Development of recursive generative grammars system for solving the problem of automatic generation of language phrases intonation patterns

This paper considers the problem of developing a new class of generative grammars, providing higher quality of string generation and internal order of the rules. The purpose is to justify theoretically and to develop the module in the previously developed system for automatic generation of transcription in various European languages, which allows adding the intonation patterns, which will enable the student to study the language, its phonetic structure and features in a more effective way. In addition, it is important to solve the problem of modality identification, the distinction between narrative sentences, questions and exclamations that will provide an opportunity to speak and express many types of emotions in a foreign language without preliminary knowledge of words and rules. This type of generative grammars - “recursive generative grammars ” - should allow solving a wider class of problems through the use of dividing the entire class of rules within generative grammars to levels and sections, what will be demonstrated in the structure of the relevant rules. Methods to accomplish the above purpose consist in hybridization of generative grammar and substring of command in the form of nonterminals of generative grammar. As a result of work a principle organizing a set of trees of the states of generated string expressions in the natural language has been developed. The developed approach allows the implementation of the algorithms for generating the strings of the natural language from concerning the application of nonterminal characters of generative grammars. The proposed approach to extending the syntax of the generative grammar rules can allow providing a more complex and qualitative transformation of generated string of the characters alphabet considering the structuring process for the trees of states on the levels of recursive generation. These tools of computational linguistics can be applied as the means of modeling and analyzing the natural languages, in particular, in order to generate meaningful speech and carry out language transformations. This will allow solving more complex problems in computational linguistics and the development of linguistic software.