Use of methods of intellectual text processing and large language models for analysis of information about legal relations in normative legal acts

Abstract

The article investigates the application of methods of machine processing of natural language texts (NLP, “natural language processing”) for automated extraction and evaluation of information about legal relations from the texts of normative legal acts with the use of large language models (LLM, “large language models”). The research is based on the hypothesis that the use of NLP and LLM technologies for solving tasks of this kind requires a reliance on fundamental ideas in the field of theory and philosophy of law. As such, the theory of legal relation developed in the works of P. P. Serkov was chosen, which is based on the idea of legal relation as a logical structure that forms a stable space and connects subjects for the materialization of subjective needs and goals on the principles of correlation and comparability, due to the impact on subjects of the ideological content of the totality of legal norms and circumstances of the regulated situation on the conditions of mutual reciprocity and parity of subjective rights and obligations. On the basis of experiments the effectiveness of the proposed methods and the significance of interrelation with the fundamental concept of legal relations are demonstrated.