A new method for creating Mivar knowledge bases in tabular-matrix form for ground intelligent vehicle control systems

Abstract

This research presents a methodology for creating mivar knowledge bases in tabular-matrix form for ground intelligent vehicle control systems. At its core, this methodology is kind of instruction for analysts facing the task of formalizing knowledge in a given subject area. The result of this formalization is a “knowledge map” created according to a special proposed template. In the future, this template allows forming a knowledge base for a given subject area in the formalism of bipartite oriented mivar networks. As an example, the subject area of ground-based intelligent vehicle control systems is used as a template. The proposed methodology of knowledge formalization makes it possible to simplify the process of creating models in Wi!Mi “Razumator-Consultant” 2.1 and also to level the probability of logical collisions when designing a knowledge model in the formalism of bipartite oriented mivar networks.

Evaluation of BERT and ChatGPT models in inference, paraphrase and similarity tasks

“…For tasks in the format of production networks, they allow finding a solution with linear computational complexity. Recently, quite a few mivar expert systems have been created in various fields: systems for plant care [2], for managing educational programs [3] at the university [4], for intelligent management systems [5] of vehicles [6] and robots [7], system of making decisions about the safety of thermolabile blood components [8], for planning of robot actions [9] and for many other areas. Mivar technologies are also used in hybrid intelligent information systems (HIIS).…”