Cross-lingual Semantic Parsing
This addresses the problem of language-agnostic semantic understanding for NLP applications, though it appears incremental with specific technical improvements.
The paper tackles cross-lingual semantic parsing by mapping source language content to target language meaning representations, introducing a model with a novel copy mechanism that outperforms baselines by at least 1.18 F1 score.
We introduce the task of cross-lingual semantic parsing: mapping content provided in a source language into a meaning representation based on a target language. We present: (1) a meaning representation designed to allow systems to target varying levels of structural complexity (shallow to deep analysis), (2) an evaluation metric to measure the similarity between system output and reference meaning representations, (3) an end-to-end model with a novel copy mechanism that supports intrasentential coreference, and (4) an evaluation dataset where experiments show our model outperforms strong baselines by at least 1.18 F1 score.