Human Behaviour Modelling Workshop – 16 Sept

June 3, 2013

My interest in human behaviour modelling is no secret, so it’s no surprise that I’m excited to be running a half-day workshop on the subject at this year’s SimTecT conference in Brisbane.

The workshop will cover key issues and major steps in human behaviour modelling, including practical “how to” advice, data collection issues, verification & validation, and some common pitfalls. Applications to Defence, Mining, and Health industries will be covered (depending on the participants). Some practical examples written in NetLogo will be given, although the techniques presented will be relevant to any simulation system.

Update: Unfortunately this workshop has been cancelled.

– Tony


Human Social Research: The Lively Science?

May 9, 2013

Michael Agar of ethknoworks.com has just written a book provocatively titled The Lively Science: Remodeling Human Social Research.

Having heard Mike speak, it’s sure to be an interesting book, and the online sample chapters confirm that belief. It’s kind of ironic that, even though I’m originally a mathematician, what Mike says about inappropriate quantification really resonates with my experience. For a complete view of human activity, numbers, pictures, and stories are all important. And if they disagree, something is surely wrong.

– Tony


“Where’s the tea?” – Simulating human behaviour

May 3, 2013


The tea

In Douglas Adams’ famous The Hitchhiker’s Guide to the Galaxy, some of the characters discuss the replacement of Arthur Dent’s brain by an electronic one:

‘Yes, an electronic brain,’ said Frankie, ‘a simple one would suffice.’
‘A simple one!’ wailed Arthur.
‘Yeah,’ said Zaphod with a sudden evil grin, ‘you’d just have to program it to say What? and I don’t understand and Where’s the tea? – who’d know the difference?’

But is that true? Can simple computer models adequately simulate human behaviour?


How will a crowd of panicked humans flow down this fire escape?

In fact, it depends on the goal of the simulation. In models of pedestrian flow, work by Dirk Helbing and others has shown that quite simple models can perform very well, particularly in simulating evacuation dynamics and similar panic-driven scenarios. In these situations, simulations not much more sophisticated than simple fluid-dynamics models can reveal the benefits of, for example, zigzag designs for evacuation routes. See Helbing & Johansson, “Pedestrian Crowd and Evacuation Dynamics” (Encyclopedia of Complexity and Systems Science, 2009).

Adding more sophisticated decision-making allows us to build agent-based models of economic behaviour. Will people purchase a particular product from a particular vendor? Will vendors alter their prices up or down to match other vendors? Alison Heppenstall, Andrew Evans & Mark Birkin provide a nice example of such modelling, by simulating the spatial variability in petrol (gasoline) prices in “Using Hybrid Agent-Based Systems to Model Spatially-Influenced Retail Markets” (Journal of Artificial Societies and Social Simulation, 2006). Relatively simple models of behaviour also suffice for epidemiological models (which were discussed at the 2012 Workshop on Verification and Validation of Epidemiological Models in Washington D.C.).


Anasazi ruins, Southwest USA

One of the most well known examples of agent-based modelling using this approach is the insightful study, by Robert Axtell et al., of ancient Anasazi population dynamics in the Southwest USA. In this case, behaviour in the model was synthesised from archaeological evidence, anthropological data, and rational decision-making – households will pack up and move out if they’ve seen too many bad harvests in a row. See “Population growth and collapse in a multiagent model of the Kayenta Anasazi in Long House Valley” (PNAS, 2002), and “Understanding Artificial Anasazi” (M.A. Janssen, Journal of Artificial Societies and Social Simulation, 2009).

Related modelling methods are used in studies of land use. Will farmers switch the crops they’ve been planting? Will they fell trees in the neighbouring forest? Will they abandon farming altogether and move to the city? Alex Smajgl and others discuss approaches to such modelling in “Empirical characterisation of agent behaviours in socio-ecological systems” (Environmental Modelling & Software, 2011). Grace Villamor, Meine van Noordwijk, Klaus Troitzsch & Paul Vlek, in their paper “Human decision making for empirical agent-based models: construction and validation” (International Congress on Environmental Modelling and Software, 2012), compare the strength and weaknesses of heuristic versus optimal decision-making in models. It may be difficult to accurately capture human heuristics, but it cannot necessarily be assumed that humans will always make the “best” decision.


