Thursday, December 16, 2021

Disaster through the Lens of Complex Adaptive Systems: Exploring Emergent Groups Utilizing Agent Based Modeling and Social Networks

For those interested in my PhD work, I'm posting the link to my dissertation in ProQuest: dissertation-link

In addition to the work I've published on the complex adaptive systems in disaster and how computational social science can contribute to disaster science and emergency management, my dissertation details the results of an agent-based model (ABM) of emergent groups in a disaster. The model shows how agent-based modeling can be linked with GIS and SNA and be used to measure the effects of emergent social networks. These emergent networks are a manifestation of the latent social capital that are available in times of crisis. When properly implemented such models can be used to both educate and inform researchers and policy-makers. 

As an example of the possible results in an ABM of social networks in disaster, the model shows that emergent group sizes follow exponential scaling patterns characteristic of complex adaptive systems. 

Number of Nodes and Edges of Emergent Groups in a Disaster Response


Dissertation Abstract:

Disasters have become more frequent and intense in the last decades and are a significant challenge to the health and well-being of local communities and regions. As a potential solution to this problem attention has been drawn to community resilience and the building of social networks that support or hinder local response and recovery. Research on disasters and community resilience has shown how the ability to leverage social capital through a community’s social networks is fundamental to the ability of individuals and communities to respond to disaster events, but there is little understanding of how the evolution of social networks can impact disaster response and recovery. A computational framework and agent-based model of disasters can provide a virtual laboratory for testing social network effects and uncover their role, function and underlying mechanisms in community resilience. Agent-based models are suited to test bottom-up dynamics and the interactions of variables that lead to the nonlinear relationships in disasters. To what extent can an agent-based model characterize the social networks that emerge in response to a no-warning disaster event such as a Nuclear Weapon of Mass Destruction impacting Manhattan Island? To explore this question this research reviews theories of disaster, primarily from sociological and anthropological research, and builds a conceptual model of disasters from which to develop an agent-based model. The agent-based model represents social networks relevant in both the normal commuting patterns of New York City and the emergent social networks responding to a Nuclear Weapon of Mass Destruction impacting Manhattan Island. Network representations of social groups along with physical representations of the community shows how individuals adapt and respond to the disaster in the initial response. Integrating agent-based models with social network analysis provides new spaces for scientific inquiry into disasters, the dynamics of social networks in resilient communities, and those areas of complexity most often explored today with qualitative methodologies. 


Tuesday, April 10, 2018

Research in Progress...A Presentation of "From Networks to Recovery"

The last number of months I have been working hard on a paper of my findings on theories of human behavior in disaster and developing the framework for a model of social networks in disaster. This period of focused work has allowed me to make incremental progress on my research, but left little time and few products to post on this blog. Last week was an exception, and I was able to take a break, attend a conference and present my work with fellow anthropologists. Today I'm sharing a presentation of my latest work on dropbox, presentation_link.

Conferences provide a welcome break to the isolation of dissertation work and a forum in which to challenge ideas. They are a tremendous opportunity to meet others in your discipline and fellow researchers working on similar topics. I was honored to have my abstract accepted and present the current results of my dissertation work at the Society for Applied Anthropology, https://www.sfaa.net/. They held their annual meeting in Philadelphia around the theme of sustainability and issues of migration and displacement, climate change, risk and disaster, and access to education, SfAA 2018 Annual Meeting Program.

The SfAA is a very active community applying anthropology in the field, and it was a privilege to hear from and talk to people working on similar research problems. In particular, there was a large body of ethnographic research on people's experiences during and recovering from hurricanes, including preliminary work on the 2017 Hurricanes Harvey and Irma in Texas and Puerto Rico. My presentation of an agent-based model of social network dynamics in hurricanes provided an interesting contrasting research method. Taken together the results of ethnographic and modeling methods illustrate how qualitative and quantitative methods can complement each other and improve understanding of human behavior in these events.


