Online exploration of cultural regions, migration and ethnicity using the geography of personal names

Paul Longley & Pablo Mateos

Recent years have seen substantially increased interest in the study of population geography through the lens of the ethnicity, migration and locality. Consistent with this, ongoing research in CASA spans issues of geo-genealogy (the present day distribution of human DNA), historic migration flows, and issues of residential segregation in cities. Our focus is at a range of scales from the very local to the global, using methods that hone improved classifications of cultural, ethnic and linguistic groups from a range of data sources. The results make it possible to improve our understanding of the heterogeneity of populations at a range of scales and this, in turn, facilitates better understanding of issues such as spatial segregation processes in cities, or historic processes of settlement and population mixing.

In empirical terms, our research utilises the telephone directories or registers of electors sourced for 26 countries in four continents. Our classification software, called Onomap, assigns individuals to ethnic groups based upon surnames and forenames. A total of 185 cultural, ethnic and linguistic groups are identified based on the probable origins of names. The classification is used to identify contemporary regions and analyse residential segregation in the UK and other countries. In addition, we undertake analysis of historical populations in order to identify the probable boundaries between cultural regions and the barriers to population exchange over the centuries.

Spatial Interaction Models for Higher Education

Alex Singleton & Ollie O’Brien

Over the last 30 years or more, spatial interaction modelling and geodemographic analyses of neighbourhood conditions have each developed as successful, though quite separate research traditions. In this talk we propose an integrated model which harnesses the analytical power of spatial interaction modelling with the human behavioural insights that arise from the use of geodemographic classifications. This innovative methodological approach is applied to the problem of modelling participation in Higher Education (HE).

Through integration of national coverage school and HE data, a national model is calibrated and tested against actual recorded flows of students into HE. The model is implemented within a Java framework and is presented as a first step towards providing a quantitative tool that can be used by HE stakeholders interested in widening access to underrepresented groups.

The Dynamics of Skyscrapers Scaling, Allometry, and Sustainability

Michael Batty

We will begin with a discussion of observed regularities in cities which are represented by statistical distributions which are power laws. Unlike many distributions in the social sciences which are regarded as being 'normal', when objects which compete with each other with respect to size are examined, these follow an inverse relationship between their frequency and their size. There are few big cities but many small one, few big firms, many small, few people with big incomes, a lot with very small, in contrast to the heights of people, few very small, few very tall, but many in the middle. This phenomena has been know for centuries but it was first formally pointed out for incomes by Pareto in the late 19th century, then popularised for word frequency distributions and cities by Zipf, and recently has received a massive amount of attention for networks. Similar analogies can be developed for tall buildings but extend the conventional wisdom in three ways: first, within the rank size regularities that we see for the distributions of tall buildings, there is considerable movement in rank and we will use the device called a rank clock to look at this. This will also be introduced through city size distributions. Second the dynamics of skyscraper building in terms of the bursts of activity that take place during timed of boom and bust (witness the Burj Tower) will be presented, and third, speculations will be made on how we can use these measures of size and scaling to illustrate what happens when buildings get bigger and change in shape: allometry.

Development of an urban growth model using high-resolution historical data

Kiril Stanilov

The presentation discusses methodological issues related to the development of a cellular automata urban growth model using parcel-level information derived from Ordnance Survey maps of West London which cover a period from 1875 to 2005. The high resolution of the data, combined with the extensive longitudinal coverage and the large areal extent of the study area, allow to trace the emergence of metropolitan form as a result of the accumulation of thousands of individual fine-grain local transformations of the built environment. The results from the analysis of the documented patterns of land use change are used to develop a long-range model of metropolitan growth, exploring key spatial factors guiding the processes of diffusion, nucleation, and conurbation within London's metropolitan fabric.

The research frontier in urban modelling: the agenda and the challenges

Alan Wilson

The research frontier in urban modelling is focused on the evolution of cities and regions. The basic tools for an attack on this challenge have been available for a considerable time – using a combination of methods that relate to statistical mechanics (Boltzmann) for spatial interaction and Lotka-Volterra-like equations (from ecology) for the dynamics – generating what can now be called BLV models. However, contemporary methods of computer visualisation are offering new insights and facilitating progress. The core ideas will be explained through two case studies: first, urban retail systems – in both BLV and agent-based modelling form; and secondly, a model of the evolution of the United States urban system from 1790 to 1870.

