LAGr: Label Aligned Graphs for Better Systematic Generalization in Semantic Parsing

TL;DR

LAGr introduces a graph-based semantic parsing framework that enhances systematic generalization by directly producing meaning representations as aligned graphs, outperforming traditional seq2seq models.

Contribution

The paper presents LAGr, a novel graph-based approach for semantic parsing that improves systematic generalization through label-aligned graph representations.

Findings

LAGr significantly outperforms seq2seq parsers in systematic generalization tasks.

Both strongly- and weakly-supervised LAGr achieve notable improvements.

LAGr effectively infers alignments in unaligned graphs using approximate inference.

Abstract

Semantic parsing is the task of producing structured meaning representations for natural language sentences. Recent research has pointed out that the commonly-used sequence-to-sequence (seq2seq) semantic parsers struggle to generalize systematically, i.e. to handle examples that require recombining known knowledge in novel settings. In this work, we show that better systematic generalization can be achieved by producing the meaning representation directly as a graph and not as a sequence. To this end we propose LAGr (Label Aligned Graphs), a general framework to produce semantic parses by independently predicting node and edge labels for a complete multi-layer input-aligned graph. The strongly-supervised LAGr algorithm requires aligned graphs as inputs, whereas weakly-supervised LAGr infers alignments for originally unaligned target graphs using approximate maximum-a-posteriori inference. Experiments demonstrate that LAGr achieves significant improvements in systematic generalization upon the baseline seq2seq parsers in both strongly- and weakly-supervised settings.