Simpler Context-Dependent Logical Forms via Model Projections
This work addresses the problem of semantic parsing with context-dependent utterances for researchers in natural language processing, though it appears incremental as it builds on existing methods with new datasets and optimizations.
The paper tackles the challenge of learning context-dependent mappings from utterances to denotations by using simpler model projections to reduce the search space, resulting in faster performance and effectiveness, including bootstrapping the full model and introducing three new datasets and a left-to-right parser.
We consider the task of learning a context-dependent mapping from utterances to denotations. With only denotations at training time, we must search over a combinatorially large space of logical forms, which is even larger with context-dependent utterances. To cope with this challenge, we perform successive projections of the full model onto simpler models that operate over equivalence classes of logical forms. Though less expressive, we find that these simpler models are much faster and can be surprisingly effective. Moreover, they can be used to bootstrap the full model. Finally, we collected three new context-dependent semantic parsing datasets, and develop a new left-to-right parser.