ACADEMIA

A researcher at the University of Leeds is using the UK’s national grid computing infrastructure, the NGS, to forecast crime rates in the city of Leeds.  Nick Malleson, a PhD student at the university, is using the NGS to carry out sophisticated modelling of possible crime scenarios that would be impossible to perform on a desk top PC.

The occurrence of crime is a complex matter and can result from the interactions of a massive number of environmental factors as well as complex human behaviours.  Malleson uses an agent-based model in which many of these factors can be taken into consideration and modelled using NGS compute resources.  An immense number of runs are required to provide sufficient material for analysis but each run was taking several days to perform on a desktop PC before Malleson switched to using the NGS.

Malleson explained “This is where the NGS was essential for the project to be feasible.  Without access to NGS resources, the project would not have been able to continue.  I can now achieve hundreds of results in a few days where I would have previously only produced one”.

So far Malleson’s model has thrown up some interesting results such as some areas of Leeds having much lower crime rates than expected despite environmental and social factors suggesting otherwise.  The model will also be able to look into the future and predict how major development schemes in the city will impact on crime figures. 

For example, the current EASEL regeneration scheme in Leeds involves many physical changes to the environment such as new houses, green spaces etc as well as social changes (rented and “affordable” accommodation etc) but what affect will this have on crime figures?  Malleson’s model could help to identify potential crime hotspots or other problems areas produced by these schemes that would be impossible to see otherwise.