Um modelo classificador da lista de e-mail do Projeto Apache que combina dicionário neurolinguí­stico com ontologia

• Ano de defesa: 2011
• Autor(a) principal: Farias, Mário André de Freitas
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Alagoas; BR; Modelagem Computacional de Conhecimento; Programa de Pós-Graduação em Modelagem Computacional de Conhecimento; UFAL
• Palavras-chave em Português: Text mining; Neuro-linguistic; Ontology; Software comprehension; Mineração de texto; Neurolinguí­stica; Ontologia; Compreensão de software; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://repositorio.ufal.br/handle/riufal/847

Electronic mailing lists and discussion groups are normally used by programmers to discuss and improve tasks to be performed during software projects development. Open Source Software (OSS) projects use this lists as the primary tool for collaboration and cooperation. In project like that, normally, the developers are around the world. Thus, means of interaction and communication are needed to ensure collaboration between them, as well as efficiency in the construction and maintenance of projects this size. Mailing lists can be an important data source to discovery information useful about patterns of behavior of developer aimed at project manager. The Neurominer is a text mining tool that determines the Preferred Representational System (PRS) of software developers in a specific context. The tool has a new approach which is a combination between the Neuro-Linguistic Programming NLP theory, text mining and statistic technique. In this context, we propose the extension of this tool by applying of techniques of ontology to dictionary, allowing the combination of sensory predicates with software engineering terms, providing a greater power in the context of the dictionary. This way, the text mining matched with NLP theory and ontology appears as natural candidate that consists a solution that aiming to improve the mining of textual information through mailing lists, in order to support software project managers in making decision. This matching showed significant outcomes, proposing a efficient and effective solution.