3. The object of the research is automatic processing of the word combinations in natural language. The research objective is a development of models and methods of the semantics of the word combinations in natural language and their use in the process of analysis to improve the qualitative characteristics of the analysis of word combinations in natural language. The research methods are formal inference systems, formal grammars, mathematical logic and methods of artificial intelligence. The equipment is a personal computer. Theoretical and practical results of the research are the following: models and methods developed by the author have reached the stage of implementation, they can be used in such fields as syntax and semantic analysis of the texts in natural languages, oriented to different further applications; search of information, represented by the texts in natural languages, according to its semantic content; creation of the intelligent interfaces of different purposes; automated abstracting of the texts in natural languages; development of the machine translation systems, based on semantics. The scientific novelty consists of the following: for the first time there has been suggested a model of superficial semantics for the words, registered in the ?-dictionary, as a superposition of the semantic functions, which represent the semantic aspects of the syntactic structure of the forms in natural languages and are the basis for getting semantic formulas. Such model gives an opportunity to formalize the word combinations in natural language as noun+noun in genitive case, noun+noun and word combinations, which have a verb; for the first time there has been suggested a method for determining the equivalence of two textually different functions, which gives an opportunity to estimate the identity when defining the context in ?-formulas for the verbs by means of step-by-step comparison of the semantic functions' arguments. The method is used in the semantic-syntactic analyzers and gives an opportunity to detect the language structures with the same superficial semantics; the further development has been given to the model of the semantic dictionary of verbs, denoting action, the characteristic feature of which is presentation of ?-calculus verbs and semantic functions, which provides a way to estimate whether the language structure is correct or not; the further development has been given to the method of syntax analysis for the word combinations in natural language with some features of superficial semantics, which, unlike the others existing, takes into account the following components of the grammar production: production of the context-free grammar, predicate formula, procedure, which carries out the compression for this production. The method offers an opportunity to improve the quality of analysis, bringing it closer to the appropriate data perception. The results of the thesis have been introduced into the training at the Department of the Computer Software at the Kharkov National University of Radioelectronics, for such academic disciplines as "Formal Systems and Knowledge Modeling", "Design of the Expert Systems", in the graduation projects of specialists and in qualification papers of the masters, the Introduction Act, dated 08.12.2010, and also in the manufacturing process of the joint venture "KODA" (Kharkov, Ukraine), the Introduction Act, dated 20.12.2010. Scientific theoretical and practical results of the thesis can be used in: the organizations, dealing with the similar issues of developing the intelligent systems for the analysis of the texts in natural language - in the National Aerospace University ("KhAI"), the National Technical University of Ukraine ("KhPI"), the National Technical University of Ukraine "Kyiv Polytechnic Institute"("KPI") ; for development of systems with the features of superficial semantics of the word combinations in natural language; in the training of specialists in the field of Software Engineering.
