# SU-RUG at the CoNLL-SIGMORPHON 2017 shared task: Morphological Inflection with Attentional Sequence-to-Sequence Models

## TL;DR

This paper presents an attentional sequence-to-sequence neural network system for morphological inflection, leveraging joint training of inflection and analysis, achieving high performance in the SIGMORPHON 2017 shared task.

## Contribution

The authors introduce a neural network model with joint training for inflection and analysis, significantly improving over baseline results in morphological tasks.

## Key findings

- Outperforms baseline models by a large margin
- Ranks 4th in the high-resource track of the shared task
- Demonstrates effectiveness of joint training in morphological modeling

## Abstract

This paper describes the Stockholm University/University of Groningen (SU-RUG) system for the SIGMORPHON 2017 shared task on morphological inflection. Our system is based on an attentional sequence-to-sequence neural network model using Long Short-Term Memory (LSTM) cells, with joint training of morphological inflection and the inverse transformation, i.e. lemmatization and morphological analysis. Our system outperforms the baseline with a large margin, and our submission ranks as the 4th best team for the track we participate in (task 1, high-resource).