Nodes coupling in a Bayesian network for the automatic classification of XML documents

The document classification is one of the classical task of information retrieval and it has involved numerous studies. In this paper, we are presenting a learning model for XML document classification based on Bayesian networks. This latter is a probabilistical reasoning formalism. It permits to represent depending relationships between the random variables in order to describe a problem or a phenomenon. In this article, we are proposing a model which simplifies the arborescent representation of the XML document that we have, named coupled model and we will see that this approach improves the response time and keeps the same performances of the classification.