# The Meaning Factory at SemEval-2017 Task 9: Producing AMRs with Neural Semantic Parsing

## TL;DR

This paper presents a neural sequence-to-sequence semantic parser for AMRs, achieving notable improvements over baseline models through data augmentation and POS-tagging, but still lagging behind state-of-the-art parsers, with ensemble methods providing marginal gains.

## Contribution

It introduces a neural character-based semantic parser with data augmentation and POS-tagging, advancing neural approaches for AMR parsing.

## Key findings

- Data augmentation improves performance
- Ensemble with traditional parser yields small gains
- Neural parser outperforms previous character-based models

## Abstract

We evaluate a semantic parser based on a character-based sequence-to-sequence model in the context of the SemEval-2017 shared task on semantic parsing for AMRs. With data augmentation, super characters, and POS-tagging we gain major improvements in performance compared to a baseline character-level model. Although we improve on previous character-based neural semantic parsing models, the overall accuracy is still lower than a state-of-the-art AMR parser. An ensemble combining our neural semantic parser with an existing, traditional parser, yields a small gain in performance.