steps

-          Use of 3D constructs to give a better perspective

-          Additional guidelines and illustrations commonly known as Morphograms

-          Accompanying text notes explaining the diagrams in detail

The authors cite studies which show that good instructions consistently improve productivity for people with average craftsmen skills. The authors then describe an algorithm that can automatically generate assembly instruction when it is provided with the geometric model of the completely assembled object and with the information of its individual components. The algorithm derives on the principles of Robotics to automate a complex and repetitive task.

On a final note, the authors discuss the strength of the algorithms by referring to studies which show that the algorithms produce visualizations that betters hand crafted versions. They also reflect the fact that other types of visualization problems can also be automated with cognitive design principles to create powerful solutions.