Formation of Input Data Array for Classification of Printed Texts in Multilevel Tintelligent Monitoring Echnolog

Abstract

The paper describes the process of solving one of the tasks of the intellectual analysis of data - the classification of text messages. The results of researches of the process of conversion of the printed text into the typical form of the input array of model synthesizer are given. The simulator of the models is an element of the monitoring intellectual system. It is used for the synthesis of modeling classifiers. They allow you to group texts according to the given requirements. The number of correctly classified texts evaluates the effectiveness of the results of improving the process of forming signs and building points of observation. It is suggested to use as an indicator of the informativi of a sign the likelihood of its occurrence in a separate text box. The expediency of the use of such indicator of the informative nature of the sign has been experimentally confirmed.