Volume 24, Issue 4 (Winter 2023)                   Advances in Cognitive Sciences 2023, 24(4): 115-131 | Back to browse issues page


XML Persian Abstract Print


Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:

Zaeim Kohan N, Nourbakhsh A. Presenting an emotional-normative agent model through combining E-X-machine and NOE architecture for investigating the phenomenon of bank rush. Advances in Cognitive Sciences 2023; 24 (4) :115-131
URL: http://icssjournal.ir/article-1-1467-en.html
1- Master’s Student in Software Engineering, Department of Computer Engineering & Information Technology, Lahijan Branch, Islamic Azad University, Lahijan, Iran
2- Assistant Professor, Department of Computer Engineering & Information Technology, Lahijan Branch, Islamic Azad University, Lahijan, Iran
Abstract:   (920 Views)
Introduction
In recent years, agent-based simulation has been successfully used in the study of various economic phenomena, which has led to the emergence of the interdisciplinary research system of agent-based computational economics.
One of the economic phenomena researchers have investigated in different ways is the problem of bank runoff. A bank raid is a situation where depositors withdraw their deposits from the bank due to fear of their deposits’ safety.
The question is whether norms, such as society’s attention to rumors, social responsibility, and prioritizing collective interests over individual desires, can reduce such phenomena’s probability.
In this research, by combining two architectures of emotional and normative agents, an emotional-normative model is proposed through which, while simulating the phenomenon of bank rush, the role of normative variables in preventing this phenomenon is investigated. As a result, the main goal of this research is to present and test an emotional-normative model and investigate the role of normative reasoning in preventing the adverse results of economic phenomena known as ecological surprises.
Methods
The modeling approach chosen to implement the simulation of the bank rush phenomenon is the combination of FSM, Emotion, and NOE models.
A Finite State Machine (FSM) is a computational model that can be used to simulate sequential logic or, in other words, to represent and control the flow of execution. FSMs can model problems in many fields, including mathematics, artificial intelligence, games, and linguistics.
In this article, according to the model presented by Sakellario et al. (3), emotions are defined as passion (the level of motivation to act). The corresponding mode is applied to the agents. These states reflect the agent’s speed in maintaining or changing how he interacts with the simulation world. This willingness to act shows the agent’s readiness to welcome other agents, situations, and events.
On the other hand, the NOE agent processes emotions based on the norm evaluation output and, similar to the OCC model, considers both the type and intensity of emotions to perform a specific action. Therefore, violating the norms in such a factor causes negative feelings (18).
Agents with norms refer to their normative memory upon entering each state to determine whether they are in a normative situation. If the answer is positive, due to the agent’s institutionalization degree of desire agent, he will behave according to the norm.
Therefore, by examining the normative nature of each state by the agent (Identification), and by gradually increasing the norm in the agent (Instantiation), and changing the agent’s behavior due to the existence of norms, the NOE model is used in the normative reasoning, and finally, Norm Fulfillment by evaluating compliance or non-compliance with the norm. The related rewards and punishments are implemented.
The three normative variables that are vital in preventing the Bank Rush phenomenon are:
• The agent’s previous experience of complying or not complying with the norm
• The agent’s behavior results in the prediction of the future
• The degree of guilt in the perpetrator
Results
The agents’ emotions are aroused by activating sub-agents and dispersing negative rumors. The value of their Cash-Need variable, which represents the amount of cash the agent feels secure with, increases rapidly (1).
To check the impact of the normative variables and determine how the model works, for simulating, the authors activate the normative variables, which in the previous bankruptcy experiments of all banks has occurred (By using 15 sub-factors and active contagion of emotions).
Moreover, the initial value of the two variables, init-norm and init-agreeable, which respectively represent the institutionalization degree of the norm in a factor and the degree of agreeableness in the agent, have been considered at the average level (0.5). Three variables- previous experience, future prediction of crossing the norm, and feeling guilty when crossing the norm- are used for the agent’s next decisions (Table 1).
Table 1. Simulation results with normative reasoning activation
Average Days Before Failure Failure Percent Init Normative Population Percent Punishment System Reward System Normative Reasoning
5 100% 20% Inactive Inactive Active
22 50% 50% Inactive Inactive Active
- 0% 75% Inactive Inactive Active






Conclusion
In normative reasoning activation mode, the most important factor influencing the simulation result is the number of agents who have accepted the desired norm at the beginning of the simulation. The results presented in Table 1 show that the higher the number of primary normative factors, the lower the probability of bank failure.
Noteworthy that the phenomenon of bank rush generally occurs in a short time. In this way, immediately after creating negative rumors, negative feelings related to it are created at the community level, and the withdrawal of money from banks starts quickly. Certainly, in such a time, there is not much opportunity for the emergence of the norm in people. The most important issue is the number of people who have already accepted the desired norm for various reasons. In this case, the very low growth of normative variables confirms this point (Fig. 1).


 
  
  
 
Figure 1. Partial changes of normative variables in the simulation interval






Ethical Considerations
Compliance with ethical guidelines
The authors have not used a questionnaire in this article, and people’s privacy has not been violated. The report’s presentation is only based on the repeated execution of the simulation, and no ethical notions were involved in this research.
Authors’ contributions
All the authors participated in this research and had an equal share in writing and editing this article.
Funding
This research was done at the personal expense of the authors.
Acknowledgments
The authors express their gratitude to all those who helped to write this article.
Conflict of interest
There is no conflict of interest among the authors.
 
Full-Text [PDF 1435 kb]   (196 Downloads)    
Type of Study: Research |
Received: 2022/09/19 | Accepted: 2022/12/28 | Published: 2023/02/19

Add your comments about this article : Your username or Email:
CAPTCHA

Send email to the article author


Rights and permissions
Creative Commons License This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

Designed & Developed by : Yektaweb