Geographic Information Science
Seminars in Geospatial Science
The Geospatial Science seminar series provides a weekly term time forum for current research and ideas, as well as speakers from the GIS industry and prospective employers. The programme comprises a mixture of invited international and UK speakers, and researchers at the postgraduate and postdoctoral levels in UCL.
Seminars will take place between 5.00pm and 7pm on Thursdays, with a research-focussed seminar from 5pm - 5.50pm, followed by an industrial seminar. All seminars take place at UCL, Gower Street, London WC1E 6BT. Room details are given below.
8th October 2015
Chadwick Building Classroom 102
Introducing the GIS Research Team at UCL
This session will introduce you to the GIS research team (PhD students, post-doctoral researchers and lecturing staff) based in UCL's Civil, Environmental and Geomatic Engineering and Geography Departments. Come and hear about the wide range of research we're doing and meet the team!
15th October 2015
Chadwick Building Classroom 102
Research Seminar: James Haworth, SpaceTimeLab, UCL
Using fitness tracking app data to estimate cycle flows
Activity tracking apps on smart
phones and GPS devices such as Strava, Garmin Connect, MapMyFitness etc. have become very popular over recent years. Their intended use is to
track physical activity and monitor training. However, these days, many people
routinely use such apps to record their daily commutes by bicycle. At the
aggregate level, this data provides a rich source of information about the
movement and behaviour of cyclists. In this talk, I will discuss some
preliminary results of using data from one app, Strava, to estimate total cycle flows on
London’s transportation network.
Industrial Seminar: Informed Solutions
22nd October 2015
Chadwick Building Classroom 102
Research Seminar: Chen Zhong, UCL Centre for Advanced Spatial Analysis
the invisible – visible: Examples of Smart-card Data Analysis
The availability of large amounts of mobility data has stimulated the research in discovering patterns and understanding regularities. In this talk, Chen will introduce her research on smart-card data analysis. She will present a set of examples of measuring of variability at individual and aggregated levels using multi-day smart-card data. Statistical analysis, correlation matrix and network-based clustering methods are applied and potential use of measured results for urban applications will be discussed.
Download PDF here.
Industrial Seminar: Steve Brace, Royal Geographical Society
29th October 2015
Chadwick Building Classroom 102
Research Seminar: Chris Gale, University of Southampton
Night-time to day time geodemographic flows in England and Wales
Geodemographic classifications provide summary indicators of the social, economic, demographic and built characteristics of small areas. It is based on the concept that similar people are more likely to live within the same locality, undertake similar activities and have similar lifestyles. The 2011 Output Area Classification (2011 OAC) is the most recent example of a freely available national classification. The 2011 OAC and previous classifications have focussed on the night-time population due to data availability. However, the latest census, for the first time, included data on the working day population alongside a new workplace areal geography. This led to the creation of a day time geodemographic classification, the Classification of Workplace Zones for England and Wales, or COWZ-EW. The 2011 OAC and COWZ-EW can be used in conjunction with origin to destination flows, made available from the 2011 UK Census, to profile the 26.7 million journeys undertaken by the population in England and Wales. Analysing these flows using geodemographics provides a way to summarise and understand the complexities and variations of journeys made the population.
Industrial Seminar: James Milner, Developer Evangelist, Esri UK
5th November 2015
Medical Sciences AV Hill Lecure Theatre
Research Seminar: Kira Kowalska, UCL
Machine learning as a tool for extracting patterns from spatial and social networks
Technology advances enable us to connect with one another ever more easily and in many new ways. The interactions shape not only who we are but where we live and work. Michael Batty in his new book “The New Science of Cities” argues that we need to view cities as systems of networks and flows in order to understand them.
In this seminar, I will introduce keys concepts from network science that enable representing city dynamics as networks. I will then propose model-based machine learning as a methodology for extracting patterns from the networks. Examples in the talk will include, but will not be limited to, mining patterns from vehicle GPS trajectories.
Industrial Seminar: Jonathan Pitts, Gulf Keystone
12th November 2015
Reading week - no seminars
19th November 2015
B03 Ricardo LT, Drayton House, 30 Gordon Street, London, WC1H 0AX
Research Seminar: Robin Lovelace, University of Leeds
The National Propensity to Cycle Tool: building an on-line interactive mapping tool to prioritise where to invest in walking and cycling
The Propensity to Cycle Tool (PCT) was proposed in late 2014 by the Department for Transport (DfT) to enable better prioritisation of funding for cycling in England and beyond (Lovelace et al. 2015). This involved analysis of current cycling behaviour, scenario-based models, and the creation of an on-line interactive tool. The talk will focus on the interactive tool, which aims to empower policy-makers, transport planners and the public to help prioritise investment in walking and cycling today. Ultimately, this is part of a broader movement to envision and enable the changes needed for Civilisation to transition away from fossil fuels to a post carbon economy.
