Learning a Reversible Embedding Mapping using Bi-Directional Manifold Alignment

Thumbnail Image

Files

2021.findings-acl.276.pdf (325.32 KB)

Links to Files

https://aclanthology.org/2021.findings-acl.276/

Permanent Link

http://hdl.handle.net/11603/22331
 http://dx.doi.org/10.18653/v1/2021.findings-acl.276

Collections

UMBC Computer Science and Electrical Engineering Department
UMBC Faculty Collection

Author/Creator

Author/Creator ORCID

Date

2021-08-01

Type of Work

conference papers and proceedings
Text

Department

Program

Citation of Original Publication

Ganesan, Ashwinkumar; Ferraro, Francis; Oates, Tim; Learning a Reversible Embedding Mapping using Bi-Directional Manifold Alignment; Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021, pages 3132–3139 August 1–6, 2021; https://aclanthology.org/2021.findings-acl.276.pdf

Rights

This item is likely protected under Title 17 of the U.S. Copyright Law. Unless on a Creative Commons license, for uses protected by Copyright Law, contact the copyright holder or the author.

Subjects

UMBC Ebiquity Research Group
UMBC High Performance Computing Facility (HPCF)

Abstract

We propose a Bi-Directional Manifold Alignment (BDMA) that learns a non-linear mapping between two manifolds by explicitly training it to be bijective. We demonstrate BDMA by training a model for a pair of languages rather than individual, directed source and target combinations, reducing the number of models by 50%. We show that models trained with BDMA in the “forward” (source to target) direction can successfully map words in the “reverse” (target to source) direction, yielding equivalent (or better) performance to standard unidirectional translation models where the source and target language is flipped. We also show how BDMA reduces the overall size of the model.