A.V. Pavlenko, A.Yu. Novikov

The huge volumes of text information being transmitted in media and electronic text storages call into existence of it automated processing to gain different objectives. The text sorting and classification are actual and significant fields of research. There are a lot of classification methods based on keywords analysis, that do not enough use the semantics of processed texts. The main idea of the proposed method is the application of neural networks to recognize the texts presented in the form of a semantic image. The ways of text formalization for neural networks learning are partly disclosed as well. The article is about aspects needed to create a neural network classifier and also contains some experiment results of its software implementation.

 

1. Novikov A.JU.Konceptualnye osnovy avtomatizacii diskursnogo analiza tekstov na osnove semantiki // Naukoemkie tekhnologii. 2010. № 9. S. 77-82.
2. Averjanov L.JA.Kontent-analiz. CH. 1. Monografija.
3. Zelenkov JU.G., Segalovich I.V.Verojatnostnaja model snjatija morfologicheskojj omonimii na osnove normalizujushhikh podstanovok i pozicijj sosednikh slov [EHlektronnyjj resurs]. 2005.
4. Barskijj A.B.Nejjronnye seti: raspoznavanie, upravlenie, prinjatie reshenijj. M.: Finansy i statistika. 2004.