# A Structured Variational Autoencoder for Contextual Morphological Inflection

arXiv: https://arxiv.org/abs/1806.03746 · HTML: https://mlsanity.com/paper/1806.03746/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 10 Jun 2018.

## Built on by (most cited first, 3 of 3)

- [Computational Morphology with Neural Network Approaches](https://mlsanity.com/paper/2105.09404/) (2021)
- [Recent advancements in computational morphology : A comprehensive survey](https://mlsanity.com/paper/2406.05424/) (2024)
- [Challenging the Semi-Supervised VAE Framework for Text Classification](https://mlsanity.com/paper/2109.12969/) (2021)

## Abstract (arXiv)

Statistical morphological inflectors are typically trained on fully supervised, type-level data. One remaining open research question is the following: How can we effectively exploit raw, token-level data to improve their performance? To this end, we introduce a novel generative latent-variable model for the semi-supervised learning of inflection generation. To enable posterior inference over the latent variables, we derive an efficient variational inference procedure based on the wake-sleep algorithm. We experiment on 23 languages, using the Universal Dependencies corpora in a simulated low-resource setting, and find improvements of over 10% absolute accuracy in some cases.