Thursday, January 08, 2015

Crowdsourcing Urban Form and Function

We have just had published a new paper entitled: "Crowdsourcing Urban Form and Function" in International Journal of Geographical Information Science which showcases some of our recent work with respect to cities and how new sources of information can be used to study urban morphology at a variety of spatial and temporal scales. Below is the abstract for the paper: 

"Urban form and function have been studied extensively in urban planning and geographic information science. However, gaining a greater understanding of how they merge to define the urban morphology remains a substantial scientific challenge. Towards this goal, this paper addresses the opportunities presented by the emergence of crowdsourced data to gain novel insights into form and function in urban spaces. We are focusing in particular on information harvested from social media and other open-source and volunteered datasets (e.g. trajectory and OpenStreetMap data). These data provide a first-hand account of form and function from the people who define urban space through their activities. This novel bottom-up approach to study these concepts complements traditional urban studies work to provide a new lens for studying urban activity. By synthesizing recent advancements in the analysis of open-source data we provide a new typology for characterizing the role of crowdsourcing in the study of urban morphology. We illustrate this new perspective by showing how social media, trajectory, and traffic data can be analyzed to capture the evolving nature of a city’s form and function. While these crowd contributions may be explicit or implicit in nature, they are giving rise to an emerging research agenda for monitoring, analyzing and modeling form and function for urban design and analysis."
This paper builds and extends considerably our prior work, with respect to crowdsourcing, volunteered and ambient geographic information. In the scope of this paper we use the term ‘urban form’ to refer to the aggregate of the physical shape of the city, its buildings, streets, and all other elements that make up the urban space. In essence, the geometry of the city. In contrast, we use the term ‘urban function’ to refer to the activities that are taking place within this space. To this end we contrast how crowdsourced data can related to more traditional sources of such information both explicitly and implicitly as shown in the table below. 

A typology of implicit and explicit form and function content

In addition, we also discuss in the paper how these new sources of data, which are often at finer resolutions than more authoritative data are allowing us to to customize the we we aggregate the data  at various geographical levels as shown below. Such aggregations can range from building footprints and addresses to street blocks (e.g. for density analysis), or street networks (e.g. for accessibility analysis). For large-scale urban analysis we can revert to the use of zonal geographies or grid systems.  
Aggregation methods for varied scales of built environment analysis

In the application section of the paper we highlight how we can extract implicit form and function from crowdsourced data. The image below for example, shows how we can take information from Twitter, and differentiate different neighborhoods over space and time.

Neighborhood map and topic modeling results showing the mixture of social functions in each area.

Finally in the paper, we outline an emerging research agenda related to the "persistent urban morphology concept" as shown below. Specifically how crowdsourcing is changing how we collect, analyze and model urban morphology. Moreover, how this new paradigm provides a new lens for studying the conceptualization of how cities operate, at much finer temporal, spatial, and social scales than we had been able to study so far.

The persistent urban morphology concept.

We hope you enjoy the paper.

Full Reference:  
Crooks, A.T., Pfoser, D., Jenkins, A., Croitoru, A., Stefanidis, A., Smith, D. A., Karagiorgou, S., Efentakis, A. and Lamprianidis, G. (2015), Crowdsourcing Urban Form and Function, International Journal of Geographical Information Science. DOI: 10.1080/13658816.2014.977905 (pdf)

Wednesday, December 17, 2014

Example Models from CSS600

Even after several years of teaching it is always amazing how quickly a semester passes. One of the courses I taught this semester was CSS 600: Introduction to Computational Social Science. This is often the first CSS class many students take here at George Mason University. We discuss a number of computational approaches which are used for social science research, coupled to  complexity theory. As an introduction to the subject, the course has the following objectives:
  1. To understand the motivation for the use of computational models in social science theory and research, including some historical aspects (Why conduct computational research in the social sciences?).
  2. To learn about the variety of CSS research programs across the social science disciplines, through a survey of social simulation models (What has CSS accomplished thus far?).
  3. To understand the distinct contribution that CSS can make by providing specific insights about society, social phenomena at multiple scales, and the nature of social complexity (What is the relation between computational social science.
  4. To provide foundations for more advanced work in subsequent courses or projects for those students who already have or will develop a long-term interest in CSS.
Part of the students final grade comes from the development of a computational model in an area of their interest (e.g., microeconomics, international relations, environmental policy, economic development, historical dynamics, finance etc..). Often, this is the first compuational model that the students have ever developed. Below you can see a number of models developed using NetLogo as part of the class.

