• JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
 
  Bookmark and Share
 
 
Master's Dissertation
DOI
https://doi.org/10.11606/D.104.2021.tde-31032021-123649
Document
Author
Full name
Gabriela Massoni
Institute/School/College
Knowledge Area
Date of Defense
Published
São Carlos, 2021
Supervisor
Committee
Stern, Rafael Bassi (President)
Cerqueira, Andressa
Prates, Marcos Oliveira
Title in Portuguese
Análise de textos por meio de processos estocásticos na representação word2vec
Keywords in Portuguese
Modelos de predição
Processamento de linguagem natural
Processos estocásticos
Representação vetorial de palavras
Abstract in Portuguese
Dentro do campo de Processamento de Linguagem Natural (NLP), o modelo word2vec vêm sendo bastante explorado no campo da representação vetorial de palavras. Ele é uma rede neural que se baseia na hipótese de que palavras semelhantes tem contextos semelhantes. Na literatura em geral, o texto é representado pelo vetor de médias das representações das suas palavras, que, por sua vez, é utilizado como variável explicativa em modelos preditivos. Um alternativa é, além da médias, utilizar outras medidas, como desvio-padrão e medidas de posição. Porém, o uso destas medidas supõe que a ordem das palavras não importa. Assim, nesta dissertação exploramos o uso de processos estocásticos, em particular, Modelos de Série Temporal e Modelos Ocultos de Markov (HMM), para incorporar a ordem cronológica das palavras na construção das variáveis explicativas a partir da representação vetorial dada pelo word2vec. O impacto desta abordagem é medido com a qualidade dos modelos preditivos aplicados à dados reais e comparado às abordagens usuais. Para os dados analisados, as abordagens propostas tiveram um resultado superior ou equivalente às abordagens usuais na maioria dos casos.
Title in English
Text mining with stochastic process in word2vec representation
Keywords in English
Natural language processing
Prediction models
Stochastic process
Word vector representation
Abstract in English
Within the field of Natural Language Processing (NLP), the word2vec model has been extensively explored in the field of vector representation of words. It is a neural network that is based on the hypothesis that similar words have similar contexts. In the literature in general, the text is represented by the mean vector of the representations of its words, which, in turn, is used as an explanatory variable in predictive models. An alternative is, in addition to averages, to use other measures, such as standard deviation and position measures. However, the use of these measures assumes the order of the words does not matter. Thus, in this dissertation we explore the use of stochastic processes, in particular, Time Series Models and Hidden Markov Models (HMM), to incorporate the chronological order of words in the construction of explanatory variables from the vector representation given by word2vec. The impact of this approach is measured with the quality of the predictive models of real data and compared to the usual ones.For the analysed data, the proposed approaches have a result that is superior to or equivalent to the usual approaches in most cases.
 
WARNING - Viewing this document is conditioned on your acceptance of the following terms of use:
This document is only for private use for research and teaching activities. Reproduction for commercial use is forbidden. This rights cover the whole data about this document as well as its contents. Any uses or copies of this document in whole or in part must include the author's name.
Publishing Date
2021-03-31
 
WARNING: Learn what derived works are clicking here.
All rights of the thesis/dissertation are from the authors
CeTI-SC/STI
Digital Library of Theses and Dissertations of USP. Copyright © 2001-2024. All rights reserved.