Processing the queries in the natural language based on semantic networks and templates

The paper considers the method of the formal mechanism for the processing of queries, which involves interactive input of the request by a user in the form of free text on some problem-oriented subset of natural language in order to obtain the necessary samples. As a basis, the electronic testing system for the marine professional preparation in Admiral Makarov State University of Maritime and Island Shipping is taken. The algorithm of the mechanism functioning is developed; it significantly reduces and simplifies the getting of the required information to the user in order to analyze and define the structure of the organization, priority posts, etc. for further solution of the staff issue. The algorithm is implemented on some formal natural language subset, which is limited by two classes of sentences: simple interrogative and imperative. Particular attention in the implementation of the software was given to the use of the effective methods for understanding and processing of semantics (meaning) of the query entered by the user. This allows with maximum accuracy to recognize the text, regardless of the complexity, possible ambiguity, used writing style, words, syntactic relationships and idioms. It is proposed to use the layered semantic networks, which allow to identify the main semantic structures of the query language 1CQ platform «1C: Enterprise 8» and link them with the syntax of the description of the request. The knowledge base for the language 1CQ, for which according to the available information about the syntax at the language the entire system of templates was built to simplify the analysis on the request in a language close to the natural, is developed.