To find out more about our program see:

Friday, November 28, 2014

Linking Cyber and Physical Spaces

We have just published a new paper in  Computers, Environment and Urban Systems entitled "Linking Cyber and Physical Spaces Through Community Detection And Clustering in Social Media Feeds". In the paper we explore how geosocial media is providing us with  a new social communication avenue and a novel source of geosocial information. 

In particular, we discuss the notion of physical presence within social media and its importance for exploring the relation between the cyber and the physical domains. We discuss how communities and groups can be detected in both the cyber and physical space, and how they can be processed to form a ‘hybrid’ geosocial view of communities using social network analysis, community detection (the Louvain method) and DenStream. To showcase these concepts and their benefits, we present the analysis of two case studies that make use of Twitter data associated with two different types of events: a planned activity during the Occupy Wall Street (OWS) Day of Action (November 17th, 2011), and the response to the Boston Marathon Bombing (April 15, 2013). We conclude with a summary and outlook. Below is the abstract of the paper:
Over the last decade we have witnessed a significant growth in the use of social media. Interactions within their context lead to the establishment of groups that function at the intersection of the physical and cyber spaces, and as such represent hybrid communities. Gaining a better understanding of how information flows in these hybrid communities is a substantial scientific challenge with significant implications on our ability to better harness crowd-contributed content. This paper addresses this challenge by studying how information propagates and evolves over time at the intersection of the physical and cyber spaces. By analyzing the spatial footprint, social network structure, and content in both physical and cyber spaces we advance our understanding of the information propagation mechanisms in social media. The utility of this approach is demonstrated in two real-world case studies, the first reflecting a planned event (the Occupy Wall Street – OWS – movement’s Day of Action in November 2011), and the second reflecting an unexpected disaster (the Boston Marathon bombing in April 2013). Our findings highlight the intricate nature of the propagation and evolution of information both within and across cyber and physical spaces, as well as the role of hybrid networks in the exchange of information between these spaces.

Research highlights include:
    • Our analysis includes two major events as captured in Twitter.
    • The themes in cyber and physical communities tend to converge over time.
    • Messages among physical space users are more consistent at the onset of the event.
    • Geolocated users are consuming information more than they produce.

      Below are some of the images from the paper. Specifically the first image is how one can think of the relationships between physical and cyber spaces.  The next image provides an overview Our geosocial analysis framework for examining cyber and physical communities.

      Our Geosocial analysis framework

      In the figure below we show an example of using DenStream for spatiotemporal clustering and how the process can capture the protest activities that were planned for the Occupy Wall Street movement’s Day of Action. Each dot corresponds to the originating location of a geolocated tweet; The color of each point indicates the time of the corresponding tweet, ranging from dark blue (early morning, 0) to dark red (late night, 1). While the circles represent a specific spatiotemporal cluster. For example the circle labeled A marked the start of the day where people congregated around Wall Street while circle labeled C shows a cluster at Foley Square.
      Physical space groups identified in the lower Manhattan area. Each dot corresponds to the originating location of a geolocated tweet; The color of each point indicates the time of the corresponding tweet, ranging from dark blue (early morning, 0) to dark red (late night, 1).

      While in the figure below we show one example of linking the cyber and physical communities. Specifically in (a), the top five communities (node degree > 100) in the cyber space retweet network (each community is designated by one color) are shown; (b) shows the physical space groups; and (c) shows the resulting  hybrid meta-network where the connections between physical groups (P nodes), and cyber space communities (C nodes) are shown.