Key findings in my research presentation include:


1. A conceptual timeline of events through the hurricane cycle.




2. Supportive functions for social networks throughout the cycle




3. Network ties are prioritized and leveraged -- shifting ego networks

4. Agent interactions need to be accounted for in three spaces; physical, social and cognitive:



At this point in my dissertation I have the empirical data and framework for a simulated social experiment on social networks through a hurricane cycle. My primary focus will be on coding the model using Java and MASON. Up to now the challenge has been to collect and format empirical data for import into the model, but in the next few months I will be working to encode realistic human behavior and export network data formatted for social network analysis (SNA).

The full presentation "From Networks to Recovery" covers the foundational work for my dissertation and is also available here. Comments and feedback are most welcome.

Friday, November 10, 2017

Technical Solutions for Caregivers - innovation through interdisciplinary work

Last weekend with fellow students at George Mason University (GMU), I participated in the well-organized 2017 Caregiving for the Caregiver Hack-a-Thon sponsored by the Lindsay Institute, http://caregivinginnovations.org/caring-caregiver-hack. The event provided a good opportunity for student technology developers to work across disciplines on a shared goal. We also had the opportunity to speak directly with the technology users, the caregiver, and get input from experts in technology and business.

Although many of these hacked solutions will never get off the ground, some may eventually be developed and released for use by caregivers. The GMU team developed a mobile application, reashore, designed to address the informational and emotional needs of caregivers struggling with the practical daily needs of those with rarer forms of dementia. The work provides an example of how a good, interdisciplinary team can apply current technology and develop innovative solutions to modern problems.

Our team was privileged to work directly throughout the Hack-a-Thon with a caregiver whose husband suffered from dementia. In our discussions we identified a numerous problems that she struggled with while caring for her husband and reduced these to three needs:

1) A big problem is that, although there are many resources for people dealing with Alzheimers, there were few resources for people dealing with the other lesser known varieties of  dementia. Ultimately, she wanted one space online to which she could find solutions for the practical problems of getting the care recipient to eat or get dressed.
2) The best place to get answers and support was a caregiver support group, and she wanted the solution to include a means for caregivers with similar perspectives to be able to find and support each other.
3) The goals of the caregiver hack were also to develop a technology tool to help caregivers stay healthy, so we added the need for a place for the caregiver to find the means for relaxation.



To solve these three problems we created a prototype web application, reashore, that could be used on Android and iOS mobile platforms. The application has three main functions that correspond to the three identified needs:










1) A crowd-sourced search capacity that allows caregivers to find, suggest and rate solutions to their daily caregiving problems. As a recommender system, it integrates some of the functionality of well-known tools like Google search and StackOverflow.


Splash Screen


Solutions Search
Virtual Space to Find and Create Groups








2) A virtual meeting place for groups and individuals to find each other and set up meetings, similar to Meetup. This capability has a SOS feature that allows a caregiver to quickly find a fellow caregiver to chat with, when it gets really tough.




Virtual SOS to for Caregiver Chat























Breathing Room

Find Services
3) A virtual breathing room in which the caregiver can find resources to relax including recommended music, videos, and games and links to resources for meditation, yoga and other relaxation services.

Lastly, a recommender services search was added as a means to enable caregivers to find good services like elder care or nursing homes and to generate ad revenue for the application's maintenance.



A lot of work is needed to make this application truly usable (improve readability, add functionality, etc.), but it demonstrates the power of interdisciplinary work to create innovative solutions in very short timeframes. 

You can see our prototype and business plan presentation for more information: reashore

Thursday, October 19, 2017

Generating Realistic Populations of Megacities and their Social Networks

In my last blog I wrote how the study of disasters needs to account for three interacting complex adaptive systems, the physical environment, the social environment and the individual cognitive environment. Agent-based experimental simulations of populations responding to disasters need  their synthetic populations to represent not just the individual characteristics of people, but also the social connections that influence their behavior. Tomorrow I will present a paper on how to generate such a population at the Computational Social Science Society of the Americas in Santa Fe, New Mexico.

