Sense representations for Portuguese: experiments with sense embeddings and deep neural language models

Sense representations for Portuguese: experiments with sense embeddings and deep neural language models

Sense representations have gone beyond word representations like Word2Vec, GloVe and FastText and achieved innovative performance on a wide range of natural language processing tasks. Although very useful in many applications, the traditional approaches for generating word embeddings have a strict d...

Full description

Saved in:
Bibliographic Details
Published in:Language resources and evaluation Vol. 55; no. 4; pp. 901 - 924
Main Authors: Rodrigues da Silva, Jéssica, Caseli, Helena de M.
Format: Journal Article
Language:English
Published: Dordrecht Springer Netherlands 01-12-2021
Springer Nature B.V
Subjects:
Computational Linguistics
Computer Science
Experiments
Language
Language and Literature
Language modeling
Linguistics
Natural language processing
Original Paper
Performance enhancement
Performance evaluation
Portuguese language
Representations
Semantic analysis
Semantics
Social Sciences
Syntax
Words (language)
Deep neural language models
Sense embeddings
Portuguese
Word sense disambiguation
Word embeddings
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Sense representations have gone beyond word representations like Word2Vec, GloVe and FastText and achieved innovative performance on a wide range of natural language processing tasks. Although very useful in many applications, the traditional approaches for generating word embeddings have a strict drawback: they produce a single vector representation for a given word ignoring the fact that ambiguous words can assume different meanings. In this paper, we explore unsupervised sense representations which, different from traditional word embeddings, are able to induce different senses of a word by analyzing its contextual semantics in a text. The unsupervised sense representations investigated in this paper are: sense embeddings and deep neural language models. We present the first experiments carried out for generating sense embeddings for Portuguese. Our experiments show that the sense embedding model (Sense2vec) outperformed traditional word embeddings in syntactic and semantic analogies task, proving that the language resource generated here can improve the performance of NLP tasks in Portuguese. We also evaluated the performance of pre-trained deep neural language models (ELMo and BERT) in two transfer learning approaches: feature based and fine-tuning, in the semantic textual similarity task. Our experiments indicate that the fine tuned Multilingual and Portuguese BERT language models were able to achieve better accuracy than the ELMo model and baselines.
ISSN:1574-020X
1574-0218
DOI:10.1007/s10579-020-09525-1