The paper covers the design, implementation and communication of the tool in roughly equal parts. The tool was designed from the ground-up as a policy support tool, simple enough for use by a range of stakeholders yet flexible enough to provide useful information under a range of user scenarios. The R new package shiny was used to implement these design principles in software, providing a framework for on-line user interaction and access to R's powerful graphical capabilities via the open source version of shiny-server running on remote Linux machines. User testing is still ongoing. The paper provides a unique insight into the inner workings of the NPCT, an opportunity to test the tool, and guidance on how others can use the underlying software, which is all open source, for improving cycling policies worldwide.
Industrial Seminar: Nick Samson, Knight Frank
26th November 2015
B03 Ricardo LT, Drayton House, 30 Gordon Street, London, WC1H 0AX
Research Seminar: Professor Tao Cheng, SpaceTimeLab, UCL
Space-Time Analytics of Mobility Data for An Improved Understanding People and Cities
Understanding the space time movement of people in mega cities is of great significance in transportation, regional planning, anthropology, ecology and public health. With the widespread use of mobile devices and sensors, abundant mobility datasets are collected and have been widely used to detect travel mode, traffic flows, and point-of-interests in cities. Despite the emergence of the concept of mobility as a service, how to make best use of mobility data to reveal insights into travellers to develop better services is yet to be fully realised. This talk showcases the use of space-time analytics to turn “big” mobility data (GPS, smart card and social media data) into space-time profiles of individuals and places, enabling an improved understanding of people and the city towards better service. It also explores the research challenges and opportunities in applying the pervasive and social computing paradigm to understand people, cities and travellers.
Industrial Seminar: Nick Austin, Deloitte
3rd December 2015
Cruciform B304, Cruciform Building
Research Seminar: Urska Demsar, University of St Andrews
Recent developments and ubiquitous use of global positioning devices have revolutionised movement analysis, as we are able to collect increasingly larger movement data sets at increasingly smaller spatial and temporal resolutions. This talk discusses 3D representations of trajectories in a Space-Time Cube (STC) environment. We introduce volumetric aggregations, the so-called space-time densities and present two case studies that use these densities for movement visualisation in two very different application areas. The first case study is in movement ecology, where we visualise the dynamics of animal space use over time. The second case study is in human-computer interaction, where we use space-time densities and 3D change detection methods to quantify interaction between eye and mouse movement in usability experiments. We finish the talk by outlining some of the open challenges for movement visualisation.
Industrial Seminar: Bart De Lathouwer, Director, Interoperability Programs, Europe, Open Geospatial Consortium
Title: OGC intro, update and latest trends
Abstract: A brief intro into the OGC, the structure, the workgroups, the initiatives and how you can engage in the process and contribute to the overall interoperability of systems. During the update session, we’ll spend some time on the latest additional to the “well-known standards” and also look ahead and see what is coming into the standards process.
Mini Bio: Bart De Lathouwer is responsible for planning and managing interoperability initiatives such as testbeds, pilots and interoperability experiments with an emphasis on activities in Europe. Since 2001, Mr. De Lathouwer has worked first as European liaison to the geospatial division of Autodesk and later as Autodesk's Product Manager for Server Technologies. In this role, he also served as member company representative to the OGC. As a company representative, he started the OGC CAD-GIS Interoperability Working Group (which evolved into the OGC 3DIM Domain Working Group) and managed the development of a core data access technology FDO (Feature Data Object) that later went open source in OSGeo.
10th December 2015
Chadwick Building Classroom 102, Gower Street, London WC1E 6BT
Research Seminar: Guy Lansley, UCL
CDRC Masters Research Dissertation Programme
Abstract: The CDRC Masters Research Dissertation Programme gives Masters students from across the UK the opportunity to undertake their dissertation on a predefined topic researching consumer data, with the sponsorship of a major retailer or large UK business. Due to the geographic nature of many retail data sets and the problems facing retailers today, much of the projects have been appropriately allocated to students of a geographical information sciences background. Previously undertaken projects have covered a wide range of themes including segmenting households by their average daily temporal profiles of gas usage as recorded by smart reader readings (sponsored by British Gas), evaluating the utility of geotagged social media to high street retailers (sponsored by Marks and Spenser), estimating the impact of click and collect services on existing retail networks (sponsored by Tesco) and topic modelling (Easyjet & Argos). With the application process for the 2016 programme due to open in January, Guy’s talk will outline the initiative and briefly summarise some of the dissertations which have previously been undertaken.
