 Subjects -> SOCIOLOGY (Total: 553 journals)
| A B C D E F G H I J K L M N O P Q R S T U V W X Y Z | | The end of the list has been reached or no journals were found for your choice. |
|
|
- Unpacking a Black Box: A Conceptual Anatomy Framework for Agent-Based
Social Simulation Models
Authors: ozge.dilaver@northumbria.ac.uk (Ozge Dilaver; Nigel Gilbert
Abstract: Ozge Dilaver and Nigel Gilbert: This paper aims to improve the transparency of agent-based social simulation (ABSS) models and make it easier for various actors engaging with these models to make sense of them. It studies what ABSS is and juxtaposes its basic conceptual elements with insights from the agency/structure debate in social theory to propose a framework that captures the ‘conceptual anatomy’ of ABSS models in a simple and intuitive way. The five elements of the framework are: agency, social structure, environment, actions and interactions, and temporality. The paper also examines what is meant by the transparency or opacity of ABSS in the rapidly growing literature on the epistemology of computer simulations. It deconstructs the methodological criticism that ABSS models are black boxes by identifying multiple categories of transparency/opacity. It argues that neither opacity nor transparency is intrinsic to ABSS. Instead, they are dependent on research habitus - practices that are developed in a research field that are shaped by structure of the field and available resources. It discusses the ways in which thinking about the conceptual anatomy of ABSS can improve its transparency.
PubDate: Tue, 31 Jan 2023 12:59:00 +000
- Using Machine Learning for Agent Specifications in Agent-Based Models and
Simulations: A Critical Review and Guidelines
Authors: m.aleebrahimdehkordi@tudelft.nl (Molood Ale Ebrahim Dehkordi; Jonas Lechner, Amineh Ghorbani, Igor Nikolic, Émile Chappin Paulien Herder
Abstract: Molood Ale Ebrahim Dehkordi, Jonas Lechner, Amineh Ghorbani, Igor Nikolic, Émile Chappin and Paulien Herder: Agent-based modelling and simulation (ABMS), whether simple toy models or complex data-driven ones, is regularly applied in various domains to study the system-level patterns arising from individual behaviour and interactions. However, ABMS still faces diverse challenges such as modelling more representative agents or improving computational efficiency. Research shows that machine learning (ML) techniques, when used in ABMS can address such challenges. Yet, the ABMS literature is still marginally leveraging the benefits of ML. One reason is the vastness of the ML domain, which makes it difficult to choose the appropriate ML technique to overcome a specific modelling challenge. This paper aims to bring ML more within reach of the ABMS community. We first conduct a structured literature review to investigate how the ABMS process uses ML techniques. We focus specifically on articles where ML is applied for the structural specifications of models such as agent decision-making and behaviour, rather than just for analysing output data. Given that modelling challenges are mainly linked to the purpose a model aims to serve (e.g., behavioural accuracy is required for predictive models), we frame our analysis within different modelling purposes. Our results show that Reinforcement Learning algorithms may increase the accuracy of behavioural modelling. Moreover, Decision Trees, and Bayesian Networks are common techniques for data pre-processing of agent behaviour. Based on the literature review results, we propose guidelines for purposefully integrating ML in ABMS. We conclude that ML techniques are specifically fit for currently underrepresented modelling purposes of social learning and illustration; they can be used in a transparent and interpretable manner.
