Visual Analogy Research

We know that visual representations are important for problem solving, under what conditions this is true is poorly understood.

In my theory visual representations are useful when the non-visual representations contain mismatches between symbols that are similar at a visual level. For example, a wall and a pole are not the same thing, but your visualization of them, perhaps as lines, can show their similarity. For papers on the theory, see:

• My Dissertation is a lengthy description of my theory. Good for comprehension and detail.
  • My dissertation proposal, [ ps | HTML (latex2html didn't do a very good job with the images, I'm afraid.) ]
• For a description of the theory as it relates to creativity, see:

  Davies, J., Goel, A. K. & Nersessian, N. J. (2003). Visual Re-Representation in Creative Analogies. The Third Workshop on Creative Systems. International Joint Conference on Artificial Intelligence.

• For the theory applied to design, see

  Davies, J., Goel, A. K., & Nersessian, N. J. (2009). A Computational Model of Visual Analogies in Design. Cognitive Systems Research: Special Issue on Analogies, 10, 204--215.

To solve this problem I am creating a language of visual primitives called Covlan. In my theory, translating a memory into this visual language will facilitate analogical problem solving in some cases.

My computer program Galatea implements core parts of the theory.

Papers on Covlan and Galatea

• Cite this as the original paper on Galatea and Covlan (which was called "Privlan" at that time)

  Davies, J., & Goel, A. K. (2001). Visual analogy in problem solving. Proceedings of the International Joint Conference on Artificial Intelligence 2001. pp 377-382. Morgan Kaufmann publishers.

• This paper explores the representation issues for a cognitive science audience:

  Davies, J., & Goel, A. K. (2003). Representation issues in visual analogy. Proceedings of the 25th Annual Conference of the Cognitive Science Society. 300--305.

• This is a journal article about Covlan, targeted to specialists in Visual Languages

  Davies, J. & Goel, A. K. (2007a). Transfer of problem-solving strategy using the cognitive visual language. The Journal of Visual Languages and Computing. 18:2 149--164.

• Interesting future work, related to modeling mental imagery, is described in this presentation given to the Georgia Tech Cognitive Science Brownbag series on October 27, 2000.

• I have applied my theory to scientific discovery, specifically to model the historical case of James Clerk Maxwell.

  The work was presented at this conference:

  Model-Based Reasoning: Scientific Discovery, Technological Innovation, Values 2001 conference.
  Submitted Abstract: [ Microsoft Word ]
  Talk: [ Microsoft PowerPoint ]

  and at my presentation given to the 2001 Cognitive Studies of Science Workshop in Charlottesville. [ html | power point ]

• These papers apply these findings to visual case-based reasoning, the first more geared toward AI and the second and third geared toward cognitive modelling:
  This one is geared toward case-based reasoning and AI.

  Davies, J. & Goel, A. K. (2003). Visual case-based reasoning: Transfer and Adaptation. The 1st Indian International Conference on Artificial Intelligence. Springer.

  These are geared toward Case-based reasoning as a cognitive model.

  Davies, J., Goel, A. K. & Nersessian, N. J. (2005). Transfer in visual case-based problem-solving. In H. Munoz-Avila & F. Ricci (Eds.) Proceedings of the 6th International Conference on Case-Based Reasoning. Springer-Verlag. Berlin Heidelberg. 163--176.

  Davies, J., Goel, A. K., & Nersessian, N. J. (2005). A cognitive model of visual analogical problem-solving transfer. Poster paper in L. P. Kaelbling & A. Saffioti (Eds.) the Proceedings of the Nineteenth Annual International Joint Conference on Artificial Intelligence. Professional Book Center, Denver, CO. 1556-1557.