Generating Logical Forms from Graph Representations of Text and Entities
This work addresses semantic parsing tasks by improving logical form generation, though it is incremental as it builds on existing GNN and pre-training techniques.
The paper tackles the problem of generating logical forms for semantic parsing by incorporating entity and relation information using a Graph Neural Network (GNN) with a decoder copy mechanism, achieving competitive state-of-the-art results without pre-training and outperforming existing methods when combined with BERT pre-training.
Structured information about entities is critical for many semantic parsing tasks. We present an approach that uses a Graph Neural Network (GNN) architecture to incorporate information about relevant entities and their relations during parsing. Combined with a decoder copy mechanism, this approach provides a conceptually simple mechanism to generate logical forms with entities. We demonstrate that this approach is competitive with the state-of-the-art across several tasks without pre-training, and outperforms existing approaches when combined with BERT pre-training.