Teresa Escrig

News and oppinion about Cognitive AI & Robotics

It is nice to receive awards for scientific excellence

leave a comment

This is an article written in collaboration with one of my postdoctoral students at the U. Jaume I.

One of my research lines a few years ago, was to create a ‘General Model’ to describe the behavior of all qualitative models based on intervals, also named – “naming” qualitative models. “Naming” qualitative models are the ones that help us to express spatial-temporal concepts such as “That horse is really fast” (velocity); “That boy is really tall” (size); “My favorite restaurant is close” (distance).  The other way we reason with qualities is by comparing:  “The black horse is faster than the brown one”; “My son is taller than yours”; “The pizzeria is closer to the Thai restaurant”.

The basis of the qualitative reasoning process is defined very easily (in fact, it’s easier than it seems when you try to explain it with words, as I am doing).  If you have the relationship (any of the spatial-temporal concepts that you are interested in applying) between the object b and the reference system RS1, and we have the relationship between the object c and another reference system RS2, and object b is included into the RS2, the basic reasoning model will obtain the relationship between object c and RS1.  As an example: given two distances between three spatial objects, a , b and c, that is Dab and Dbc, we want to find the distance which is not initially given, i.e. Dac.

I can hear the voices of those asking… “and why is this important”?  “What is this useful for”?  The general model will allow us to create a unique algorithm to be able to represent and reason with all different spatial-temporal qualitative naming concepts.

We corrected this article so many times, that we finally obtained the Practical Applications of Agent and Multi-Agent Systems’ 2012 AWARD OF SCIENTIFIC EXCELLENCE.  Congratulations Esther!

If you want to read the whole article, paperDCAI-2012_red

Leave a Reply