Discretion in Applying Provisions of Law: Linguistic Prospects For AI and Machine Learning

Abstract

The note presents the will of actions in using the provisions of the law as an assessment when making an opinion, or as a choice of a more favorable and correct conclusion from the number of varieties allowed by law. This legal appearance was presented in the guise of a targeted cause-and-effect graph, in which all actions are associated with a friend with another, are considered the reasons or consequences of a friend of a friend. Any of its vertex points is an event, an edge is an association, the orientation of the edges is from the corollary to the base. The features of the euro legislation on AI are considered. The essence of the artificial origin of the mind has been studied for making conclusions at personal discretion. The reprimand was made on the model of AI and machine learning, which have every chance to be: multifunctional, dynamic, expressed in the definitions of mathematical functions, risk models, credit optimization; static, expressed in relationships between data objects (chart of accounts, affairs between subjects). The Abstraction and Reasoning Corpus dataset has been demonstrated as a benchmark for robust AI. Methods of controlled machine learning have been discovered, encompassing authorized learning ("student's task"), inductive programming ("life experience"), predictive power ("scientist's task"), study campaign ("teacher's task"). The classification of arguments, methodologies and layouts has been determined for a discretionary judgment process with a pertinent case of the Euro Court of Human Rights.