• 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
Document
Author
Full name
Thales Sinelli Lima
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Carlos, 2019
Supervisor
Committee
Amancio, Diego Raphael (President)
Chalco, Jesús Pascual Mena
Liang, Zhao
Travieso, Gonzalo
Title in Portuguese
Segmentação em tópicos utilizando redes complexas
Keywords in Portuguese
Processamento de língua natural
Redes complexas
Segmentação de tópicos
Abstract in Portuguese
Redes complexas têm sido utilizadas para representar sistemas reais em várias áreas. Um dos sistemas estudados utilizando redes são as línguas naturais, com ênfase em estudos sobre as estruturas textuais. Nessa dissertação introduzimos modelos baseados em relações semânticas de palavras ou sentenças para a segmentação de tópicos. Aqui são abordados os principais conceitos utilizados no desenvolvimento dessa dissertação, assim como os resultados principais obtidos pelos métodos propostos. Vemos nos resultados obtidos que utilizando uma combinação de técnicas baseadas em redes previamente propostas juntamente com abordagens mais modernas de NLP obtemos resultados melhores que os modelos atuais baseados em redes.
Title in English
Complex network approach to topic segmentation
Keywords in English
Complex networks
Natural language processing
Topic segmentation
Abstract in English
Complex Networks have been used to represent real systems of several areas. One of such system is natural language, with enpahsis on text structure. In this thesis, we introduce models based on semantic relations between words or sentences for topic segmentation. Here, we discuss the main concepts used during the development of this dissertation, as well as the main results achieved by the models proposed. The results obtained by combining previous complexnetwork based models with modern NLP methods perform considerably better than the current complex network baseline.
 
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
2019-11-01
 
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.