Extraction of implicit information from the texts in natural language: problems and methods

A semantic-oriented linguistic processor, which provides deep analysis of text in natural language and forms knowledge structures, is considered. A significant direction in development of such processors is connected with the extraction from the texts the named entities, their properties, and links, presented in implicit forms. The methods of such extraction at different levels of text analysis (lexical-grammatical, syntactical-semantic and conceptual) are proposed.