Visual-Imagery-Based Analogical Construction in Geometric Matrix Reasoning Task
This work addresses figural analogical reasoning for AI systems, but it is incremental as it builds on existing human strategies without introducing a fundamentally new approach.
The authors tackled solving Raven's Progressive Matrices using computational models based on analogies and image transformations, achieving a result of solving 57 out of 60 problems on the standard version.
Raven's Progressive Matrices is a family of classical intelligence tests that have been widely used in both research and clinical settings. There have been many exciting efforts in AI communities to computationally model various aspects of problem solving such figural analogical reasoning problems. In this paper, we present a series of computational models for solving Raven's Progressive Matrices using analogies and image transformations. We run our models following three different strategies usually adopted by human testees. These models are tested on the standard version of Raven's Progressive Matrices, in which we can solve 57 out 60 problems in it. Therefore, analogy and image transformation are proved to be effective in solving RPM problems.