Neural networks are a field of Computer Science related to Artificial Intelligence. The field aims at implementing mathematical models that are similar to biological neural structures. It is also capable of adapting its parameters as a result of interactions with the external environment, gradually improving their performance in the solution of a particular task. By using neural networks in computer information retrieval systems one can assign a dynamic character to those systems by allowing the representation of documents to be reevaluated and modified according to specifications of relevance attributed by users to retrieved documents. This work presents the main initiatives in applying neural networks concepts to information retrieval systems and evaluates its applicability to large document databases as is the case of the Web.
Neural networks; Information retrieval; Adaptive systems
