Máquinas de classificação para detectar polaridade de mensagens de texto em redes sociais

Detalhes bibliográficos
Ano de defesa: 2015
Autor(a) principal: Von Lochter, Johannes
Orientador(a): Almeida, Tiago Agostinho de lattes
Banca de defesa: Não Informado pela instituição
Tipo de documento: Dissertação
Tipo de acesso: Acesso aberto
Idioma: por
Instituição de defesa: Universidade Federal de São Carlos
Câmpus Sorocaba
Programa de Pós-Graduação: Programa de Pós-Graduação em Ciência da Computação - PPGCC-So
Departamento: Não Informado pela instituição
País: Não Informado pela instituição
Palavras-chave em Português:
Palavras-chave em Inglês:
Área do conhecimento CNPq:
Link de acesso: https://repositorio.ufscar.br/handle/20.500.14289/7903
Resumo: The popularity of social networks have attracted attention of companies. The growing amount of connected users and messages posted per day make these environments fruitful to detect needs, tendencies, opinions, and other interesting information that can feed marketing and sales departments. However, the most social networks impose size limit to messages, which lead users to compact them by using abbreviations, slangs, and symbols. Recent works in literature have reported advances in minimizing the impact created by noisy messages in text categorization tasks by means of semantic dictionaries and ontology models. They are used to normalize and expand short and messy text messages before using them with a machine learning approach. In this way, we have proposed an ensemble of machine learning methods and natural language processing techniques to find the best way to combine text processing approaches with classification methods to automatically detect opinion in short english text messages. Our experiments were diligently designed to ensure statistically sound results, which indicate that the proposed system has achieved a performance higher than the individual established classifiers.