While neural, encoder-decoder models have had significant empirical success in text generation, there remain several unaddressed problems with this style of generation. Encoder-decoder models are largely (a) uninterpretable, and (b) difficult to control in terms of their phrasing or content. This work proposes a neural generation system using a hidden semi-markov model (HSMM) decoder, which learns latent, discrete templates jointly with learning to generate. We show that this model learns useful templates, and that these templates make generation both more interpretable and controllable. Furthermore, we show that this approach scales to real data sets and achieves strong performance nearing that of encoder-decoder text generation models.
@article{arxiv.1808.10122,
title = {Learning Neural Templates for Text Generation},
author = {Sam Wiseman and Stuart M. Shieber and Alexander M. Rush},
journal= {arXiv preprint arXiv:1808.10122},
year = {2019}
}