PubDate: Tue, 31 Jan 2023 12:58:00 +000
- A Methodology to Develop Agent-Based Models for Policy Support Via
Qualitative Inquiry
Authors: v.nespeca@tudelft.nl (Vittorio Nespeca; Tina Comes Frances Brazier
Abstract: Vittorio Nespeca, Tina Comes and Frances Brazier: Qualitative research is a powerful means to capture human interactions and behavior. Although there are different methodologies to develop models based on qualitative research, a methodology is missing that enables to strike a balance between the comparability across cases provided by methodologies that rely on a common and context-independent framework and the flexibility to study any policy problem provided by methodologies that focus on capturing a case study without relying on a common framework. Additionally, a rigorous methodology is missing that enables the development of both theoretical and empirical models for supporting policy formulation and evaluation with respect to a specific policy problem. In this article, the authors propose a methodology targeting these gaps for ABMs in two stages. First, a novel conceptual framework centered on a particular policy problem is developed based on existing theories and qualitative insights from one or more case studies. Second, empirical or theoretical ABMs are developed based on the framework and generic models. This methodology is illustrated by an example application for disaster information management in Jakarta, resulting in an empirical descriptive agent-based model.
PubDate: Tue, 31 Jan 2023 12:57:00 +000
- Agent-Based Simulation of Land Use Governance (ABSOLUG) in Tropical
Commodity Frontiers
Authors: marius.vonessen@gmail.com (Marius von Essen; Eric F Lambin
Abstract: Marius von Essen and Eric F Lambin: Well-designed land use governance that involves multiple stakeholders is crucial to reducing deforestation in tropical commodity frontiers. The effectiveness of different policy mixes is difficult to assess due to long implementation times and challenges to conducting real-world experiments. Here we introduce an agent-based simulation of land use governance (ABSOLUG) to examine the interactions among governments, commodity producers, and civil society and assess the impacts of different land use governance approaches on deforestation. The model represents a generic commodity producing landscape in the tropics with a central marketplace and features four groups of agents: largeholders, smallholders, NGOs, and a government. The objective of largeholders and smallholders is to generate profits through the production of commodity crops. Statistical evaluation through local and global sensitivity analyses shows that the model is robust, and few parameters show threshold behaviors. We used a hands-off and a proactive-government scenario to evaluate the model operationally. The hands-off scenario was inspired by high rates of tropical deforestation in the second half of the 20th century and the pro-active government scenario by a few recent cases of forest transition countries. The hands-off scenario led to quasi-complete deforestation of the landscape at the end of the simulation period. Deforestation in the proactive-government scenario decreased and eventually stopped in the second half of the simulation period, followed by reforestation.
PubDate: Tue, 31 Jan 2023 12:56:00 +000
- Emergency Warning Dissemination in a Multiplex Social Network
Authors: haizhong.wang@oregonstate.edu (Charles Koll; Michael Lindell, Chen Chen Haizhong Wang
Abstract: Charles Koll, Michael Lindell, Chen Chen and Haizhong Wang: Disasters vary in many characteristics, but their amount of forewarning—the amount of time remaining until the disaster strikes—is a crucial factor affecting the dissemination of emergency warnings. People canbe warned by public safety officials through broadcast channels, such as commercial TV and radio, that transmit simultaneous warnings to mass audiences. In addition, however, warnings are also transmitted by peers through informal warning networks that operate through contagion from one person to another. This paper establishes an interdisciplinary agent-based model with Monte Carlo simulations to assess the relative effects of these broadcast and contagion processes in a multiplex social network. This multiplex approach models multiplechannels of informal communication—phone, word-of-mouth, and social media—that vary in their attribute values. Each agent is an individual in a threatened community who, once warned, has a probability of warning others in their social network using one of these channels. The probability of an individual warning others is based on their warning source and the time remaining until disaster impact, among other variables. We model warning dissemination using simulation parameter values chosen from empirical studies of disaster warnings along with the spatial aspects of the Coos Bay, OR, USA and Seaside, OR, USA communities. Results indicate that the initial broadcast size has a negative correlation with the critical percolation threshold, which varies from approximately 1–5%, depending on the size of an initial broadcast. A sensitivity analysis on the model parameters indicates that, along with initial broadcast size and sharing probability, forewarning and confidencein the warning significantly affect the total number of warning recipients. The results generated from this study identify areas for future research and can inform community officials about the effects of event and community characteristics on the dissemination of emergency warnings in their communities.
