Demos: Concept Algebra Toolbox (CAT)

algebra1algebra2A set of Matab tools for complexity measurement and algebraic analysis of discrete-featured concepts

v1.0 by Jacob Feldman

Instructions for use:

1. Download the archive (zip file) and decompress it wherever you want it. It will create a directory called Algebra.
2. In Matlab, cd into the Algebra directory.
3. To use the graphical user interface (recommended): at the Matlab prompt, type ‘complexity_tool’. The GUI should appear.
4. For documentation and explanation of syntax, type ‘help algebra’ at the Matlab prompt. This gives a list of available command-line routines with brief explanations.

Click here to download the CAT (v1.0): zip

The concept algebra and its associated complexity measure (algebraic complexity) are described in Feldman (2006) "An algebra of human concept learning" (J. Mathematical Psychology, 50, 339-368). You can also find synopses and elaborations Feldman (2003) "How surprising is this simple pattern? Quantifying 'Eureka!'" (Cognition, 93, 199-224; see Appendix I) and Aitkin, C. D. and Feldman, J. (2006) "Subjective complexity of categories defined over three-valued features." (Proceedings of the Cognitive Science Society)

Please email the author (jacob@ruccs.rutgers.edu) with questions or comments.

Screenshot of complexity_tool in action, showing a 5-object concept defined over a 2x2x3-valued feature grid:

algebra3