DEVELOPMENT OF A METHODOLOGY FOR USING MULTI-AGENT SYSTEMS FOR DATA DISTRIBUTION

Balabanov Aleksandr, Kukushkina Natalia, Voronkin Evgeniy

The article discusses the existing methods of building a multi-agent system architecture for data distribution. The considered material allows us to assert that there is a fairly extensive variability of choice, but there is no universal solution for any tasks. The article provides brief extracts from a complete comparative analysis of each technique, which allows us to sufficiently evaluate each of them. In our opinion, the methodology chosen by us during the analysis has not only a sufficient set of necessary characteristics, but also its use lays a good foundation for further extensibility and scalability of the architecture.

Keywords:multi-agent system, information system architecture, comparative analysis, construction methodology, machine learning, artificial intelligence, hybrid methodology.