CPSC 533C: Assignment 1

by Siamak Tazari

The images below are taken from the paper "A Contextual-Logic Extension of TOSCANA"1 by Peter W. Eklund, B. Groh, Gerd Stumme and Rudolf Wille. Contextual Logic - developed at Darmstadt University of Technology, Germany - is an attempt to formalize classical logic (as opposed to mathematical logic), which is based on concept, judgement and conclusion. Since 1980, much work has been done in this area and the developed methods were successfully used in numerous practical applications (for more information, see http://www.upriss.org.uk/fca/fca.html ). One of its main strengths is that it transforms data into concept lattices and concept graphs, which can be very accessible to the human mind - if visualized properly. The images below represent the same information about the domestic flight schedule in Austria, based on a user query. The first one is a "raw" concept graph, whereas the second one is a map-based visualization of the data in the graph. Whereas the concept graph can be constructed automatically, coming up with an alternative visualization as the second one here, can be a challenge and non-trivial. But as we see in this example, it is a very important step and drastically changes the usefulness of the data.

It is clear that the diagram above is too complicated to use. If you want to find out how to get from one place to another at a certain time, you have to follow a whole lot of arrows - sometimes in the opposite direction - and the probability of getting lost in the diagram is very high. Furthermore, for people who are not much familiar with such graphs, this representation is not usable at all.

This picture very clearly and beautifully illustrates the very same data in a way which is easily understood by anyone without the need of much (or any) explanation. If one wants to plan a trip, she can do so by just looking at this diagram with the plain eye and following a few lines - no computers (who sometimes never understand what you are really looking for) or other help is needed, neither is there the difficulty of looking up entries in tables and matching them to find connections. I think this is a good visualization because it is simple, compact, easily understood and non-confusing.

1 Peter W. Eklund, B. Groh, Gerd Stumme, Rudolf Wille , "A Contextual-Logic Extension of TOSCANA", In B. Ganter, G. W. Mineau, Conceptual Structures: Logical, Linguistic, and Computational Issues. Proc. ICCS '00, pp. 453-467. Springer, Heidelberg, 2000.
http://www.aifb.uni-karlsruhe.de/Publikationen/showPublikation_english?publ_id=529