Blog > International Conference On E-social Science 2009-06-24 To 2009-06-26

International Conference on e-Social Science 2009-06-24 to 2009-06-26

Contents

  • Introduction
  • People
  • Notes from 2009-06-26
  • Notes from 2009-06-25
  • Notes from 2009-06-24
  • References

Introduction

People

Notes from 2009-06-26

Closing session

  • Peter Halfpenny gave thanks to all the organisers and hosts and funders and participants and reminded us to
feedback by filling out the evaluation form…

Session 3d: Tools for understanding individual behaviour

  • A Data Warehouse approach to Public Safety Monitoring
    • Diederick de Vries
  • Designing authoring tools for the creation of online behavioural interventions
    • Adrian Osmond
  • Good session to end with and well chaired by Meik Poschen :-)

Session 3c: Ethics and e-Research

  • Trying to see the bigger picture: a review of the opportunities for linking eHealth and social Science datasets to enhance understanding of risk in suicide in Scotland

Chat with Pascal Ekin and Karim Chine about the state of the world and science funding

  • What is wrong with the system?
  • How do we make it better?
  • Do funders know what is best to invest in?
  • Do funders focus too much on reports and are they in danger of creating IT infrastructure that does not get used?
  • Do we need them to worry most about data preservation and making sure data is available in case there is some emergency need…
    • How do we get them to think about this kind of thing?
  • Why don't Pascal and Karim post their thoughts on a blog?
    • Are they worried this will haunt them?
    • Do they think it won't make a difference
    • Anyway, I value their views and the chat we had :-)

Chat with Pascal Ekin

  • I tried to explain what I am trying to do with modelling the world with a social simulation focus with agents representing individual people…
    • Pascal said that he knew of some work at MIT trying to do this back in the late 1980's and that he would dig up some references for me.
    • Pascal also suggested trying to run things in reverse.
  • We discussed tipping points and computer games and how to parallelise the model…
  • I wonder if Pascal is interested enough and can find the time to collaborate on this specifically…

Keynote 3: Networking Knowledge

Introduction and Thanks

Notes from 2009-06-25

Chat with Pascal Ekin

  • Pascal pointed me to the following as a good resource that explains Shibboleth security:
  • I can see that it can be slow given all the very many http requests made
    • Pascal was mentioning that Shibboleth could be slow causing time out issues for things like database requests.

Session 1c: e-Infrastructure Developments II

  • Spatial Data e-Infrastructure
    • Richard Sinnott
    • I was there to help with questions, but there weren't any…
  • Secure Data Service: specifications and challenges of using potentially disclosive data
    • Reza Afkhami
  • Internet proxy technology for securing a satellite image service (SIDS)
    • Pascal Ekin
    • GEMEDA was perhaps the first grid application in the UK
    • GEMS
    • Statistical classification of Urban/Rural areas using WPS

Chat with Yukio Maeda

Chat with David Fergusson

  • David was wondering what data about people could be captured from an image based remote sensing or something like that.
  • I think the answer is that there is not much really that is best captured in this way, counting the number of people in an area using remote sensing will require very high resolution data and remote sensing will find it very difficult to see through buildings and the like.

Meeting with Edwin Horlings

  • Eric Meyer introduced us when following up from our chat from yesterday.
  • I showed Edwin CASOS
  • He had seen the web pages…
  • I described what Social Simulation work I am doing, that the basis is modelling individual people as agents and described the Social Simulation Tutorial yesterday
  • …

Agreed to write 4 pages on Social Simulation for Elizabeth Vander Meer

  • Social simulation is a term given to a number of things. It is used in studies of individual insects colonies, and for made up societies that are not supposed to represent reality per se. The social simulation of interested here, is that which has base level agents that represent people. These people agents interact with each other and the environment and these interactions are modelled explicitly in a somewhat stochastic fashion.
  • Social simulation modelling becomes complex quickly and is perhaps best built in small incremental steps. At the base of many models is birth and death for which mortality and fertility rates are used. A gender distinction between males and females is also normal from the outset and each gender can have specified different age specific mortality and fertility rates.A simple model with this basis may be run to create a simulation. The simulation is run for a specified period of time at a specified time step (temporal resolution or tick). The output from the simulation is either some graphics such as a set of age-gender plot images, or the data from which such things can be generated. A seed or start population can be completely made up, or it can be based on age and gender distributions as recorded in some survey data.
  • A basic model can be extended in many ways by adding extra attributes, but as geographers, key is to explicitly model things spatially with two or three spatial dimensions. So, individual person agents need locating within the environment and are given the ability to move around in it and perceive all or part of it.
  • The modelling is stochastic in that it is based on pseudo random sequences, but these can be set with random seeds and implementations can be designed such that results are replicable based on some small amount of configuration (passing in of parameters). The models can be run a number of times with different random seeds in what is sometimes called an ensemble, producing usually a range of results that can summarised. Both trends and extreme results are especially interesting, but the variation in simulations is generally ineresting. The less variation there is, the more stable the model is and the less uncertain results of subsequent runs is.
  • Additionally sensitivity of results to different mortality and fertility rates can be studied. Running ensembles that vary these type of inputs and analysing the results is known as a parameter study.
  • In a spatial model mortality and fertility rates can be allowed to vary spatially and it can be that fertility may be specified to be dependent on males and females being co-located. Whether this is appropriate may depend on the spatial and temporal resolution and the nature of society being modelled. Is the nature of the society such that the only ways for someone to get pregnant is by being female and having sex with a male?
  • …