The emotions of fear and joy: The Rescue by John Everett Millais

The choice between heuristic and optimal decision-making in models is complicated further when the humans being modelled make decisions on emotional grounds. Stacy Marsella and his team at the University of Southern California have had considerable success in modelling human emotion. One very successful use of their approach has been the tactical language training software marketed by Alelo, which also incorporates game technology. For details, see Johnson & Valente, “Tactical language and culture training systems: Using AI to teach foreign languages and cultures” (AI Magazine, 2009). Further development of this approach is likely to have several interesting applications.

For practical purposes, then, we can simulate human brains by electronic ones. But they will not necessarily be simple.

A significantly expanded version of this post will appear as an article in the Summer 2013 issue of the Society for Modeling & Simulation International (SCS) Magazine.


Catholics in Australia

March 25, 2013

Prompted by the selection of the new Pope, I’ve been reflecting on Catholicism in Australia. The Catholic community within Australia has remained roughly steady in relative size over time, while other religious communities have grown or shrunk. To understand this better, I decided to take a look at the Australian census data for 2006 and 2011. By modelling aging over five years and factoring in known probabilities of death at different ages, I was able to compare the 2011 numbers against what might have been “expected.” This bar chart shows the resulting anomalies:

Not shown in this chart are 346,377 children under 5 as at 2011, who were either born in Australia, or who were part of immigrant families:

Age Number
0 60,092
1 68,550
2 69,815
3 72,671
4 75,249

In addition to these babies, the chart shows that for ages 5–16 and 26–53 there is a net influx, due largely to immigration, of 193,000 Catholics. These figures are consistent with the fact that roughly 20% of immigrants are Catholic.

For ages 17–25, the chart shows a net outflow (after immigration) of 50,000 young people leaving the Catholic Church. The peak anomaly is at age 21 which, given the five-year period, corresponds to people leaving at age 18 or 19. The chart also shows smaller net outflows for ages 54 and up, but these probably reflect inaccuracies in my death-rate estimates, which are based on the Australian population as a whole.

The chart below summarises these factors, which led to an increase of the Catholic population from 5,126,883 to 5,439,267 over the period (a slight relative drop from 25.8% to 25.3% of the Australian population). Not reflected in these numbers is the equally interesting demographic shift towards non-Anglo-Celtic groups. The census data also allows that phenomenon to be explored, but that is beyond the scope of this blog post.

– Tony


Would you trust these eyes?

January 30, 2013

A recent blog post at Science 2.0 reported on a fascinating study by a team of researchers from Charles University in Prague. These researchers published in PLOS One their findings on the relationship between eye colour and perceived trustworthiness of faces.

Which of these two people looks most trustworthy?


Photos by Randen Pederson (L) and “Garrett” (R) – click for details

The team (Karel Kleisner, Lenka Priplatova, Peter Frost, and Jaroslav Flegr) showed photographs of 40 male and 40 female faces to their subjects, and asked them to rate how trustworthy the faces looked. They found a significant relationship – brown-eyed faces were perceived as more trustworthy.


Kleisner, Priplatova, Frost, and Flegr found (after controlling for “dominance” and ”attractiveness”), a very significant (p < 0.001) link between eye colour and perceived trustworthiness, in both male and female faces (image from their paper)

The genius of Kleisner et al., however, was not to leave it at that, but to repeat the experiment after altering the eye colours on the photographs. This revealed that eye colour per se had no effect. Rather, the perceived trustworthiness was linked to aspects of facial shape – aspects that normally correlate with eye colour.

