Graph-Based Decoding for Task Oriented Semantic Parsing
This work addresses semantic parsing for task-oriented dialogue systems, offering a more data-efficient alternative to existing methods.
The authors tackled semantic parsing by reformulating it as a dependency parsing task using graph-based decoding instead of the standard sequence-to-sequence approach, finding that their method is competitive in standard settings and offers significant improvements in data efficiency and with partially-annotated data.
The dominant paradigm for semantic parsing in recent years is to formulate parsing as a sequence-to-sequence task, generating predictions with auto-regressive sequence decoders. In this work, we explore an alternative paradigm. We formulate semantic parsing as a dependency parsing task, applying graph-based decoding techniques developed for syntactic parsing. We compare various decoding techniques given the same pre-trained Transformer encoder on the TOP dataset, including settings where training data is limited or contains only partially-annotated examples. We find that our graph-based approach is competitive with sequence decoders on the standard setting, and offers significant improvements in data efficiency and settings where partially-annotated data is available.