Incorporating fuzzy logic with neural networks for document retrieval

Abstract Classical methods such as Boolean searches, the vector space model and probabilistic retrieval have been applied to the field of document retrieval. However, these methods cannot handle the increasing demands of end-users in satisfying their needs. Recently, there has been specific interest in adapting advanced computer techniques in the field of document retrieval. The most recent attempt investigates the possibility of applying the neural-network paradigm as a means of providing end-users with a more powerful retrieval mechanism. Most of the work done thus far has been focused on incorporating neural networks into the existing classical retrieval methods. This paper demonstrates how to apply neural networks with fuzzy logic for document retrieval. In particular, the Fuzzy Kohonen Neural Network (FKNN) is used as an example to illustrate the versatility of fuzzy neural networks as applied to the document-retrieval process. In this paper, the training, pattern recall, ranking, relevance feedback and performance issues of the FKNN document-retrieval process are discussed.