Effective estimation of deep generative language models
T Pelsmaeker, W Aziz - arXiv preprint arXiv:1904.08194, 2019 - arxiv.org
Advances in variational inference enable parameterisation of probabilistic models by deep
neural networks. This combines the statistical transparency of the probabilistic modelling …
neural networks. This combines the statistical transparency of the probabilistic modelling …
A neural generative model for joint learning topics and topic-specific word embeddings
We propose a novel generative model to explore both local and global context for joint
learning topics and topic-specific word embeddings. In particular, we assume that global …
learning topics and topic-specific word embeddings. In particular, we assume that global …
Leveraging natural supervision for language representation learning and generation
M Chen - arXiv preprint arXiv:2207.10617, 2022 - arxiv.org
Recent breakthroughs in Natural Language Processing (NLP) have been driven by
language models trained on a massive amount of plain text. While powerful, deriving …
language models trained on a massive amount of plain text. While powerful, deriving …
Neural baselines for word alignment
Word alignments identify translational correspondences between words in a parallel
sentence pair and is used, for instance, to learn bilingual dictionaries, to train statistical …
sentence pair and is used, for instance, to learn bilingual dictionaries, to train statistical …
Generative latent neural models for automatic word alignment
Word alignments identify translational correspondences between words in a parallel
sentence pair and are used, for instance, to learn bilingual dictionaries, to train statistical …
sentence pair and are used, for instance, to learn bilingual dictionaries, to train statistical …
[PDF][PDF] Topic representation learning on sequential data for text understanding
L Zhu - 2023 - wrap.warwick.ac.uk
Throughout history, texts have been essential in communication between humans and
machines, as well as among humans themselves. Texts are used for multiple purposes …
machines, as well as among humans themselves. Texts are used for multiple purposes …
[PDF][PDF] Alignment-based neural networks for machine translation
TAN Alkhouli - 2020 - publications.rwth-aachen.de
After more than a decade of phrase-based systems dominating the scene of machine
translation, neural machine translation has emerged as the new machine translation …
translation, neural machine translation has emerged as the new machine translation …
Generative Probabilistic Alignment Models for Words and Subwords: a Systematic Exploration of the Limits and Potentials of Neural Parametrizations
AKN Ho - 2021 - theses.hal.science
Alignment consists of establishing a mapping between units in a bitext, combining a text in a
source language and its translation in a target language. Alignments can be computed at …
source language and its translation in a target language. Alignments can be computed at …
[PDF][PDF] Global Under-Resourced MEedia Translation (GoURMET)
R Dobreva, RA Birch - gourmet-project.eu
This deliverable reports the work conducted within WP3 on structure induction at sentence
level for low-resource neural machine translation (NMT). It focuses on three main tasks …
level for low-resource neural machine translation (NMT). It focuses on three main tasks …
Latent variable models for machine translation and how to learn them
P Schulz - 2020 - eprints.illc.uva.nl
This thesis concerns itself with variation in parallel linguistic data and how to model it for the
purpose of machine translation. It also reflects the paradigm shift from phrase-based to …
purpose of machine translation. It also reflects the paradigm shift from phrase-based to …