The paper, co-authored with my fellow researchers at George Mason University, discusses how a mixed method of iterative proportional fitting and network generation to build a synthesized subset population of the New York megacity and region. Our approach demonstrates that a robust population and social network relevant to specific human behavior can be synthesized for agent-based models.

Read our paper here:
Generation of Realistic Megacity Populations and Social Networks for Agent-Based Models

Or view the presentation:
Generating Megacity Populations and their Social Networks

The study area includes New York City and the larger megacity region.




The New York Megacity and region includes very dense areas and rural ones.



We are creating a 1:1 population based on the US 2010 Census, and we use Iterative Proportional Fitting (IPF) for the population synthesis. This is a method in which the attributes of an individual unit are taken from a data set with fixed marginal totals. For each agent in the model a random sample is taken from a probability distribution of the relevant attributes existing in the population data. This process is repeated until all the attributes are assigned to the agent population.



In the last stage of the process we create social networks based on family, work and school connections. If a network tie in larger workplace and school populations, we created a small-world network from the potential pool of network connections. The result are thousands of network connections that extend individual and family social ties beyond their households. This shows a family with two working parents and their two children.



My colleague, Talha Oz, has shared the code on Jupyter's nbviewer:



Friday, October 13, 2017

Organizing Theories for Disaster Study in Computational Social Science

What are frameworks for understanding disasters? And, how can a Complex Adaptive Systems framework provide better understanding?



Check out my presentation at GMU's Computational Social Science (CSS) Friday Seminar:
Organizing Theories for Disaster Study in Computational Social Science

Or view the slides:
Disasters in a Complex Adaptive Systems Framework


Friday, September 22, 2017

Synthetic Populations for ABMs

Agent-based models are being used for computer experiments in epidemiology, transportation, migration, climate change, and urban studies. Researchers use the models to experiment on simulated human behavior with a synthesized population in a controlled environment. What population synthesis methods are currently being used in ABMs, and how have these synthetic populations been used?

Population synthesis is the process of creating agent representations of the model population based on available data. Sample-based methods are more traditional, but new methods also create synthetic populations sample-free.

Sample-based
Sample-based methods either involve synthetic reconstruction or combinatorial optimization (reweighting) based on existing datasets on population characteristics such as census data.

In synthetic reconstruction the joint-distribution of relevant population attributes are used to create a fitted population and generate individual units on that population. The most common method is Iterative Proportional Fitting (IPF). A procedure in which the attributes of an individual unit are taken from a contingency table with fixed marginal totals. For each agent in the model a random sample is taken from a probability distribution of the relevant attributes existing in the population. This process is repeated until all the attributes are assigned to the agent population. The result is a reconstruction of the original population.

In combinatorial optimization a sample population is generated and then repeatedly modified until it meets a threshold of required constraints. First a set of randomly selected households are taken from an existing population dataset. A random household from this sample and one from the large dataset is assessed for fit. If there is a better fit, the households are switched. The assessment and potential switch is repeated, and the sample population gradually improves its fit to a set of population constraints. The result is a sample population

Sample-free
A sample-free method in synthetic reconstruction involves generating individual units and placing them into households or other groupings until the entire population is used. The method draws the individual's attributes at the most disaggregated level from joint distributions. After all the individuals are generated, the population is compared to the joint distributions and inconsistencies are handled by shifting attribute values. The last step is to gather the individuals into households or groupings. Look for this technique to be applied to migratory groups, areas undergoing rapid change and other underrepresented, marginalized populations.


Populations and Generator Tools:
Population generative tools are available now with existing synthetic populations or for use on new population data.

PopGen
Uses Iterative Proportional Updating (IPU) that, unlike IPF, controls for both the agent and agent grouping at the same time. Used for creating realistic human populations for prediction of anatomical, physiological and phase 1 metabolic variation in the population in response to exposure or dosage.

