Russ Bernard's Plenary talk at the UK social networks conference...

...was a great polemic against those that seek to created a quantitative/qualitative divide in the social sciences, arguing for the naturalness of Mixed Methods.  Can be seen as arguing for the relative primacy of Evidence over theoretical frameworks, so fitting well with the methodology of the CPM, which could be summarised as: "You have to have a very VERY good reason to ignore evidence".  In descriptive simulation one often combines narrative accounts to inform the programming of the micor level (agents) and check this against quatitative evidence at the macro level.

At the UK Social Networks Conference in Manchester

Gave a talk centered around a proof concerning computability on a class of systems, showing that there is not possible network measure that will reflect underlying node importance.  I suggest that simulation could be used to stage abstraction to a social network.

The Impossibility of a General Intelligence

I just caught Drew McDermot's talk as part of the AISB 2010 symposium: "Towards a Comprehensive Intelligence Test (TCIT): Reconsidering the Turing Test for the 21st Century Symposium".  He contended that Turing never intended this as a general test of intelligence, just a test that would establish that a machine was intelligent in some sense.

More than this I have argued that there is no such thing as a general intelligence -- intelligence is very different from computation.  See:

• Edmonds, B. (2000). The Constructability of Artificial Intelligence (as defined by the Turing Test). Journal of Logic Language and Information, 9:419-424. (http://cfpm.org/cpmrep53.html)
• Edmonds, B. (2008) The Social Embedding of Intelligence: How to Build a Machine that Could Pass the Turing Test. In Epstein, R., Roberts, G. and Beber, G. (Eds.) Parsing the Turing Test.  Springer, 211-235. (http://cfpm.org/cpmrep95.html)

My Invited Talk at SNAMAS

Revealing the weakness of SNA and possibly fixing it, using MAS

Slides at:  http://bruce.edmonds.name/talks/be-snamas.ppt

A social network model consists of the representation of the target domain in terms of some system of nodes and arcs, plus how inferences about this are going to be made which can be interpreted back to the target.  This is a not an analytic results but a contingent theory that can only be validated against independent empical evidence.  The approach consists of several stages: (1) the collection of data about the structure and processes in the target; (2) the representation of this in a social network structure; (3) the inference of properties of the network using measures and other results; (4) the interpretation of these inferences back in terms of the target.  Properly considered the theory requires all stages and not simply stage (2) (I will call such a SNAT - a social network analysis theory).

To validate such a SNAT would require studies to see if there is independent evidence that the outcomes in the target system actually do correspond to the inferences from such a process (as interpreted to the target and given the data collection processes) for the range of targets that correspond to the declared scope of the theory.  Unfortunately this is rare, and more frequently a SNAT is only weakly validated against the intuitions of the same researcher that constructed the SNAT.  Partly this is due to expense of SNA and independent validation studies, but it also seems to be a result of the way SNA is divided between theoreticians and users.  The theoreticians looks at measures and other techniques that can be made of a given network system usually without any reference to observed case-studies - stage (3).  The users study observed examples and apply the techniques (frequently wrapped in software to make them more accessible) of the theoreticans to dervice conclusions about what their target systems - stages (1), (2) and (4).  Nobody checks that the combined SNAT, all four stages put together, actually works, i.e. subject it to an independent validation.

To demonstrate how SNAT are an inherently difficult and empirical approach, two cases in "Artificial Social Network Analysis" are exhibited.  That is where a MAS is studied using SNA methods.

• In an apparently simple MAS, where almost all information about the nodes, their behaviour, the social network etc. is known beforehand (everything except the initialisation of the environment), it is proved that there is NO measure that will reliably correspond to the asymptotic importances of the nodes.  Given that one can not devise a reliable measure in this ideal and very simple case, this indicates that uses of SNA measures etc. that assume a priori that a given measure is a useful indication of a property of the target system is deeply flawed.
• In a plausible simulation of a P2P file-sharing system, given information that is analagous to what a researcher of social networks "in the wild" would infer, it is evident that the wrong conclusions might well be made.  Given this is a simulation it is possible to to check whether the SNAT holds, and despite appearences is found to be lacking.  If this is the case for a plausible simulation, how can we take unvalidated SNA analyses of observed systems seriously?

Thus MAS in the form of simulations can be used to probe weaknesses in SNA approaches, showing doubtful assumptions as well as making clear the empirical and contingent nature of SNAT.

It is suggested that the root problem is the drastic nature of the abstraction step in SNA, from a complex social system to a relatively "thin" mathematical structure - a network.  However such abstraction can be staged using MAS simulations.  This has the advantage that the chain of reference from model to model is maintained and testable, but at the cost of far more work.

 Given at SNAMAS, part of AISB 2010, Leciester, March 29th 2010.

Reading: Krotoski (2009) Social influence in Second Life

Krotoski, Aleksandra K. (2009). Social influence in Second Life: Social Network and Social Psychological Processes in the Diffusion of Belief and Behaviour on the Web. PhD Dissertation. University of Surrey, Department of Psychology, School of Human Sciences.

This thesis examines which social psychological and social network analytic features predict attitude and behaviour change using information gathered about 47,643 related avatars in the virtual community Second Life. Using data collected over three studies from online surveys and data accessed from the application’s computer servers, it describes why the structure of a social system, an individual’s position in a social group, and the structural content of an online relationship have been effective at predicting when influence occurs.

Available at:

   http://alekskrotoski.com/post/academic-dissertation---social-influence-in-second-life-social-n