Document space dimension reduction by nonlinear Hebbian neural network

This paper deals with information retrieval of text documents, and their clustering into some other feature space. The aim of this paper is to reduce the dimension of the document space by the nonlinear Hebbian neural network. As can be seen from the results, not only dimension reduction of document space is made, but also clustering of these documents into clusters. We used here the nonlinear Hebbian neural network, which is feed-forward neural network with unsupervised learning.