Feed-forward and Self-organizing Neural Networks for Text Document Retrieval

The aim of this paper is to survey the feed-forward and self-organizing neural networks for the text document retrieval models, which retrieve text documents in a natural language. These models come from linguistic and conceptual approach of the text document analysis, where problems of document representation and document database creation are being solved. The proposed structure of the feed-forward and self-organizing neural network models solve the problem of the document retrieval by a user’s query which is transformed into the set of keywords. However, learning algorithm and neural network invariance, which comes from utilization of the chosen neural networks, enable the decrease of computational complexity for the language analysis of text document retrieval process.