Агентно-базований підхід до моделювання колективної роботи
Loading...
Date
2021-12-10
Authors
Гороховський, Семен
Радзієвська, Олександра
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
У роботі розглянуто використання агентно-базованого підходу для моделювання соціальних взаємодій у команді на прикладі розв’язання Wilderness Survival: A Consensus-Seeking Task. Змодельовано процес обговорення задачі та досягнення компромісу агентами. Проаналізовано, наскільки результати моделювання збігаються з гіпотезою про те, що на успіх роботи команди більший вплив мають соціальні здібності учасників, ніж інтелектуальні здібності окремих членів команди. Звернуто увагу на ситуацію, коли не всі агенти однаково зацікавлені в загальному успіху команди. Створено агентну платформу для проведення експериментів та аналізу того, як різні риси учасників команди впливають на її загальний результат.
In the modern world, it is no longer enough to simply create a product that performs its function, but it should perform it better than thousands of competitors. However, the problem is that human intellectual abilities are limited and many complex tasks are beyond the capabilities of a single person. The natural way of raising our intellectual level is to build teams to share our experience, knowledge, and worldview to create something beyond the capacity of the individual. Thus it is not surprising that according to a recent ranking, collaborative skills are considered most essential in the 21st century [2]. To cope with all challenges and create high-quality products, there should be a team whose members are experts in communication, discussion, problem-solving, and critical thinking. In addition, it is important to manage the team effectively. To do so, it is necessary to know more about the social processes which take place inside a team. Agent-based modeling can be an effective tool to gain such insights. Agent-based modeling is a powerful instrument for simulating different processes, including social ones. This technology was formed under the influence of many other fields such as artificial intelligence, sociology, game theory, and so on. In this article, a model which simulates human interaction in the framework of "Wilderness Survival: A Consensus-Seeking Task" is used to demonstrate the core principle of agent-based modeling. The group of agents complete a test by themselves and afterwards discuss their answers to reach a consensus and achieve the best score. It will be analyzed which human character traits are more important for successful collaborative work. Situations in which some team members are not interested in the team success will be identified. Also, a user interface is provided to enable running custom experiments to better understand how specific character traits impact the team results.
In the modern world, it is no longer enough to simply create a product that performs its function, but it should perform it better than thousands of competitors. However, the problem is that human intellectual abilities are limited and many complex tasks are beyond the capabilities of a single person. The natural way of raising our intellectual level is to build teams to share our experience, knowledge, and worldview to create something beyond the capacity of the individual. Thus it is not surprising that according to a recent ranking, collaborative skills are considered most essential in the 21st century [2]. To cope with all challenges and create high-quality products, there should be a team whose members are experts in communication, discussion, problem-solving, and critical thinking. In addition, it is important to manage the team effectively. To do so, it is necessary to know more about the social processes which take place inside a team. Agent-based modeling can be an effective tool to gain such insights. Agent-based modeling is a powerful instrument for simulating different processes, including social ones. This technology was formed under the influence of many other fields such as artificial intelligence, sociology, game theory, and so on. In this article, a model which simulates human interaction in the framework of "Wilderness Survival: A Consensus-Seeking Task" is used to demonstrate the core principle of agent-based modeling. The group of agents complete a test by themselves and afterwards discuss their answers to reach a consensus and achieve the best score. It will be analyzed which human character traits are more important for successful collaborative work. Situations in which some team members are not interested in the team success will be identified. Also, a user interface is provided to enable running custom experiments to better understand how specific character traits impact the team results.
Description
Keywords
агентна парадигма, мультиагентна система, Python, Mesa, стаття, agent-based modeling, multiagent system, collaborative work, Python, Mesa, article
Citation
Гороховський С.С. Агентно-базований підхід до моделювання колективної роботи / Гороховський С. С., Радзієвська О. В. // Наукові записки НаУКМА. Комп'ютерні науки. - 2021. - Т. 4. - С. 60-63. - https://doi.org/10.18523/2617-3808.2021.4.60-63