PopGen for SimTRAVEL
Adds person-level attributes in addition to census data distributions in population synthesis. These populations are designed for application to urban planning and analysis of transportation, routes, activities, vehicles, emissions and land-use. Arizona State University (ASU) is integrating it into UrbanSim.
http://urbanmodel.asu.edu/popgen.html

Synthetic Populations and Ecosystems of the World (SPEW)
Provides a synthetic population and ecosystem from available data of over 80 countries in American Community Survey (ACS), International Public Use Microdata Samples (IPUMS) and other population samples using simple random sampling. Carnegie Mellon University (CMU) plans to add moment matching and iterative proportional fitting in future versions. Moment matching is a statistical technique used to estimate population parameters by deriving equations that describe the population characteristic's expected mean.
http://www.stat.cmu.edu/~spew/resources/

Virginia Bioinformatics Institute Synthetic Data 
Synthetic populations of Portland, Oregon, Montgomery, Virginia, West Africa, and Washington, D.C. that have been applied to studies of infectious disease, incarceration rates, and emergency management.
http://ndssl.vbi.vt.edu/synthetic-data/

RTI U.S. Synthetic Household Population
Provides a representation of households and persons in U.S. populations from publicly available data sources. These data are placed on a map and represent distribution variations within census blocks. Used for representations of the demographic characteristics of a population including age, gender, race, income, and educational attainment. The map and underlying data is available online for free to track infectious disease, study transportation networks or optimize supply chains.

The data for SPEW, RTI US Synthetic Household and other populations from the Models of Infectious Disease Study (MIDAS) can also be found here:


Further References:
Arentze, Theo, Harry Timmermans, and Frank Hofman. 2007. “Creating Synthetic Household Populations: Problems and Approach.” Transportation Research Record: Journal of the Transportation Research Board 2014 (December): 85–91.

Barthelemy, Johan, and Philippe L. Toint. 2013. “Synthetic Population Generation Without a Sample.” Transportation Science 47 (2): 266–79. doi:10.1287/trsc.1120.0408.

Beckman, Richard J., Keith A. Baggerly, and Michael D. McKay. 1996. “Creating Synthetic Baseline Populations.” Transportation Research Part A: Policy and Practice 30 (6): 415–29.

Deming, W. Edwards, and Frederick F. Stephan. 1940. “On a Least Squares Adjustment of a Sampled Frequency Table When the Expected Marginal Totals Are Known.” The Annals of Mathematical Statistics 11 (4): 427–444.

Huang, Zengyi, and Paul Williamson. 2001. “A Comparison of Synthetic Reconstruction and Combinatorial Optimisation Approaches to the Creation of Small-Area Microdata.” Department of Geography, University of Liverpool.

McNally, Kevin, Richard Cotton, Alex Hogg, and George Loizou. 2014. “PopGen: A Virtual Human Population Generator.” Toxicology 315 (January): 70–85.

Müller, Kirill, and Kay W. Axhausen. 2010. “Population Synthesis for Microsimulation: State of the Art.” In . Monte Verità, Ascona, Switzerland: ETH Zürich, Institut für Verkehrsplanung, Transporttechnik, Strassen-und Eisenbahnbau (IVT).

Williamson, Peter, Michael Birkin, and Phillip H. Rees. 1998. “The Estimation of Population Microdata by Using Data from Small Area Statistics and Samples of Anonymised Records.” Environment and Planning A 30 (5): 785–816.

Wise, Sarah. 2014. “Using Social Media Content to Inform Agent-Based Models for Humanitarian Crisis Response.” George Mason University. http://digilib.gmu.edu/xmlui/handle/1920/8879.

Tuesday, September 12, 2017

Hurricane Response & Crisis Mapping

Crowdsourcing sharable maps for humanitarian relief efforts are becoming the new normal.

Three examples:
Harvey Relief http://harveyrelief.handiworks.co/relief-map
Irma Response http://google.org/crisismap/2017-irma
US Wildfires http://google.org/crisismap/us-wildfires


These maps include layers of geospatial data as KML files that can be downloaded for analysis. The Harvey Relief map includes crowd-sourced locational data for individuals to find aid.
The google crisis map layers include links to the original source KML data.



These maps can be used to inform humanitarian efforts around the globe.