Modelo de Arquitetura da Informação para Ontology Learning no ambiente de Nota Fiscal de Consumidor Eletrônica (NFC-e)

Introduction: Ontology Learning works together with Machine Learning, Data Mining and Text Mining algorithms and provides a powerful method for representing, reusing and sharing domain knowledge. Objective: This article aims to present and discuss the evaluation of the results of the Information Architecture used to organize data that will be retrieved by Text Mining and used by Machine Learning in the construction of an ontology about the beer product. Methodology: The literature review will address topics such as: Machine Learning, Information Architecture, Ontology and Ontology Learning, a summary of the method and evaluation of information organization for ontology engineering using Term Mining in the context of Electronic Consumer Invoices (NFC-e). The research begins with a planning for the bibliographic survey and analysis of examples that deal with adopted resolution models. The exploratory bibliographic survey elaborates a literature review within the terms in the context of the research, whose sources of information used were books, articles, journals, theses and dissertations. Results: The methodology adopted to obtain the data was based on the first layer of 'Term Mining' for Ontology Learning that presented a response of each AI element as a guideline for the construction of the ontology, without the necessary presence of the domain expert. Conclusion: The AI ​​Model for Ontology Learning focuses on the form of requirements acquisition, which contributes significantly to the electronic invoice environment, generating results and facilitating the understanding of the process.