They write: “brown-eyed faces tended to have a rounder and broader chin, a broader mouth with upward-pointing corners, relatively bigger eyes, and eyebrows closer to each other. This was also the pattern of a trustworthy face.

These results are consistent with earlier work by Alexander Todorov, Sean Baron, and Nikolaas Oosterhof, who found that not only are certain facial shapes perceived as trustworthy, but that these faces generate a measurable response in the Amygdala, detectable with functional Magnetic Resonance Imaging. I’m not sure why this brain response exists, but presumably it is one of the things which confidence tricksters exploit.


Would you trust this brown-eyed man? Frank Abagnale, whose life was portrayed in the film Catch Me If You Can (photo by “Marcus JB”)

– Tony


Vaccination and decision-making

January 26, 2013

Vaccination is important. In Australia, for example, vaccination coverage among 5-year-olds for the terrible disease polio ranges from 93% in Tasmania and the ACT, down to 89.5% in Western Australia (the latter number is a disappointing memorial to the legendary Sugar Bird Lady).


Fortunately, the number of polio cases have been dropping as vaccine coverage improves (source).

Much of the world has now eradicated polio. Afghanistan, Pakistan, and parts of Africa still have poor polio vaccination coverage, and are continuing to see cases. Taliban murders of vaccination workers are a factor in this, sadly.


Polio vaccination coverage worldwide (source)

Frederick Chen, Amanda Griffith, Allin Cottrell, and Yue-Ling Wong, in an interesting paper reported on the Science 2.0 blog, use an online game to investigate vaccination choices. Chen et al. find that “people’s behaviour is responsive to the cost of self-protection, the reported prevalence of disease, and their experiences earlier in the epidemic.”

These results imply that vaccination rates are likely to be disappointing – perhaps dangerously so – for some of the killer diseases which the Western world has half-forgotten. Dana McCaffery and other infants did not need to die of whooping cough, for example.


Whooping cough is deadly. So is diphtheria.

The methods used by Chen et al. are also an interesting way of obtaining human behavioural data in other domains. Here is some related work.

– Tony


Language and Culture

January 20, 2013

Australian PhD student Sara Ciesielski (Graduate School of Humanities & Social Sciences, University of Melbourne) recently had the honour of presenting a “two minute thesis” for PhD Comics, on her project “Language Development and Socialization in Sherpa”:

Novelist Dorothy L. Sayers, in her short story “The Entertaining Episode of the Article in Question” (in Lord Peter Views the Body, 1928), writes about the gender-laden French language: “Now, in France, every male child is brought up to use masculine adjectives about himself. He says: Que je suis beau! But a little girl has it rammed home to her that she is female; she must say: Que je suis belle! … When I am at a station and I hear an excited young woman say to her companion, ‘Me prends-tu pour un imbécile’ – the masculine article arouses curiosity.

Referring to the same issue is Luce Irigaray’s famous book with an untranslatable title: Ce sexe qui n’en est pas un. The French language also incorporates status differences in the distinction between the pronouns tu and vous. However, other languages, such as Thai, have a far more complex pronoun structure.


Non-verbal Thai communication also requires care

Pioneer linguist Edward Sapir pointed out examples of languages where males and females used quite distinct sublanguages. Japanese is one such language, thus leading to more extreme cases of the issue pointed out by Dorothy Sayers.

Even in British English, studies show that “mauve,” “beige,” “pink,” “maroon,” “lovely,” “nice,” and “cute” are used more often by females than males. However, in 2010, Internet legend XKCD conducted a survey of colour term usage, half-expecting to see this clichéd pattern:

In fact, results were more like this:

More complex patterns of language usage mark professions, ethnic communities, and other subcultures. Writer Annie Dillard once said: “The mind – the culture – has two little tools, grammar and lexicon: a decorated sand bucket and a matching shovel.” The decorations vary from group to group. And occasionally, the shovel is used as a weapon.

