This paper focuses on latent representations that could effectively decompose different aspects of textual information. Using a framework of style transfer for texts, we propose several empirical methods to assess information decomposition quality. We validate these methods with several state-of-the-art textual style transfer methods. Higher quality of information decomposition corresponds to higher performance in terms of bilingual evaluation understudy (BLEU) between output and human-written reformulations.
@article{arxiv.1909.12928,
title = {Decomposing Textual Information For Style Transfer},
author = {Ivan P. Yamshchikov and Viacheslav Shibaev and Aleksander Nagaev and Jürgen Jost and Alexey Tikhonov},
journal= {arXiv preprint arXiv:1909.12928},
year = {2022}
}
Comments
arXiv admin note: substantial text overlap with arXiv:1908.06809