A unified theory explains how people learn concepts and chunks
A Unified Account of Concepts and Chunks
Computation and LanguageArtificial Intelligence
Summary
People learn by recognizing categories (concepts) and familiar patterns (chunks), but these ideas have been studied separately. The authors review a system called Cobweb that models how concepts form, and they extend it to include chunks too. They created a tool named Trellis to test this idea using grammar rules, showing it can learn language-like patterns from examples. This work helps understand how we combine different kinds of knowledge in our minds.
What this means in practice
- •For natural language engineers: Use Trellis to build systems that learn and represent complex language patterns from example sentences.
- •For software developers in ai: Incorporate the unified theory of concepts and chunks to enhance machine learning algorithms for pattern recognition across data types.
Authors
Karthik Singaravadivelan, Pat Langley
Abstract
Cognitive psychology has studied how people encode, use, and learn concepts that describe categories, and how they represent, recognize, and acquire chunks for familiar patterns of elements. The literatures on these two topics are nearly disjoint, which poses a challenge for unified theories of cognition. In this paper, we review Cobweb, a computational account of categorization and concept formation, and propose an extended theory that incorporates chunks and their acquisition. The theory makes no commitments about modality, applying to any experience that decomposes into elements and relations among them. We also present \trellis/, an implementation of this theory, and illustrate its application to learning context-free grammars, which we adopt as a testbed because they involve both concept-like and chunk-like elements. In addition, we report experimental results on three synthetic grammars that demonstrate the system's ability to represent syntactic knowledge, use it to parse and generate sentences, and learn compositional structures from sample parses. We conclude by discussing related work on concepts and chunks, along with directions for future research in the area.