Методика иерархической разметки текстовых данных на основе онтологического представления сценариев обработки персональных данных

Abstract

Intensive digitalization in all spheres of human activity constantly increases the amount of personal data collected and processed for various services. It is necessary to automate the process of formalization and structuring of user agreements written in natural language, because most users agree with their terms without realizing the potential consequences due to the complexity of these documents. This paper proposes a text data markup technique that takes into account possible semantic links between markup elements and allows annotating training samples for text classifiers. The development and testing of a software tool that implements the proposed methodology has been performed. The developed tool is planned to be used for further research in the field of formalization of user agreements.