Use of cluster analysis for documents processing in retrieval system

The role of information retrieval systems becomes every year more and more actual. The e-information doubles each 7–9 years, therefore, the solution of the problem of obtaining relevant information from large volume of data is very important. The main stages of creation of the information retrieval system are described. The news from a portal ria.ru for 2011 is used as practical material. The problems arising in processing a large amount of data are described; the mechanisms of their solution are proposed. Search quality is evaluated by two key parameters: the accuracy and completeness. The most important factor is response time. The mechanism of reduction of the response time without loss of search quality is offered. This mechanism is based on the synthesis of cluster analysis and genetic algorithm.