Semantic-oriented system for processing not-formalized information and presenting the results in natural language

The author considers a system for automatic formalization of different kinds of information and messages (personal data, job applications, resume, mass media reports) looking as the texts in natural language. For accomplishing the above-mentioned task the 3-Step methodology has been suggested. At the first stage the information object and their relations are extracted from the text. At the second stage these structures are analyzed by means of expert systems in order to obtain new data (total job period, knowledge of foreign languages, etc.). At the third stage the knowledge structures and obtained data are presented as the natural language components (word combinations, sentences). The suggested system is designed for personnel and recruiting agencies, as well as other services working with the flows of textual documents.