Learning Disentangled Representations of Negation and Uncertainty

Autor: Vasilakes, Jake, Zerva, Chrysoula, Miwa, Makoto, Ananiadou, Sophia
Rok vydání: 2022
Předmět:
Druh dokumentu: Working Paper
Popis: Negation and uncertainty modeling are long-standing tasks in natural language processing. Linguistic theory postulates that expressions of negation and uncertainty are semantically independent from each other and the content they modify. However, previous works on representation learning do not explicitly model this independence. We therefore attempt to disentangle the representations of negation, uncertainty, and content using a Variational Autoencoder. We find that simply supervising the latent representations results in good disentanglement, but auxiliary objectives based on adversarial learning and mutual information minimization can provide additional disentanglement gains.
Comment: Accepted to ACL 2022. 18 pages, 7 figures. Code and data are available at https://github.com/jvasilakes/disentanglement-vae
Databáze: arXiv