Legal Tech: Documents’ Validation Method Based on the Associative-Ontological Approach

Abstract

Trend of Legal Tech is actively developing, thus, allowing an automation of various legal tasks solved both by professional lawyers and ordinary (common) users. Because of the legal documents properties, natural language processing (NLP) technologies are widely used in Legal Tech development. One of the tasks in Legal Tech, whose solution is necessary for both professionals and non-professionals, is a validation documents’ texts, including certain checking for the presence of mandatory structural elements in them. This article considers an implementation of a method and an algorithm for validating, for instance, the documents called “Consent to the personal data processing” in the Russian Language legal practice based on machine learning and using an associative-ontological representation of the text. Such validation occurs by checking documents through a set of rules, at that, each rule describes the documents’ structural elements. Associative ontological representation of the text makes such rules human-readable, and simplifies their adjustment and fine-tuning to the changing legislation norms. Results of experimental verification of the proposed algorithm on a set of texts of real legal documents show its effectiveness when applied to Legal Tech systems.