h2 {link:4th International Conference on e-Social Science|http://www.ncess.ac.uk/events/conference}

h3 People
* Delegates
** Alphabetical by Surname, Firstname
*** ...
* Collaborators
** {link:People 2008-06-20|http://portal.ncess.ac.uk/access/wiki/site/~a.g.d.turner@leeds.ac.uk/2008-06-20.html#people}
** {link:People 2008-06-19|http://portal.ncess.ac.uk/access/wiki/site/~a.g.d.turner@leeds.ac.uk/2008-06-19.html#people}
** {link:People 2008-06-18|http://portal.ncess.ac.uk/access/wiki/site/~a.g.d.turner@leeds.ac.uk/2008-06-18.html#people}
** [Ties van Ark]
** [Mike Batty]
** [Mark Birkin]
** [Andrew Crooks]
** [Andy Evans]
** {link:Wolfgang Gentzsch|http://www.linkedin.com/in/wolfganggentzsch}
** [Peter Halfpenny]
** [Rene Jordan]
** [Belinda Wu]
** [Ken Kahn]
** [Howard Noble]
** [Richard Milton]

h3 Notes
* {link:Workshop 5 Agent Based Modelling for the Spatial Social Sciences|http://www.ncess.ac.uk/events/conference/programme/workshop5/}
** Participants
*** 20 people including:
**** [Ties van Ark]
**** [Mike Batty]
**** [Mark Birkin]
**** [Andrew Crooks]
**** [Andy Evans]
**** [Peter Halfpenny]
**** [Andy Hudson-Smith]
**** [Rene Jordan]
**** [Ken Kahn]
**** [Richard Milton]
**** [Howard Noble]
**** [Belinda Wu]
**** ...
** Introduction
*** [Mike Batty]
**** Some interesting history
**** An outline of workshop activities
***** Mainly to consist of presentations with questions and discussion
***** The morning talks were roughly divided into two sections
****** Theory
****** Practice
** Theory
*** Agent Based Aggregate Equilibrium Models
**** {link|Mike Batty|http://www.casa.ucl.ac.uk/people/MikesPage.htm}
**** Showed a nice demo of tube evacuation 
**** http://gisagents.blogspot.com
**** http://www.digitalurban.blogspot
**** I decided to plan to show some of the transportation visualisations for airtraffic and mass transit that I looked at yesterday [2008-06-17]
*** Agent based Modelling and Microsimulation
**** [Mark Birkin]
**** What is the difference?
**** Mark has shown a picture of the HSAR :-)
***** There is an interesting data issue with that slide:
****** Basically it should not be put online if it really is the data!
***** I'm glad Mark did this and I'll hopefully remember to refer back to this later...
**** {link:APPSIM|http://www.canberra.edu.au/centres/natsem/research-models/projects_and_models/appsim}
** Break
*** Set up for my presentation
** Applications
*** Reconstruction of the entire UK population using microsimulation
**** [Andy Turner]
**** {link:Power point presentation slides|sakai:/moses/Publications/4thInternational_e-SocialScienceConference/AndyTurnerAgentModeliingWorkshopPresentation2.ppt}
**** This talk was originally titled "Reconstructing Data Sets for Microsimulation"
**** I would have preferred the following title "MoSeS Reconstruction of Individual Level Census Data for the UK in 2001 from Aggregate Statistics and Samples of Annonymised Records"
**** In general it went OK
**** I showed some visualisations of transport data that Simon showed me yesterday :-)
***** ...
**** I was glad I had prepared some pictures for the presentation as they lightened it somewhat :-)
***** If I give another presentation of this I will want to do it more graphically and include some maps and graphs of results...
**** Questions/Feedback
***** Andy Evans asked how the Genetic Algorithm (GA) approach would compare with Simulated Annealing (SA)
****** I am not sure what is possible with SA, but I think it is easier to do a broad search of the solution space before and during conversion with a GA
******* Mainly because in the way the GA that I developed works, we can swap multiple units in and out.
******** Simulated Annealing works by swapping individual records
********* This might be fine if the search space is very smooth, but I don't think this is the case.
********** I wonder if I can show how rough the solution space is.
****** [Ties van Ark] liked my pictures because they were funny
****** [Mercedes Castilierro Arguello] also liked them :-)
******* Maybe I should produce a dynamic graphic to show the dynamic simulation model working...
*** Agent-Based Modelling for Housing
**** [Andrew Crooks]
**** {link:PDF Presentation Slides|http://www.casa.ucl.ac.uk/andrew/research/talk.pdf}
**** {link:Andrew Crooks' blog item for this: Agent Based Modelling for Housing|http://gisagents.blogspot.com/2008/06/agent-based-modelling-for-housing.html}
*** Agent based modelling of Crime
**** [Nick Malleson]
**** ...
*** Break
**** Nice chat with [Ties va Ark]
***** What will Leeds be like in 40 years?
****** Considered the difference between what we think will be and what we want to be.
**** Pointed [Andy Hudson-Smith] to {link:Mason|http://www.cs.gmu.edu/~eclab/projects/mason/} as I have heard it is good for 3D and indeed Georgios rated it :-)
***** Where are you Georgios, I miss you :-)
**** Gillian Sinclair gave me an {link:NGS|http://www.ngs.ac.uk} present of some bumf with my name and Geodemographic modelling on.
***** This is what Katy Weeks helped produce in March/May :-)
*** Visualisation of Agent Based Models (?)
**** Andy Hudson-Smith
**** Shall we all get on Second Life?
***** It sounded like an invite, but I don't know how it works still yet.
****** Perhaps it is time to learn
**** Questions and Feedback
***** I mentioned again that I rate the work on 3DIM and City GML and so on that I saw in Potsdam at the OGC TC  ***** I considered that we perhaps have something to learn from those doing cartographic generalisation
***** Perhaps we are seeing a demand for interactive scale based visualisation
Everything needs calculating on the fly although caches and dummy views can be very useful for keeping the users focus

