Exploring Neural Methods for Parsing Discourse Representation Structures
This work addresses the problem of semantic parsing for formal semantics, which is incremental as it applies neural methods to an existing domain.
The paper tackles the challenge of using neural methods to parse Discourse Representation Structures (DRSs) for English sentences, achieving high accuracy and outperforming traditional parsers, with performance boosted by techniques like De Bruijn-indices and silver training data.
Neural methods have had several recent successes in semantic parsing, though they have yet to face the challenge of producing meaning representations based on formal semantics. We present a sequence-to-sequence neural semantic parser that is able to produce Discourse Representation Structures (DRSs) for English sentences with high accuracy, outperforming traditional DRS parsers. To facilitate the learning of the output, we represent DRSs as a sequence of flat clauses and introduce a method to verify that produced DRSs are well-formed and interpretable. We compare models using characters and words as input and see (somewhat surprisingly) that the former performs better than the latter. We show that eliminating variable names from the output using De Bruijn-indices increases parser performance. Adding silver training data boosts performance even further.