Multilevel Natural Language Processing for Intelligent Information Retrieval and Text Mining

This paper considers the application of methods for multilevel natural language processing to information retrieval and text mining. It is shown how to use linguistic information about text structure and to what kind of problems this information is applicable. The information is obtained by syntactic, semantic, and discourse parsing. The results of developing methods for multilevel processing of Russian-language content and their application in semantic and question-answering search, information extraction from texts, text classification, and psycholinguistic analysis of texts are presented.