Application of machine learning for E-justice

Abstract

Decision support systems (DSS) in law enforcement have a long history. Starting from the late 50s, they have been developed through several architectural approaches. Still, having a proven capability of DSSes to improve legal practice, the real-world application is limited due to multiple issues, including lack of trust, interpretability, validity, scalability, etc. The paper develops a service-based decision support platform for machine learning applications for eGovernance and internal policy modelling and presents a case study of the application of the platform for the case of migration law enforcement. We have developed a decision support platform a number of micro services that connect with each other asynchronously via the REST protocol. The artificial intelligence core of the platform was built upon a knowledge base, which includes machine learning models and methods. In this work we have developed a method of structuring, analysis of legal data models based on machine learning. In the course of computational experiment, the efficiency of the developed method was proved and the interpretation of the obtained results was performed to provide recommendations for the enhancement of administrative regulation.