For more details please visit: https://www.cdrc.ac.uk/retail-masters/
Industrial Seminar: Damien McCloud, Arup
21st January 2016
Research Seminar: Guy Lansley, UCL Geography
How do we Tweet? Unlocking insight from geo-tagged Tweets in London
Abstract: Recent years have seen an increased use of social media data as a cheaper alternative to more traditional methods of market research. Twitter generate a huge quantity of georeferenced data every day and some of the data is available through their Application Programming Interface (API). Yet the representativeness of Twitter data is not well known and the quality of Tweets' textual data is impeded by the imposition of a 144 character limit. This seminar explores how location based insight can be achieved from geo-tagged Tweets by analysing a large volume of Tweets from London and through linkage to other data sets. The work presents techniques to estimate the typical usage of the service and the demographics of users. The research also explores text mining approaches to segment Tweets into 20 key topics, and 100 subtopics. This Twitter classification has been used to summarise what users in London Tweet, and how this varies by time, location and also by the users’ characteristics. The results identify how urban activity and geodemographics influence how individuals behave on Twitter across the typical weekday.
Industrial Seminar: Sandrine Balley, London Borough of Hackney
28th January 2016
Research Seminar: Huanfa Chen, SpaceTimeLab, UCL
Guidelines and Development of an Effective Online Police Patrol Strategy
Abstract: Police patrolling is one of the most important methods for crime prevention and emergency response in urban areas. An effective routing strategy for daily operations is necessary to maintain the effects of hotspot policing and reduce crime and disorder. Existing patrol routing strategies are not suitable as they fail to meet challenges including minimising the average time lag between two consecutive visits to hotspots, as well as coordinating multiple patrollers and imparting unpredictability in patrol routes. Clear guidelines and evaluation measures for police patrol routing strategy are essential to facilitate the development of effective strategies. In this research we propose a framework of guidelines and evaluation measures for police patrol routing strategy to meet the challenges of police patrol. Following these guidelines, we develop an innovative heuristic-based and Bayesian-inspired online strategy for routing police patrols. This strategy aims to make the optimal choice of next-step hotspot, utilising a Bayesian framework to combine multiple decision variables. Using real-world data and a benchmark patrol strategy, agent-based simulations are implemented to testify the efficiency, flexibility, scalability, unpredictability, and robustness of the proposed strategy and the usability of the proposed guidelines.
Industrial Seminar: Crossrail GIS Team
4th February 2016
Research Seminar: Kelvin Wong, UCL Civil, Environmental and Geomatic Engineering
What is 3D GIS?
Abstract: This talk will provide you with a short introduction into the exciting and wonderful world of the third dimension. 3D GIS is the future of our field, but what can it help us do and what are the advantages over 2D? What are the challenges we face in academia and industry?
Industrial Seminar: Jo Cook, Astun Technology
11th February 2016
Research Seminar: Gareth Boyes, UCL Civil, Environmental and Geomatic Engineering
Industrial Seminar: Integraph
25th February 2016
Research Seminar: Internal Careers Advice Seminar
Industrial Seminar: Ian Sheppard and Pippa Day, WSP Group
3rd March 2016
Research Seminar: Internal Careers Advice Seminar
Industrial Seminar: British Red Cross and Medicins Sans Frontiers
7th March 2016 - UCL/AGI GIS Careers Fair
This fair brings together established GIS professionals, new entrants and current and prospective GIS students and includes talks about how GIS is used in various industries and skill-sets required for employment, as well as mock interviews, CV reviews and networking. Details can be found http://www.eventbrite.co.uk/e/ucl-geographical-information-science-gis-careers-event-2016-tickets-16296024849
17th March 2016
Research Seminar: Kurtis Garbutt
Industrial Seminar: TBC
21st January 2014
Damien McCloud (Arup)
GIS at Arup
28th January 2014
Iain Willis (EQECAT)
To be confirmed
To be confirmed
4th February 2014 (11:00-12:00 - Location TBC)
Professor Bin Jiang (University of Gävle, Sweden)
Head/Tail Breaks for Visualizing the Fractal or Scaling Structure of Geographic Features
Things surrounding us are unevenly distributed, implying that instead of more or less similar things, there are far more small things large ones, e.g., far more small cities than large ones in a country (Zipf 1949), far more small city artifacts than large ones in a city (Salingaros and West 1999, Jiang and Liu 2012). This phenomenon is often referred to as fractal or scaling in the literature (Mandelbrot 1982, Bonner 2006). To better visualize the fractal or scaling structure, we developed a new classification scheme, namely head/tail breaks, which recursively divides things (or large things) into two imbalanced parts: a minority of large things in the head and a majority of small things in the tail (Jiang 2013). Unlike conventional classification methods like natural breaks (Jenks 1967), this new classification scheme helps efficiently and effectively visualize the underlying fractal or scaling structure in a simple and unique way. More importantly, head/tail breaks adds deep implications to statistical mapping (Jiang 2013), map generalization (Jiang, Liu and Jia 2013), cognitive mapping (Lynch 1960, Jiang 2013), and perception of beauty (Alexander 2002, Jiang and Sui 2014). The head/tail breaks derived hierarchy (or the number of the classes), namely ht-index (Jiang and Yin 2014), can be used, as an alternative to fractal dimension (Mandelbrot 1982), for quantifying complexity of geographic features, or fractals in general.
