Social Modeling for Multi-agent Systems

2016 ◽  
Author(s):  
Gita Sukthankar
Keyword(s):  
Multi Agent Systems ◽  
Social Modeling ◽  
Agent Systems ◽  
Multi Agent

scholarly journals Application of social modeling using agent based approach in scientific and technical development, implementation of R&D and maintenance of innovative potential

2019 ◽  
Vol 81 (3) ◽  
pp. 339-359
Author(s):  
V. I. Abramov ◽  
A. N. Kudinov ◽  
D. S. Evdokimov
Keyword(s):  
Social Sciences ◽  
Fine Tuning ◽  
Technical Development ◽  
Macro Level ◽  
Multi Agent Systems ◽  
Social Modeling ◽  
Agent Based ◽  
Agent Systems ◽  
Multi Agent ◽  
Innovative Potential

Agent based models (ABM) and multiagent systems (MAS) can be used to solve problems in many fields of research - from natural and computer to economics and social sciences. Many natural and social phenomena can be represented in form of complex simulations so over time agent models and multi-agent systems have proven to be a really powerful tool in areas such as economics and trade, health, urban planning and social sciences. In addition multi-agent systems can be represented as an artificial society similar to a human one and consisting of entities with characteristics similar to human ones, for example in terms of autonomy and intelligence. ABM are based on the principle of objective orientation as well as the evolution (training) of agents in the process of modeling various variants of the proposed events. Despite the apparent simplicity of the rules of interaction between agents the results are usually non-obvious and quite meaningful. ABM can be developed both at the micro level and represent models with multiple agents at the macro level. The concept of multi-agent systems which immediately gained followers and support in both scientific circles and industrial communities, first started talking in the mid-1980s. Over the past thirty years, the methodology of IAU creation has been constantly improved: technologies and tools for its promotion and use in the management of large-scale network structures (such as defense systems, energy, health, transport, logistics, urban management, collective robotics, etc.) have been actively developed. The scope of application of MAS is very wide. The analysis of implemented MAS proves that currently the tool is the most advanced technology for managing any objects built on the principles of self-organization. However, despite all the evidence of positive prospects for the introduction of AOM technology the number of examples of its successful application to date is small. In this regard creation of new platforms for discussion of international experience and improvement of the approach to simulation modeling in general is especially important for further dissemination of AMB and MAS. Creation of an open consortium for agent-oriented modeling as well as promotion of development, communication and dissemination of research results as well as implementation of educational activities together will contribute to the development of agent based modeling. The analysis and review of existing methodology of social modeling with use of agent based approach in the application to scientific and technical development, implementation of R&D and maintenance of innovative potential showed that models characterized by complex multi-level processes and interactions of agents have more capacious software structures which depend more on the "fine" tuning of the agents themselves. Such models can contain and use a voluminous set of data, and in the field of economic research tend to focus on the analysis and forecasting of various socio-economic processes at the macro level.

Sign in / Sign up

Export Citation Format

Share Document