In this project we are extending the work of Mitchell and Hofstadter on modeling analogy-making, and integrating this work into image understanding and retrieval systems.