– Tony


Emotion and Intelligence

January 14, 2013

A recent blog post in Science 2.0 refers to the 2004 book The First Idea: How Symbols, Language, and Intelligence Evolved from Our Primate Ancestors to Modern Humans by the late Stanley I. Greenspan and by Stuart Shanker. Drawing particularly on personal studies of child development, Greenspan and Shanker claim that “our highest level mental capacities, such as reflective thinking, only develop fully when infants and children are engaged in certain types of nurturing learning interactions.

They go onto to argue that the various stages of child development involve an intertwined growth of emotional and cognitive skills, and that these cannot be separated.

This raises the question as to whether (strong) Artificial Intelligence is possible. Can an unemotional thinking entity, like Data in Star Trek, actually exist? Such issues are explored further in a 2002 book edited by Robert Trappl, Paolo Petta, and Sabine Payr.


Unemotional thinkers in fiction, like Star Trek’s Data, actually do display various emotions – if not, the reader/viewer would lose interest

Greenspan and Shanker’s theories also have implications for child-rearing. If they are correct, emotionally rich interactions with caregivers are essential for the development of intelligence. For example, they argue (in contrast to Noam Chomsky and Steven Pinker), that language does not develop “spontaneously,” but is critically dependent on those interactions. Greenspan and Shanker write: “A child’s first words, her early word combinations, and her first steps towards mastering grammar are not just guided by emotional content, but, indeed, are imbued with it.


Emotionally rich interaction (photo: Robert Whitehead, 2006)

– Tony


School Shootings – Can Potential Shooter Profiles be Identified?

January 8, 2013

In light of the recent shootings, in Newtown, Connecticut. New debates have been sparked on the idea of gun control and identification of potentially unstable individuals who could commit such crimes.

An article on Science Daily website, School Shootings: What We Know and What We Can Do, highlighted some recent research studying past events and tragedies to accumulate a profile of potential shooters and how these individuals can be identified ahead of time. The article uses research by Dr. Daniel J. Flannery on explaining how shooters demonstrate similar features such as depression, low self-esteem, narcissism and a fascination with death. However these key aspects and similarities across shootings are not strong enough to produce conclusive profiles which could allow for future prevention of such tragedies.

Other research has produced similar findings, Leary, Kowalski, Smith and Philips (2003) analysed multiple shootings between 1995 to 2001. They found that depression, low self-esteem and narcissism were all present in the individuals involved in the shootings. However they all also shared one more common attribute and that was social rejection. Social rejection alone cannot fully explain these acts of violence as most people navigate through life and at some stage are exposed to social rejection. However this social rejection coupled with psychological problems or a fascination with death may lead to acts of violence occurring.

Unfortunately research in this area is inconclusive and therefore specific attributes and characteristics have not been idenitifed to put in place preventative measures to reduce the chances of such tragedies occuring again.

– Stefano


Emergence, Intelligence, Networks, Agents: WEIN

December 1, 2012

Networks are ubiquitous, and often large. The WWW contains over a trillion pages. The Internet contains over 900 million hosts. Facebook has over 900 million active users. The human brain contains over 80 billion neurons. The world has over over 7 billion people.

The behaviour of such networks is determined not just by the behaviour of individual nodes, but also by the network topology. The famous six degrees of separation often hold, for example. Network dynamics and self-organization are also important.

From the nodes and links of these large networks, fascinating things emerge. Nobody expected what happened when a small internal network at CERN was scaled up to planetary size, for example. And somehow, the interactions of our billions of neurons make us intelligent.

The 5th International Workshop on Emergent Intelligence on Networked Agents (WEIN’13) will be exploring some of these phenomena at AAMAS in Saint Paul, Minnesota next May.

The WEIN 2013 Call for Papers is out, and submissions are due January 30. It is likely to be an interesting event, and one that should help to build bridges between the multi-agent system and complex network communities.


Bridging disciplines: the PLoS One Map of Science

– Tony