Dr. Bin Jiang is Professor in GeoInformatics and Computational Geography at University of Gävle, Sweden. He is also affiliated to Royal Institute of Technology (KTH) at Stockholm via KTH Research School. He worked in the past with The Hong Kong Polytechnic University and the University College London’s Centre for Advanced Spatial Analysis. He is the founder and chair of the International Cartographic Association Commission on Geospatial Analysis and Modeling. He has been coordinating the NordForsk-funded Nordic Network in Geographic Information Science. His research interest is geospatial analysis and modeling, in particular topological analysis of urban street networks in the context of geographic information systems. He is Associate Editor of international journal Computer, Environment and Urban Systems.
4th February 2014 (16:30-17:30)
Chris Phillips (MapAction)
To be confirmed
To be confirmed
11th February 2014
Jo Cook (Astun Technologies)
To be confirmed
To be confirmed
25th February 2014
Richard Goodman (Intergraph) and Ralph Diment (Intergraph)
To be confirmed
To be confirmed
4th March 2014
David Edem (Tullow Oil) and Nat Evatt (Tullow Oil)
To be confirmed
To be confirmed
11th March 2014
Speaker to be confirmed
To be confirmed
To be confirmed
18th March 2014
Jouke van Dijk (University of Groningen, The Netherlands)
To be confirmed
To be confirmed
8th October 2013
Professor Tomoki Nakaya (Ritsumeikan University, Japan)
I introduce two case studies to demonstrate the possibility to use three-dimensional mapping of events in a space-time cube with space-time statistical data transformation for an effective spatial reasoning about spatio-temporal patterns of disease and crime outbreaks. In the case of crime analysis about snatch-and-run in Kyoto, 2003-2004, the methodology revealed that transient clusters alternatively appeared in a pair of hotspot regions. It suggests a new type of “displacement” phenomenon of crime. In the case of historical disease mapping about Typhoid fever outbreaks in Kyoto, 1928-9, the modern analytical approach to the old material provides a clearer vision on the spatio-temporal structure of detected outbreaks, and reveals a distinct discrepancy between the described outbreaks on the old maps and statistically detected clusters as space-time gaps of cluster domains. The meaning of the space-time gap may reflect social discrimination of the public health officers against residents of the deprived areas/urban slums at those age. Finally, further possibilities related to space-time cluster detection will be briefly argued.
Tomoki Nakaya is Professor of geography at the College of Letters, Ritsumeikan University in Japan and co-director of the Institute of Disaster Mitigation for Urban Cultural Heritage. He was educated as geographer in the Department of Geography, Faculty of Science at the Tokyo metropolitan University and obtained Ph.D. from the university in 1997. He specialised in spatial statistics and mathematical modelling in human geography. One of his recent major research themes is to integrate spatial or space-time statistics and GIS-based visualisation for better understanding of various geographical phenomena with special interest in geography of crime and health.
15th October 2013
Patrick Rickles (UCL)
No Title or Abstract
22nd October 2013
Sarah Wise (UCL)
Modeling Hotspots: An Agent-Based Approach to Simulating the 2012 Colorado Wildfire Evacuation
Crisis response is a time-sensitive problem with multiple concurrent and interacting sub-processes. In such a high-stress environment, an individual's heterogeneity, social relationships, and spatiality all influence her chance of survival. Traditional methodologies have trouble incorporating these features of the evacuee's experience, but agent-based modeling offers a way to explore these complex systems. As an example, I have developed a simulation of the 2012 wildfires which swept across the US state of Colorado. These wildfires prompted the evacuation of over 30,000 citizens from the city of Colorado Springs alone. I demonstrate how individual characteristics, social relationships, spatial explicitness, information gathering, and decision-making processes can be applied to this system in order to explore how the evacuation proceeded and how future work might utilize such a system to help plan for disaster evacuation efforts.
29th October 2013
Alistair Leak (UCL) and Kira Kowalska (UCL)
In what may be considered the age of information social data provides new and potentially valuable insight into the routine and behaviour of real-world populations. Utilising the space-time component of recorded Twitter data, the project seeks to address weaknesses in current demographic products, namely the night-time representation of populations. Alistair and Kira will provide an overview of their work within the Uncertainty of Identity project including the use of heuristics to infer age, gender and ethnicity, and the development of a Twitter based population record comparable in structure to that of the Electoral Roll. The premise of the above being that a name is a statement of an individuals identity. The presentation will conclude with a benchmark of the Twitter derived population versus the 2007 Electoral Roll and 2011 population estimates. This work will contribute in the development of a demographic classification of online social media.
