Compositional generalization with a broad-coverage semantic parser
Pia Weißenhorn, Lucia Donatelli, Alexander Koller
Abstract
We show how the AM parser, a compositional semantic parser It is the first semantic parser that achieves high accuracy on both naturally occurring language and the synthetic COGS dataset. We discuss implications for corpus and model design for learning human-like generalization. Our results suggest that compositional generalization can be best achieved by building compositionality into semantic parsers.
Topics & Concepts
Principle of compositionalityComputer scienceParsingGeneralizationNatural language processingArtificial intelligenceProgramming languageMathematical analysisMathematicsNatural Language Processing TechniquesTopic ModelingMultimodal Machine Learning Applications