• 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
10.11606/D.100.2018.tde-04052018-192006
Document
Author
Full name
Barbara Barbosa Claudino da Silva
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Paulo, 2018
Supervisor
Committee
Paraboni, Ivandre (President)
Carvalho, Ariadne Maria Brito Rizzoni
Ferreira, Fernando Fagundes
Machado-Lima, Ariane
Title in Portuguese
Reconhecimento de traços de personalidade com base em textos
Keywords in Portuguese
Big Five
Modelo dos cinco grandes fatores
Personalidade
Processamento de linguagem natural
Abstract in Portuguese
Apresentamos uma pesquisa na área de Processamento de Linguagem Natural, para reconhecimento de personalidade com base em textos da língua portuguesa. Neste trabalho utilizamos textos provenientes da rede social Facebook, em conjunto com o modelo de personalidade dos Cinco Grandes Fatores, para construir um córpus rotulado com as personalidades de seus autores e, após a identificação das características mais relevantes para o reconhecimento de personalidade, construir modelos computacionais utilizando essas características. Utilizando-se métodos provenientes de léxicos, como o dicionário LIWC ou atributos psicolinguísticos, e métodos provenientes do próprio texto, como bag of words, representação distribuída de palavras e de documentos foram desenvolvidos modelos para reconhecimento de personalidade sem a necessidade de outros métodos mais comumente utilizados para essa tarefa, como inventários ou entrevistas com psicólogos. Os resultados dos métodos de representação distribuída são ligeiramente superiores do que os resultados utilizando o dicionário LIWC, com a vantagem de não exigirem recursos dependentes de um idioma específico
Title in English
Personality traits recognition through texts
Keywords in English
Big Five
Big five model
Natural language processing
Personality
Abstract in English
We present a research proposal in the Natural Language Processing field, to recognize personality through texts in the portuguese language. Using texts from the social network Facebook we built a corpus labeled with authors Big-5 personality traits, and after identifying the most relevant atributes to recognize personality, we built computational models based on those attributes. The model was expected to recognize personality without the help of any other methods commonly used in this task, such as inventories or interviews with psychologists. Using lexical methods such as the LIWC dictionary or psycholinguistic attributes, and methods from the text itself, such as bag of words, distributed representation of words and documents, we obtained models for personality recognition without the need of other methods most commonly used for this task. The results of distributed representation methods are slightly better than the results using the LIWC dictionary, with the advantage of not requiring features dependent on a specific language
 
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.
defesa.pdf (1.26 Mbytes)
Publishing Date
2018-05-17
 
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 © 2001-2020. All rights reserved.