12th November 2013
Michalis Vitos (UCL), Julia Attenbuchner (UCL) and Gill Conquest (UCL)
Citizen science in the rainforest
The UCL Extreme Citizen Science (ExCiteS) Research Group aims to develop theories, tools and methodologies that will enable any community – no matter their geographical location or standard of technical, numerical or textual literacy – to start a Citizen Science project to deal with issues that concern them. For such projects to succeed requires the engagement of communities, overcoming many technical, human-technology interaction and cultural challenges; our interdisciplinary group achieves this by drawing on the knowledge of geographers, anthropologists, computer scientists, Human-Computer Interaction (HCI) experts, designers, electronic engineers, ecologists and other fields.
19th November 2013
Monsuru Adepeju (UCL)
Space-Time Pattern Analysis of Big Crime Datasets: Hotspots Detection for Predictive Policing
Predicting where and when crime is likely to occur is an important aspect of police operation, especially toward the end of effective deployment of limited police resources. Crime hotspot detection is generally used, relying on historical crime datasets, to produce patterns of crime over large temporal steps such as monthly or seasonally. These patterns are then used to project the locations where crimes are likely to occur in near future at the same or larger scales. However, the use of large temporal scales are not suitable for proactive day-to-day crime intervention, especially at the local level. In this study, we combine both the global and local approach of hotspots detection to proactively identify areas of consistently high crime concentration at a finer temporal scale, i.e. daily, over a certain period of time. This approach provides better visualisation and profiling of crime patterns for better predictive policing.
22nd November 2013
Professor Manfred Fisher (Vienna University of Economics and Business)
A Bayesian approach to identifying and interpreting regional convergence clubs in Europe
This study suggests a two-step approach to identifying and interpreting regional convergence clubs in Europe. The first step involves identifying the number and composition of clubs using a space-time panel data model for annual income growth rates in conjunction with Bayesian model comparison methods. A second step uses a Bayesian space-time panel data model to assess how changes in the initial endowments of variables (that explain growth) impact regional income levels over time. These dynamic trajectories of changes in regional income levels over time allow us to draw inferences regarding the timing and magnitude of regional income responses to changes in the initial conditions for the clubs that have been identified in the first step. This is in contrast to conventional practice that involves setting the number of clubs ex ante, selecting the composition of the potential convergence clubs according to some a priori criterion (such as initial per capita income thresholds for example), and using cross-sectional growth regressions for estimation and interpretation purposes.
Manfred M. Fischer is the Professor of Economic Geography & GIScience at the Vienna University of Economics and Business. He is also an Adjunct Professor in the State Key Laboratory of Resources and Environmental Information Systems at the Chinese Academy of Sciences in Beijing. He received his MSc and PhD degrees in Geography and Mathematics from the University of Erlangen, and his habilitation degree from the University of Vienna. He is one of the founding editors of the Journal of Geographical Systems, and the book series Advances in Spatial Science, and co-editor of the Springer Handbook of Regional Science. For nearly four decades Professor Fischer has been involved in the development and application of (mathematical and statistical) models, methods and techniques in regional science and related disciplines, a time period during which he has published 40 books, and about 250 refereed journal articles and book chapters. For a full list of his publications see his Curriculum Vitae. In recognition of his achievements he was elected a Fellow of the International Academy of Sciences for Europe and Asia (1995), the Austrian Academy of Sciences (1996), the Royal Netherlands Academy of Arts and Sciences (1999), and the Regional Science Association International (2006). In 2013, he also received the 2012 European Investment Bank Prize of the European Regional Science Association.
26th November 2013
Jens Kandt (UCL)
Bio-social relations and the study of health disparities. Genes, place and social class.
Health disparities have long been of interest in health geography and have often been investigated with a strong focus on social and environmental aspects of place. While many of these studies have yielded contradicting evidence with respect to an independent contribution of place in shaping health, latest findings in epigenetics suggest that health research could be enhanced with biological information on local populations. In this presentation, I introduce current approaches in the study of spatial health disparities and discuss how data on surnames may be used to incorporate the biological dimension into the geographic inquiry on health. Drawing on preliminary observations made while working on three datasets – a genotyped sample of 4,300 individuals throughout the UK, the 1881 Census of Great Britain and the 2007 UK Enhanced Electoral Roll – I will suggest ways to gauge the potential of these datasets to specify biological characteristics of populations alongside other, contextual aspects of place.
3rd December 2013
Garavig Tanaksaranond (UCL)
Visualisation of traffic in space-time.
Visualisation is an effective tool for studying traffic congestion using massive traffic datasets collected from traffic sensors. Existing techniques can reveal where/when congested areas are formed, developed, and moved on one or several highway roads, but it is still challenging to visualise the evolution of traffic congestion on the whole road network, especially on dense urban networks. To address this challenge, this thesis proposes various visualisation techniques: the wall map, the isosurface, and the network constrained isosurface.
