Thursday, September 07, 2023

Agent-Based Modeling of Consumer Choice

At the upcoming International Conference on Geographic Information Science (GIScience 2023) Boyu Wang and myself have a new paper entitled "Agent-Based Modeling of Consumer Choice by Utilizing Crowdsourced Data and Deep Learning." In the paper we explore how through mining Yelp reviews can inform an agents choices of restaurants. The model itself was created in Mesa and uses Mesa-Geo and  more details about the model can be found at https://github.com/wang-boyu/yelp-abm.  If this sounds of interest, below you can see the abstract to the paper, some fugues including the graphical user interface of the model and a link to the paper.

Abstract: People’s opinions are one of the defining factors that turn spaces into meaningful places. Online platforms such as Yelp allow users to publish their reviews on businesses. To understand reviewers' opinion formation processes and the emergent patterns of published opinions, we utilize natural language processing (NLP) techniques especially that of aspect-based sentiment analysis methods (a deep learning approach) on a geographically explicit Yelp dataset to extract and categorize reviewers' opinion aspects on places within urban areas. Such data is then used as a basis to inform an agent-based model, where consumers' (i.e., agents') choices are based on their characteristics and preferences. The results show the emergent patterns of reviewers' opinions and the influence of these opinions on others. As such this work demonstrates how using deep learning techniques on geospatial data can help advance our understanding of place and cities more generally.


Keywords: aspect-category sentiment analysis, consumer choice, agent-based modeling, online restaurant reviews.

An overview of proposed agent-based model logic.

Average star rating vs. average sentiment by aspect category for 200 randomly selected restaurants in the City of St. Louis, MO.

The prototype agent-based model (a) with simulated (b) and actual visiting patterns (c).

Full reference:

Wang, B. and Crooks, A.T. (2023), Agent-Based Modeling of Consumer Choice by Utilizing Crowdsourced Data and Deep Learning, in Beecham, R., Long, J.A., Smith, D., Zhao, Q., and Wise, S (eds), Proceedings of the 12th International Conference on Geographic Information Science (GIScience 2023), Dagstuhl Publishing, Dagstuhl, Germany., pp. 81:1-81:6. (pdf)


Wednesday, August 30, 2023

ABM Online Courses

Often I get asked about how to learn about agent-based modeling (ABM). While we have a book on this with respect to GIS and ABM, the other day, Jiaqi Ge posted a question about free ABM online courses on the SIMSOC mailing list and I though it would be worth summarizing the responses here as the resources are quite useful.

Jiaqi shared some really good resources like the Santa Fe Institutes "Introduction to Agent-Based Modeling" and "Fundamentals of NetLogo" along with the University of Geneva's Coursera course "Simulation and modeling of natural processes". 
 
Others also responded to the question. For example, Wander Jager responded with online modules developed from the Action for Computational Thinking in Social Sciences (ACTiSS) team. Jen Badham responded with an extended tutorial about model design and creating models in Netlogo while Dino Carpentras responded with several general videos on YouTube on ABM which he has created. Hopefully readers will find these useful and also you might want to see our Github pages on GIS and ABM

Wednesday, June 28, 2023

Editorial: Urban analytical approaches to combating the Covid-19 pandemic

While there has been a lot written about COVID-19 Angela  Yao, Bin Jiang, Jukka Krisp, Xintao Liu, and Haosheng Huang and myself have just recently wrapped up a special issue in Environment and Planning B and how it can be studied through the lens of urban analytics.  After a call for papers for the special issue, we published 10 papers that cover a wide spectrum of analytical methods have been used to study the pandemic. These ranged from how policies impacted pedestrian patterns to how data could on the disease could be visualized along with many things in between. Below you can see papers:

Accompanying these papers is an editorial entitled "An overview of urban analytical approaches to combating the Covid-19 pandemic," In this editorial we situate these papers in the larger literature of urban analytics and Covid-19. Also in the editorial, we explore what can be learned from the current research on Covid-19 and finally we identify gaps and future research opportunities for urban analytics in combating epidemic outbreaks.

A framework of the Covid-19 pandemic dynamics in urban systems.


Covid-19 research themes and topics through the lens of geography and urban analytics.

Full Reference:

Yao, X.A, Crooks, A.T., Jiang, B., Krisp, J., Liu, X. and Huang, H. (2023), An overview of urban analytical approaches to combating the Covid-19 pandemic, Environment and Planning B, 50 (5), pp. 1133–1143. (pdf)


Saturday, May 20, 2023

Simulation & Optimization Techniques for the Mitigation of Disruptions to Supply Chains

Our last paper at the Annual Modeling and Simulation Conference (ANNSIM) is entitled "Simulation and Optimization Techniques for the Mitigation of Disruptions to Supply Chains" where we (Raj Patel, Abhisekh Rana, Sean Luke, Carlotta Domeniconi, Hamdi Kavak, Jim Jones and myself) build upon our previous work which explored how the actions of criminal networks and agents might impact supply chains. 
 
