Simpler but More Accurate Semantic Dependency Parsing
This work improves parsing accuracy for natural language processing tasks, but it is incremental as it builds on existing methods.
The paper tackled semantic dependency parsing by extending an LSTM-based syntactic parser to generate graph structures, achieving state-of-the-art performance with a 0.6% labeled F1 improvement over the previous complex system and up to 1.9% with richer inputs.
While syntactic dependency annotations concentrate on the surface or functional structure of a sentence, semantic dependency annotations aim to capture between-word relationships that are more closely related to the meaning of a sentence, using graph-structured representations. We extend the LSTM-based syntactic parser of Dozat and Manning (2017) to train on and generate these graph structures. The resulting system on its own achieves state-of-the-art performance, beating the previous, substantially more complex state-of-the-art system by 0.6% labeled F1. Adding linguistically richer input representations pushes the margin even higher, allowing us to beat it by 1.9% labeled F1.