This presentation will explain how the visualisation techniques were developed and how the massive amounts of traffic data were organised to improve multidimensional query response time. The development of Graphic User Interface (GUI), which allows users to interact with the traffic data and also to manipulate the visualisation, will be also explained.
10th December 2013
Chris Gale (UCL)
Creating a New Open Geodemographic Classification of the UK
The current Output Area Classification (2001 OAC) is a free open geodemographic classification of the UK. Created using the 2001 UK Census, it represents a geodemographic view of the UK from over a decade ago. The extent of how representative it remains today is debatable. The perceived degradation of reliability over time, complicated by spatial variance, has led to a decline in users of the classification. The release of 2011 UK Census data allows for a new version of the classification, the 2011 Output Area Classification (2011 OAC), to be created. The 2011 OACs methodology will be made freely available and utilise open-source programs where possible to make the classification fully reproducible. This seminar will discuss the processes involved in creating the 2011 OAC and examine some of its outputs.
15th January 2013
Chris Phillips (Map Action)
Mapping in Disaster Response Situations.
The talk will introduce MapAction (who they are, what they do and why they do it) and will cover the role of mapping and how great /critical/all important maps are in disaster response situations. Chris will use his real-world experience of disaster situations, in particular Haiti, as case studies throughout the talk, and will also discuss what’s happening in the wider arena of humanitarian mapping.
22nd January 2013
Damien McCloud (Arup)
GIS at Arup
Damien will be talking about the use of GIS at Arup, which is a multi-national firm of designers, engineers, planners, consultants and technical specialists (responsible for, amongst many other things, the roof of Kings Cross Station). You can find some introductory information about their use of GIS here.
29th January 2013
Jo Cook (Astun Technology)
Open Source and Astun Technology
Jo will give a brief summary of what open source is, and what it's not, and talk about what they do at Astun Technology. She will discuss the technologies that she thinks are key in open source at the moment, then will talk about some ways forward for people interested in taking things further.
Jo has previously worked as an archaeologist with Oxford Archaeology, and started working with Open Source technologies as a way to overcome financial constraints on buying software. She became involved in the Open Source Geospatial Foundation (OSGeo) in 2006, and has been heavily involved ever since, founding the UK chapter, and having a stint on the Board of Directors. She made the decision to concentrate full time on open source GIS about 2.5 years ago, and now works for Astun Technology, where they have a suite of mapping products all built on the open source stack
5th February 2013
Richard Goodman (Intergraph) and Ralph Diment (Intergraph)
No Title or Abstract
19th February 2013
Mike Worboys (University of Greenwich, London)
Have formal approaches a role in GIS?
In this talk, I survey some past and current contributions of mathematics and logic to geographic information systems and science. We will also look at some possibilities on the horizon. I hope to make a convincing argument that formal approaches have and will continue to play a crucial part in the development of GIS.
26th February 2013
Nart Tamash (Crossrail) and Dan Irwin (Crossrail)
GIS at Crossrail
Dan Irwin and Nart Tamash will provide an overview of the application of GIS within the design, construction, and management of the largest infrastructure project in Europe. It will include an overview of the project, show how GIS fits into the framework of the organisation, explain how it is used across a variety of technical areas, and elaborate upon the future use of GIS both within the project and the transport industry as a whole.
5th March 2013
AGI Panel Discussion
Members of the AGI (Association for Geographic Information) will form a panel to answer your questions on career options related to Geographic Information, based on their own experiences. From a variety of backgrounds, panel members will give a brief summary of their career, the route that brought them into GI-related work, and the skills which helped them get there. The panel will then try to answer your questions on pathways into the GI industry.
12th March 2013
David Edem (Tullow Oil)
GIS and Tullow Oil
David Edem is the Global GIS Manager for Tullow Oil, one of the largest independent oil and gas exploration and production companies in Europe. GIS is vital to the oil and gas industry, as companies map their assets, competition, and areas they wish to investigate for new opportunities. David will be covering how GIS has shaped his career and how he's used it to make a visible impact on shaping how Tullow Oil operates.
9th October 2012
Paul Richards (Kingston University London)
Geo-policing classifications have been inspired by geodemographic classifications, the domain structure of Indices of Multiple Deprivation and the British Crime Survey spreadsheets. The presentation will outline the context, motivation and purpose of the geo-policing classifications as well as showing how they may be created using Self Organising Map software.