Keynote 2 : CIM-Earth: A Community Integrated Model of Energy and Resources Trajectories for Humankind

  • Ian Foster
  • Right up my street as a computational geographer trying to develop a model of the world based on social simulation :-)
  • external link: http://arete.uchicago.edu/projects/CIM_EARTH.shtml
  • external link: http://cimearth.org/
  • Looking for European collaborators
  • Feedback
    • I made a comment to make Ian aware that in this we are trying to do similar things based on social simulation and modelling individual people as agents and that I am keen to collaborate.
      • I highlighted the Social Simulation Tutorial which Ian had already spotted

Session 1b: Collaboratories

Paper Session 1 a: e-Infrastructure developments 1

  • Room: Maternussaal
  • Chair: Andy Turner
  • Introduction part 1
    • Welcome to this first session on e-Infrastructure developments which focuses on security with two presentations:
      1. Details experience of implementing security solutions for supporting access to data.
      2. About architectural design patterns for achieving end to end secuity in workflows.
    • Professor Richard Sinnott and Sardar Hussein from the UK National Centre for e-Science based at the University of Glasgow. are going to present starting with Rich who is an expert on security in e-Science and has experience of a wide range of application domains.
  • Feedback
    • Requested feedback via identi.ca and Twitter
  • Timing, we started at 10:05…
  • Supporting Security-oriented Interdisciplinary Research: Crossing the Social, Clinical and Geospatial Domains
    • Richard Sinnott
    • My default question:
      • What takes you the most time now in general in putting security in place for an application? Is it:
        • Finding out what is wanted (specifying the VO)…
        • Gaining the necessary trust…
        • Designing/Choosing/Implementing the solution…
    • Good clear presentation
    • Question from Junaid Arshad about usability
      • Basically Rich's reply is that for the user it is designed to be as easy as it can be
        • The go to their Virtual Organisation which knows what they need and they authenitcate to all of this via a login to their host institution.
          • What could be more simple :-)
  • Introduction 2
    • Next up is Sardar is specialising in security in workflows and is a PhD student in Rich's group.
  • Argh!!!! Sardar's laptop did not play and it took us 10 minutes to switch!!!!
  • Architectural design patterns for security oriented workflows in the social science domain
    • Sardar Hussain
    • Another good clear presentation and thankfully Sardar managed to catch us up some time, but we still finished nearly 10 minutes late.
    • Hopefully anyone with questions or feedback to Sardar will address these during the conference or soon…

Keynote 1: e-Infrastructures and e-Social Science

  • Kostas Glinos
  • Feedback
    • Ralph Schroeder asked for more details about evaluating impacts
    • Kostas would like these measured more quantitatively in terms of socio-economic and environmental metric and concedes that currently the evaluation is qualitative and that this is something that is important and work is on-going to make improvements…
    • Karime Chine asked some probing questions :-)

Welcome

Notes from 2009-06-24

  • Asked to and agreed to chair Paper Session 1 a tomorrow from 10:00 to 11:00am
    • Read the first paper:
      • Supporting Security-oriented Interdisciplinary Research: Crossing the Social, Clinical and Geospatial Domains
        • Rich Sinnott et al.
        • Thinking about relevant questions and how to introduce and comment…
    • Considering how to use IT to get feedback…
    • Key to chairing the sessions well is:
      • to make every effort to ensure the presentations run to time
      • to give a good introduction and summary
      • thank the speakers, attendance and organisers of the session

Chat with Mustafizur Rahman

  • He is interested to work with me to model with the Transport for London Data
    • This is very interesting, in particular I am interested in the difference between supply and demand and measuring spare capacity and capability of the public transport network…

Chat with Eric Meyer

  • There is a meeting later in the year with groups specialising in the history of science and modelling social systems
  • I mentioned the following to Eric, but could not at the time remember its name:
  • I am interested in the meeting and it may provide an opportunity for me to extend GENESIS work…

Social Simulation Tutorial

Preparation

Travel Information

  • I am booked into the conti hotel from 2009-06-23 to 2009-06-27
  • It figure it is best to get a train and walk to and from Cologne and the airport.
  • Airport Information
  • I am booked on flights:
    • 2009-06-27
      • KLM 1551 Amsterdam to Leeds departs 21:55 to 21:30
      • KLM 1812 Cologne to Amsterdam departs 18:45 to 19:55
    • 2009-06-23
      • KLM 1813 Amsterdam to Cologne departs 20:50 to 21:55
      • KLM 1548 Leeds to Amsterdam departs 17:25 to 19:45

References