On-line Spatio-Temporal Analysis of Network Data and Road Developments

Tao Cheng

The results of recent traffic surveys and analysis of road network performance in London indicate a decline in traffic flows and, perversely, a decline in speeds and increase in congestion. The current annual cost of congestion on London’s main roads is estimated to be in the range of £1.8 to £3 billion. Analysis of road network performance is intricate. This is because the road network is essentially an open system with many factors impacting upon it, and because travellers can adapt their behaviour and impact upon monitored performance outcomes in many different ways. These factors have different patterns of influence in both time and space, and analysis of the distinct cause-effect patterns is complicated by the non-linearity of their effects, including the possibility of abrupt growth in congestion once it sets in. Modelling spatial-temporal dependency of the factors is the bottleneck in analysis of the network performance. The challenge is to model dependency in both space and time seamlessly and simultaneously, in order that the accuracy of analysis can be improved. Another challenge is to fully accommodate the topology (links and hierarchies) and geometry (distances and directions) of real road networks in the analysis. These are also fundamental challenges to modelling the complexity of other types of networks. This presentation will report the progress made in an on-going project to tackle these challenges through innovative combination of novel machine learning methods (Support Vector Machines – SVM) with advanced statistical space-time series analysis (STARIMA). The modelling procedure integrates spatio-temporal prediction, outlier detection and 3D visualization of traffic data in Central London.

Twitter Tags and Real-Time Visualisation of Complex Geographic Data with MapTube

Richard Milton

As an extension of our previous "Mood Map" concept, where an online form was used to collect survey data geocoded by postcode, we show how "Twitter Tags" can be used to create a real-time map of UK snow depth during January 2010. We then compare this Twitter map to the measured snow depths using synoptic observation data from the UK Meteorological Office. While surveys using online forms or Twitter tags can generate hundreds of thousands of responses, producing a representative visualisation of data of this magnitude is a challenge for current web-based systems. Many mash-up websites avoid any processing of the data and simply present what has been input, for example by placing clickable pins of the most recent responses onto the map, or by displaying response counts by area. With the current moves towards more open and accessible Government data (http://data.london.gov.uk and http://data.gov.uk), a lot of high quality information is now available. Raw response counts can be processed to show percentages of the population for each area and other demographics or official statistics used to enhance the information. We show how our MapTube web site can combine these new sources of data to produce scientific visualisations and incorporate real-time crowd sourced information from surveys.

Tales of Things and electronic Memory – Creating and Mapping The Geography of Everything

Andrew Hudson-Smith

“Spimes are manufactured objects whose informational support is so overwhelmingly extensive and rich that they are regarded as material instantiations of an immaterial system. Spimes begin and end as data. They’re virtual objects first and actual objects second.” Bruce Sterling, Shaping Things, (2005).

The TOTeM project is located within the emerging technical and cultural phenomenon known as ‘The Internet of Things’. The term is attributed to the Auto-ID research group at MIT in 1999, and was explored in depth by the International Telecommunication Union who published a report bearing the same name at the United Nations net summit in 2005. The term, ‘Internet of things’, refers to the technical and cultural shift that is anticipated as society moves towards a ubiquitous form of computing in which every device is ‘on’, and every device is connected in some way to the Internet. The specific reference to ‘things’ refers to the concept that every new object manufactured will also be able to part of this extended Internet, because they will have been tagged and indexed by the manufacturer during production. It is also envisaged that consumers will have the ability to ‘read’ the tags through the use of mobile ‘readers’ and use the information connected to the object, to inform their purchase, use and disposal of an object.

Over the last 6 months the project has been developing technology to tag objects with any online media via a online service and custom written iPhone application linked to a back-end database of objects.

The talk explains the concept of TOTeM and explores the implications of adding location to any tagged item. This has the potential to allow us to know the last known location of any object in the world, opening up many questions on space, place, location, security, and privacy.

TOTeM technology allows us to connect anything with any media anywhere, through this talk we explore the potential and present the first steps in creating a Geography of Everything.