An Artificial Neural Network Perspective on Knowledge Representation from Databases: The Use of a Multilayer Perceptron for Data Clusters Cartography

III order to display the thematic organisation of knowledge space, we compare and evaluate two different cartographic approaches: principal components analysis (PCA), and artificial neural networks (ANNs) using a multilayer perceptron (MLP) in "self-association" mode. This kind of MLP can be used to perform a PCA if the activation function is set to the identity function. This allows us to look for the non-linear activation function which best fits the data structure. We present an evaluation criteria and the results and maps obtained with both methods.