* {link:Open Agent Based Modeling Consortium|http://www.openabm.org/}

Eduardo Pignotti
Innovators do better than imitators or vice versa?
Interesting question:
 I guess it probably helps to do some of both
 It's not really a host/parasite thing 

Wants help with PolicyGrid Evaluation
30minute session in Syndicate Room C/D any time tomorrow or Friday am.

constructing to learn
Modelling4All
 Modelling4All.nsms.ox.ac.uk
Ken Kahn (Oxford), Howard
Looks good.
Flickr for modelling, BAMZOOKi
Scratch (Based on StarLogo)
Jotit Manyeyes

** Miscellaneous 
*** Internet connection problems
**** Something made the internet/web browsing very slow on my laptop
***** I later found out that rebooting my laptop fixed the problem!
****** I should not have struggled for so long...

* Wine Reception Social Event
** A good chat with Eric Meyer and Ralph Schroeder on the way to the event :-)
*** I am looking forward to AllHands in Edinburgh and the Oxford eResearch 2008 conferences towards the end of the summer.
*** For Oxford eResearch 2008 I mentioned that it might be that I can progress my contribution sufficiently to not only get review from MoSeS colleagues, but to get them contributing also.
** Mapping discourses chat
*** I tried to argue that I have a common thinking or map with two very nice people
*** What is a map, how does it translate, how is it distinct from a geographic map
*** When we are learning/dealing with maps it is important to learn to appreciate the states, boundary conditions, limitations and other nature of the track.
*** Tautologies abound when describing maps :-)
*** Taking the contrary view my friend argued that our ideas of map and discourse are very different
*** I think in the end we agreed to disagree about some things but there was a common understanding and appreciation I think
** Andrew Crooks chat
*** He found the House Price data I provided on behalf of the land registry, not so useful in the end
*** It was either too incomplete, or did not have enough information for what he wanted
*** CASA are keen to try to disaggregate further the Individual, Household and Communal Establishment data to assign people to individual buildings or addresses
*** We had a think about auxiliary data that might help guide this
**** Telephone directory
**** Electoral role
**** GP data
**** Benefits data
**** PLASC
*** We considered random assignment as an initial start.
** Kenny Baird
*** Learning resources development are keen to support us
I need to think about what they can usefully do and then try to let them know
I suggested developing:
**** a tutorial for wikimapia
**** a tutorial for using the NCeSS Sakai portal
**** a tutorial for reconstructing the 2001 individual and household level MoSeS data
I also thought about
a tutorial for the GLS

We had an interesting discussion about:
review of user generated content and the quality of expert data
tracking people to give them more freedom
how technology can sometimes come to the rescue and is a useful backup although mentally get lost and then re-found can be reassuring and that its benefits are sometimes missed or not appreciated in contemporary society.

Chat with Alex Voss
He plans to send me the code books for the Taiwanese census data so we can consider how to do dynamic simulation with individual and household level data
The metadata for the Taiwanese census outputs is being written in English
I will take a look at this
Alex suggested a meeting or to start communicating
It would be very good to have this as an example of international collaboration

Chat with Neil Chue Hong
He wanted to know that we were still making progress with GLS




* General feedback on my blogging during the day:
** Kenny and Andrew Crooks and Andy Hudson-Smith have all said they like my blog
** Kenny likes it as it is about what I have done rather than what I think or plan to do
*** Although I do have a bit now which gives advanced notice of events attendance.
** Andrew Crooks likes it and wonders how many people read it.
*** He suggested that I improve the RSS feeds as one new feed a month is too little in his opinion