Research

How mental representations are constructed and how they evolve are central problems for cognitive science. Representation decisions help determine what computations are hard or easy. Structured, relational representations are a hallmark of human cognition. Developmental studies show that children do not perform as well as adults in tasks that require noticing relational similarity. But improve over time with experience (and age). What drives this development? This question is the primary motivation for my research.

Gentner and her colleagues argue that comparison and language are the two primary forces that drive representational change. In my doctoral thesis, I explore these ideas further by presenting a computational model to illuminate the roles of comparison and language in driving representational change.

Comparison Driven Representational Change

Click on the image above to see a flow chart of the model. The slideshow on the left highlights some interesting images from my dissertation. You can read the full version here.

Publications

Conference Proceedings

Kandaswamy, S, Forbus, K. and Gentner, D. (2014). Modeling Learning via Progressive Alignment using Interim Generalizations. Proceedings of the 36th Annual Conference of the Cognitive Science Society (CogSci). Quebec city, Canada.

Kandaswamy, S. and Forbus, K. (2012). Modeling Learning of Relational Abstractions via Structural Alignment. Proceedings of the 34th Annual Conference of the Cognitive Science Society (CogSci). Sapporo, Japan.

Doctoral Dissertation

Kandaswamy, S. (2016). Comparison Driven Representational Change Doctoral dissertation, Northwestern University, Department of Electrical Engineering and Computer Science, Evanston, Illinois. [Dissertation].

Workshops

McLure, M., Kandaswamy, S., and Forbus, K. (2015). Finding Textures in Sketches using Planar Ising Models . Proceedings of the 28th International Workshop on Qualitative Reasoning. Minneapolis, MN.

Lovett, A., Kandaswamy, S., McLure, M., and Forbus, K. (2012). Evaluating qualitative models of shape representation. Proceedings of the 26th International Workshop on Qualitative Reasoning. Los Angeles, CA.

Research Software

CogSketch : Open domain sketch-understanding software. Used as a cognitive science research instrument and as a sketch-based educational software. (My Contribution: Design and implementation of the ink processing modules: corner detection, edge segmentation, shape detection, etc.)

Case Mapper : Tool designed for cognitive scientists to experiment with computational models of structure-mapping (SME, MAC/FAC) (My Contribution: Enhanced with more features (V6 to V7) and Maintained the software)

Other

A NetLogo model for comparing the effects of NeoClassical vs Keynesian economic theories on immigration. (for course ECCS 372)