[PDF] Integrating Weakly Supervised Word Sense Disambiguation into Neural Machine Translation | Semantic Scholar (2024)

[PDF] Semantic Reader

This work proposes CmBT (Contextually-mined Back-Translation), an approach for improving multi-sense word translation leveraging pre-trained cross-lingual contextual word representations (CCWRs), and shows that the system improves on the translation of difficult unseen and low frequency word senses.

This paper proposes to build knowledge and supervised based Multilingual Word Sense Disambiguation systems and address the annotation scarcity problem for MWSD by transferring annotations from rich sourced languages.

This paper presents CluBERT, an automatic and multilingual approach for inducing the distributions of word senses from a corpus of raw sentences that attains state-of-the-art results on the English Word Sense Disambiguation tasks and helps to improve the disambiguated performance of two off- the-shelf WSD models.

This work proposes Extended WSD Incorporating Sense Embeddings (EWISE), a supervised model to perform WSD by predicting over a continuous sense embedding space as opposed to a discrete label space, which allows EWISE to generalize over both seen and unseen senses, thus achieving generalized zero-shot learning.

This work introduces a simple and scalable approach to resolve translation ambiguity by incorporating a small amount of extra-sentential context in neural neural models to translate ambiguous source words better than strong sentence- level baselines and comparable document-level baselines while reducing training costs.

The capabilities of LLMs to translate “ambiguous sentences” are studied, and two ways to improve their disambiguation capabilities are proposed, through a) in-context learning and b) fine-tuning on carefully curated ambiguous datasets.

This paper presents an evaluation benchmark on WSD for machine translation for 10 language pairs, comprising training data with known sense distributions, and builds upon the wide-coverage multilingual sense inventory of BabelNet, the multilingual neural parsing pipeline TurkuNLP, and the OPUS collection of translated texts from the web.

This thesis proposes a method to decide whether two occurrences of the same noun in a source text should be translated consistently, and designs sense-aware MT systems that select the correct translations of ambiguous words by performing word sense disambiguation (WSD).

This work uses contextualized word embeddings for an unsupervised Arabic WSD that is based on linguistic markers and uses sentence-BERT Transformer pre-trained models, which yields encouraging results that outperform other existing un supervised neural AWSD approaches.

This work investigates the statistical relation between word frequency rank and word sense number distribution and proposes a Z-reweighting method on the word level to adjust the training on the imbalanced dataset.

...

...

This paper investigates a new strategy for integrating WSD into an SMT system, that performs fully phrasal multi-word disambiguation, and provides the first known empirical evidence that lexical semantics are indeed useful for SMT, despite claims to the contrary.

While a baseline NMT system disambiguates frequent word senses quite reliably, the annotation with both sense labels and lexical chains improves the neural models’ performance on rare word senses.

It is shown that the word-translation system can be used to improve performance on a simplified machine-translation task and can effectively and accurately prune the set of candidate translations for a word.

This work demonstrates that WSD systems can be adapted to help SMT, thanks to three key achievements: it considers a larger context for WSD than SMT can afford to consider, and it adapts the number of senses per word to the ones observed in the training data.

This paper validates the hypothesis and proposes a simple and flexible framework, which enables the NMT model to only focus on the relevant sense type of the input word in current context and achieves substantial improvements on every test set over competitive baselines.

A sense-based translation model to integrate word senses into statistical machine translation based on a nonparametric Bayesian topic model that automatically learns sense clusters for words in the source language is proposed.

A multisense embedding model based on Chinese Restaurant Processes is introduced that achieves state of the art performance on matching human word similarity judgments, and a pipelined architecture for incorporating multi-sense embeddings into language understanding is proposed.

Training on a large, open-domain corpus (Europarl) and including word senses as contextual features in maxent-based translation models yields significant improvements in machine translation from English to Portuguese.

Empirical evidence is provided that existing NMT systems in fact still have significant problems in properly translating ambiguous words, and methods are described that model the context of the input word with context-aware word embeddings that help to differentiate the word sense before feeding it into the encoder.

...

...