      We hope you enjoy the paper.

      Full Reference:
      Croitoru, A., Wayant, N., Crooks, A.T., Radzikowski, J. and Stefanidis, A. (2014), Linking Cyber and Physical Spaces Through Community Detection And Clustering in Social Media Feeds, Computers, Environment and Urban Systemsdoi:10.1016/j.compenvurbsys.2014.11.002

      Wednesday, October 29, 2014

      Happy Halloween: Zombie Agent-Based model

      In preparation for Halloween,  last night in class we explored a simple agent-based model of zombie attack. At each time step, a human moves to a nearby unoccupied space, and a zombie moves to the nearest human. If a zombie and an uninoculated human occupy the same space, a fierce battle ensues, in which the probability that the human will kill the zombie is pkH-z, and the probability that the zombie kills the human and converts them to their horrific undead form is pkZ-h. 

      Zombies, however, are not attracted to inoculated humans and ignore them. If recovery? is enabled, then there is a chance (given by recoveryRate) that a zombified person will see the errors of their cannibalistic ways and return to human form. All these factors working together provide some interesting population dynamics, illustrated by the “Totals” population count plot on the screen. 

      The model is programed in NetLogo and also demonstrates how a simple as .asc file can be used for the spatial envirment for the zombies and humans to interact within.

      You can download the code from here.

      Tuesday, September 23, 2014

      Geosimulation and Big Data: A Marriage made in Heaven or Hell?
      Call for papers: AAG 2015 – Geosimulation and Big Data: A Marriage made in Heaven or Hell?

      In recent years, human emotions, intentions, moods and behaviors have been digitized to an extent previously unimagined in the social sciences. This has been in the main due to the rise of a vast array of new data, termed ‘Big Data’. These new forms of data have the potential to reshape the future directions of social science research, in particular the methods that scientists use to model and simulate spatially explicit social systems. Given the novelty of this potential “revolution” and the surprising lack of reliable behavioural insight to arise from Big Data research, it is an opportune time to assess the progress that has been made and consider the future directions of socio-spatial modelling in a world that is becoming increasingly well described by Big Data sources.

      We invite methodological, theoretical and empirical papers that that engage with any aspect of geospatial modelling and the use of Big Data. We are particularly interested in the ways that insight into individual or group behavior can be elucidated from new data sources – including social media contributions, volunteered geographical information, mobile telephone transactions, individually-sensed data, crowd-sourced information, etc. – and used to improve models or simulations. Topics include, but are not limited to:
      • Using Big Data to inform individual–based models of geographical systems;
      • Translating Big Data into agent rules;
      • Elucidating behavioral information from diverse data;
      • Improving simulated agent behavior;
      • Validating agent-based models (ABM) with Big Data;
      • Ethics of data collected en masse and their use in simulation.
      Please e-mail the abstract and key words with your expression of intent to Nick Malleson by 28th October, 2014. Please make sure that your abstract conforms to the AAG guidelines in relation to title, word limit and key words and as specified at An abstract should be no more than 250 words that describe the presentation’s purpose, methods, and conclusions as well as to include keywords.

      • Alison Heppenstall, School of Geography, University of Leeds
      • Nick Malleson, School of Geography, University of Leeds
      • Andrew Crooks, Department of Computational Social Science, George Mason University
      • Paul Torrens, Department of Geographical Sciences, University of Maryland
      • Ed Manley, Centre for Advanced Spatial Analysis, University College London
      • 28th October, 2014: Abstract submission deadline. E-mail Nick Malleson by this date if you are interested in being in this session. Please submit an abstract and key words with your expression of intent.
      • 31st October, 2014: Session finalization and author notification
      • 3rd November, 2014: Final abstract submission to AAG, via All participants must register individually via this site. Upon registration you will be given a participant number (PIN). Send the PIN and a copy of your final abstract to Nick Malleson. Neither the organizers nor the AAG will edit the abstracts.
      • 5th November, 2014: AAG registration deadline. Sessions submitted to AAG for approval.