16th October 2012
Paolo Battino (UCL)
The use of virtual globes, and in particular Google Earth, is increasingly popular in both formal and informal education. Basic animated tours authored with Google Earth are relatively easy to produce and those produced so far have been educationally valid in visualization terms. However, an earlier study revealed that students tend to use tours passively. This project aims at investigating how Google Earth Tours (GETs) can be best combined with related student activities, i.e. how they can be integrated in an Active Learning approach. In this seminar we will see how we approached this challenge in terms of devising an evaluation strategy to 1) evaluate the learning outcome ("how much" they learned) and 2) evaluate the user activity ("how" they learned). We will describe how we balanced qualitative and quantitative evaluation methods, how we developed a web-based logging tools specifically for Google Earth and how we managed to apply eye-tracking (working out-of-the box on a screen-wide scale only) to a geographical scale, with no use of extra proprietary software.
23rd October 2012
Suresh Veluru (City University London) and Muhammad Adnan (UCL)
Uncertainty of Identity: Identities in email addresses and social network data
This presentation proposes the methods of classifying email addresses and social network data. Email addresses can hold clues to cultural, ethnic and linguistic identity. Previous research has identified the cultural, ethnic and linguistic characteristics that can be ascribed to individuals on the basis of forename and surname conventions (Mateos et al 2011). Clues to family names are also present as substrings in e-mail addresses, and here we propose methods of categorising e-mail addresses based on the semantics of surnames. The fission of different lifestyles and advent of new media and ubiquitous computing are embedding technology ever deeper into human identity. There has been a tremendous rise in the growth of online social networks all over the world in recent times. Here we present the analysis performed on the Twitter data to identify the aspects of cultural and ethnic identity.
P Mateos, P A Longley, D O’Sullivan 2011. Ethnicity and population structure in personal naming networks.
PLoS ONE (Public Library of Science), 6(9) e22943, 1-12
30th October 2012
Michael Goodchild (University of California; Santa Barbara and University of Washington)
Traditional geographic information provided by authoritative sources results from a lengthy and labor-intensive process of synthesis. Census data, for example, is compiled from billions of raw observations, and little detail is available about the process by which synthesis is achieved. By contrast, the vast amounts of geographic information that are now appearing on the Web are largely produced by non-experts, and any integration is likely achieved by software rather than the intervention of experts. Several examples are cited. Asserted geographic facts are clearly of variable quality, and three general strategies are advanced for addressing quality control: Linus's Law, social strategies, and strategies that make use of fundamental geographic knowledge.
13th November 2012
James Howarth (UCL)
Space-time analysis and forecasting of road network data
Traffic congestion is a major problem in many large cities, with huge economic and environmental cost. One of the ways in which congestion can be alleviated is through accurate forecasting of traffic variables such as travel times and flows, which can be used to inform traffic control strategies and provide predictive information to road users. In this seminar, the modelling and forecasting of travel times in London is discussed. I begin with a description of how the space-time dependency between road links on the network can be quantified, then discuss some of the forecasting methodologies that have been used within the STANDARD project at UCL.
20th November 2012
Patrick Rickles (UCL)
Achieving Interdisciplinarity with GIS
Geographic Information Systems have been used by various disciplines to add richness to their analyses; making it a great integrator for interdisciplinary projects. My research focuses on how to identify what people need to know about GIS, to facilitate its use in their work, and how to effectively teach them. In this seminar, I will discuss the challenges of interdisciplinary research, a few case studies of projects involving multiple disciplines, and some concepts of how to teach GIS to ensure its successful uptake.
27th November 2012
Mirco Musolesi (University of Birmingham)
Interdependence and Predictability of Human Mobility and Social Interactions
The study of the interdependence of human movement and social ties of individuals is one of the most interesting research areas in computational social science. Previous studies have shown that human movement is predictable to a certain extent at different geographic scales. One of the open problems is how to improve the prediction exploiting additional available information. In particular, one of the key questions is how to characterise and exploit the correlation between movements of friends and acquaintances to increase the accuracy of the forecasting algorithms. In this talk I will discuss the results of our analysis of the Nokia Mobile Data Challenge dataset showing that, by means of multivariate nonlinear predictors, it is possible to exploit mobility data of friends in order to improve user movement forecasting. This can be seen as a process of discovering correlation patterns in networks of linked social and geographic data. I will also show how mutual information can be used to quantify this correlation; I will demonstrate how to use this quantity to select individuals with correlated mobility patterns in order to improve movement prediction. Finally, I will show how the exploitation of data related to friends improves dramatically the prediction with respect to the case of information of people that do not have social ties with the user. More information about this project can be found here.
Dr. Mirco Musolesi is a Senior Lecturer at the School of Computer Science at the University of Birmingham. He received a PhD in Computer Science from University College London in 2007 and a Master in Electronic Engineering from the University of Bologna in 2002. From October 2005 to August 2007 he was a Research Fellow at the Department of Computer Science, University College London. Then, from September 2007 to August 2008 he was an ISTS Postdoctoral Research Fellow at Dartmouth College, NH, USA, and from September 2008 to October 2009 a Postdoctoral Research Associate at the Computer Laboratory, University of Cambridge. Before joining Birmingham, he was a Lecturer at the University of St. Andrews. His research interests lie in the broad area of networked systems with a current focus on large-scale data analysis, mobile sensing, social computing and network science (in particular, social and spatial networks). More information about his research profile can be found here.