PubDate: Tue, 31 Jan 2023 12:55:00 +000
- A Geospatial Bounded Confidence Model Including Mega-Influencers with an
Application to Covid-19 Vaccine Hesitancy
Authors: anna.haensch@tufts.edu (Anna Haensch; Natasa Dragovic, Christoph Borgers Bruce Boghosian
Abstract: Anna Haensch, Natasa Dragovic, Christoph Borgers and Bruce Boghosian: We introduce a geospatial bounded confidence model with mega-influencers, inspired by Hegselmann and Krause (2002). The inclusion of geography gives rise to large-scale geospatial patterns evolving out of random initial data; that is, spatial clusters of like-minded agents emerge regardless of initialization. Mega-influencers and stochasticity amplify this effect, and soften local consensus. As an application, we consider national views on Covid-19 vaccines. For a certain set of parameters, our model yields results comparable to real survey results on vaccine hesitancy from late 2020.
PubDate: Tue, 31 Jan 2023 12:54:00 +000
- Confirmation Bias as a Mechanism to Focus Attention Enhances Signal
Detection
Authors: michael.vogrin@uni-graz.at (Michael Vogrin; Guilherme Wood Thomas Schmickl
Abstract: Michael Vogrin, Guilherme Wood and Thomas Schmickl: Confirmation bias has been traditionally seen as a detrimental aspect of the human mind, but recently researchers have also argued that it might be advantageous under certain circumstances. To test this idea, we developed a minimally complex agent-based model in which agents detect binary signals. Compared to unbiased agents, biased agents have a higher chance to detect the signal they are biased for, and a lower chance to detect other signals. Additionally, detecting signals is associated with benefits, while missing signals is associated with costs. Given this basic assumptions, biased agents perform better than unbiased agents in a wide variety of possible scenarios. Thus, we can show that confirmation bias increases the fitness of agents in an evolutionary algorithm. We conclude that confirmation bias sensitizes agents towards a certain type of data, which allows them to detect more signals. We discuss our findings in relation to topics such as polarization of opinions, the persistence of first impressions, and the social theory of reasoning.
PubDate: Tue, 31 Jan 2023 12:53:00 +000
- Conditions and Effects of Norm Internalization
Authors: batzke@uni-kassel.de (Marlene Batzke; Andreas Ernst
Abstract: Marlene Batzke and Andreas Ernst: Norm internalization refers to the process of adoption of normative beliefs by individuals, thus representing a link between individual and social change. However, there are several questions regarding norm internalization which need to be answered. These include understanding under which circumstances norm internalization does occur by considering the effects of internalizing either a certain norm or even conflicting norms. To investigate the conditions and effects of norm internalization, we developed a theoretical agent-based model called “DINO”, comprising a norm internalization process grounded on a psychological model of decision-making, considering different types of norms, goals, and habits as well as inter-individual differences. Our conceptualization of personal norms introduces a new level of complexity, allowing for more than one norm to be internalized and either approved or disapproved. Our conceptual model was implemented within the framework of a 3-person Prisoner’s Dilemma game. Results showed that playing with cooperative others generally facilitated norm internalization. Norm internalization encouraged norm compliance and affected behavioural stability and payoff equality. We discuss how our results relate to empirical findings and theoretical literature, providing a bridge between theory development and empirically testable hypotheses and between psychological micro-level phenomena and social dynamics.
PubDate: Tue, 31 Jan 2023 12:52:00 +000
- Review of: Escape from Model Land. How Mathematical Models Can Lead Us
Astray and What We Can Do About It
Authors: bruce@edmonds.name (Bruce Edmonds
Abstract: Review of: Escape from Model Land. How Mathematical Models Can Lead Us Astray and What We Can Do About It by Erica Thompson, reviewed by Bruce Edmonds
PubDate: Tue, 31 Jan 2023 12:51:00 +000
|
Open Access journal