This paper extends this research to incorporate both disruption and mitigation modeling into the same simulation.  By using evolutionary computation optimization techniques (e.g., Covariance Matrix Adaptation Evolution Strategy) we demonstrate how we can optimize both the disruption and mitigation scenarios in a pharmaceutical supply chain (which we call PharmaSIM). Our results demonstrate how  evolutionary computation techniques could be used to not only identify worst-case disruption scenarios but to also optimize the allocation of the mitigations to counter their effects.  If this sounds of interest, below we provide the abstract to the paper, some of the figures we use to support our discussion and results. While at the bottom of the post we provide the full reference to the paper along with a link to a preprint of it. 

Abstract
The COVID-19 pandemic has clearly highlighted the importance of supply chains to the function of the world economy. Moreover, the global nature of most modern supply chains along with their complexity has left them vulnerable to a wide-ranging set of disruptive scenarios. This increase in complexity has also led to a corresponding increase in disruptions to supply chains from criminal networks. In this paper, we demonstrate how a generic pharmaceutical supply chain network can be successfully modeled using discrete event simulation. We outline how disruptions by criminal networks and mitigation strategies to counter them can be effectively incorporated into the same model. Finally, we show how optimization techniques, such as evolutionary computation, can be used to not only identify worst-case disruptions and find mitigations for them, but also be used to identify mitigation strategies that are effective against a diverse set of damaging disruption scenarios.

Keywords: Simulation, Optimization, Supply Chains, Disruptions, Mitigation. 


Topology of the generic pharmaceutical supply chain (PharmaSIM) model.

Fitness after evolutionary optimization of attack configurations and corresponding safety stock allocation for different budgets.

Fitness by generation for the coevolution of attack vectors and mitigation configurations.

Full Reference:

Rana, R., Patel, R., Luke, S., Domeniconi, C., Kavak, H., Jones, J. and Crooks, A.T.  (2023), Simulation And Optimization Techniques for the Mitigation of Disruptions to Supply Chains, The Annual Modeling and Simulation Conference (ANNSIM), Hamilton, ON. (pdf)

Spiral Software Development Process for ABM

Readers of this blog might gather that we are constantly developing agent-based models to study and better understand a wide range of problems but unlike in say the software industry, agent-based model development is rather ad hoc in terms of a standardized software development process. To this end  Maxim Malikov, Fahad  Aloraini, Hamdi Kavak and William Kennedy from George Mason University and myself  have a paper entitled "Developing a Large-Scale Agent-Based Model Using the Spiral Software Development Process" which we will be presenting at the upcoming Annual Modeling and Simulation Conference (ANNSIM).

In the paper, we review the unique aspects of agent-based models and discuss the challenges faced in the development of our own large-scale agent-based model, which simulates the impact of a disaster on the infrastructure and the population of a city. This project combines the expertise of teams with multiple disciplines, and therefore must be able to adjust to novel input from these teams over the life of the project.  Furthermore, we describe our solution to these challenges in the form of a variation of the Spiral model of software development and the ways this approach helped us address the exploratory nature of agent-based modeling.  

If this sounds of interest, below we provide the abstract to the paper, some of the figures we use to support our discussion. At the bottom of the post we provide the full reference to the paper along with a link to a preprint of it. 

Abstract:

As the level of complexity of agent-based models grows, so does the complexity of their development. At the time of writing, the discipline of agent-based modeling does not have an established standard for the software development process to support this increasing complexity. We hope to address this need by introducing our variation of the Spiral model of software development and demonstrating an application of this process through a simple use case. We argue that the Spiral model of software development is a flexible approach that can be tailored to fit the needs of almost any project type. Further, our agent-based modeling variation of the Spiral model is an effective approach that is capable of guiding and supporting large interdisciplinary teams participating in a project, while providing sufficient flexibility to account for the uncertainty in the requirements that may arise during the development period.

Keywords: Software development, Agent-based Modeling, Spiral Development, Disaster.

 

Spiral model with adjustments made to account for the specifics of complex agent-based models. Adopted from Boehm (2000).


Prototype 1 of the city infrastructure simulation. This graphical user interface shows agent and infrastructure changes after a disaster.

Full reference:

Malikov, M., Aloraini, F., Crooks, A.T., Kavak, H. and Kennedy, W.G. (2023), Developing a Large-Scale Agent-Based Model Using the Spiral Software Development Process, The Annual Modeling and Simulation Conference (ANNSIM), Hamilton, ON. (pdf)