4th December 2012
Christian Nold (UCL)
Mapping the Material Practices of Citizen Sensing
This presentation discusses Citizen Sensing as an entanglement of devices & subjective experience and power relations. Using a case study of participatory noise pollution to identify the ontological and epistemological conflicts about what is being mapped and to what ends. The aim of this talk is to provide a real-world example of the way that subjectivity as a material affect is engaged in a political struggle over legitimacy. This approach aims to bridge a cultural and technical understanding of Citizen Sensing, by suggesting a critical material practice which can engage communities around real world issues.
11th December 2012
Monsuru Adepeju (UCL)
Prospective Surveillance of Emerging Crime Patterns in Central London.
Understanding the dynamics of “repeated” crimes in space-time as well as their relationships with other social economic factors could provide clues to understanding how crimes emerge over a period of time within a geographical region. Here the Space Time Scan Statistical (STSS) method is applied to analyse the crime patterns in Camden borough of London, in prospective manner to detect ‘emerging’ space-time clusters of crimes. The case study shows the possibility of detecting crime clusters as they emerge simultaneously in both space and time before becoming statistically significant. The effectiveness of each prospective cluster’s surveillance was evaluated by comparing its ‘emergence’ date with its respective retrospective cluster ‘start date’. The overall significance of the results obtained in this analysis demonstrates that this technique is able to proactively identify areas of potential crime outburst so as to facilitate immediate prevention strategies in such areas.
11th October 2011
Dan Lewis (UCL) and Muhammad Adnan (UCL)
GIS Partnerships and Ph.D. research
Ph.D. research in GIS-related topics is focused upon the unique, complex and sometimes difficult issues that arise when asking the fundamental question ‘where?’ Much research in the area is funded by research councils, foundations and charities with interests in the science of problem-solving. But because GIS is also fundamentally an applied problem-solving technology, it can also attract research funding from organisations in the commercial, public service and government sectors. This seminar will feature the work of two recent research projects in UCL: work by Dan Lewis, who has been co-funded by Southwark Primary Care Trust in the health sector, as well as the Economic and Social Research Council; and the research of Muhammad Adnan, who has completed a Ph.D. part time while working on a Knowledge Transfer Partnership with Local Futures Group Ltd.
18th October 2011
Jiaqiu Wang (UCL) and Ioannis Tsapakis (UCL)
The STANDARD Project: quantitative analysis of traffic congestion and traffic prediction in London
This presentation will focus upon short-term prediction of travel time using neural networks to accommodate space-time autocorrelation. A space-time neural network is proposed to accommodate integrated space-time autocorrelation in transport networks. An experiment is used to demonstrate that the space-time neural network model can yield accurate and robust travel time predictions based on real data.
25th October 2011
Zorica Nedovic-Budic (University College Dublin)
GeoWeb and the potential synergy of information and knowledge production sources for spatial planning and policy
Spatial planning and policy is highly dependent on integration of data from variety of sources to ensure its information and knowledge-base (aka planning intelligence). The move from standalone decision-support and geographic information systems (GIS) to WebGIS, and on toward spatial data infrastructures (SDI), has in many ways revolutionalised the planning and policy processes. However, the promise of democratisation that the new technologies would bring about has not been easy to achieve. Public participation GIS and locally controlled (bottom up) systems have been promoted by academics, but never fully integrated with other more formal and institutionalised sources of information and its derivative knowledge. The development of Web 2.0 has enabled contribution of user content via Internet. The phenomena of volunteered geographic information (VGI) and more broadly conceived GeoWeb practices – ranging in scope from Google Maps and Wikimappia to more focused OpenStreetMap and other topical (e.g., environmental) applications – offer complementary sources to the official spatial data infrastructure (SDI). These new sources are more open for two-way exchange of spatial data and production of knowledge to better inform and improve the process of planning and policy formation and decision-making. While their potential to enhance the access and completeness of geospatial information is intuitively recognised, the reality of the synergy is less understood. Why various web users contribute geographic content and how we incorporate and use it in the planning and policy processes are important questions to examine before the potential is captured and channelled into useful and sustainable networks.
1st November 2011
Ross Maciejewski (Arizona State University)
Syndromic Surveillance Using Geovisual Analytics
As data sources become larger and more complex, the ability to effectively explore and analyze patterns amongst varying sources becomes a critical bottleneck in analytic reasoning. Incoming data contains multiple variables, high signal to noise ratio, and a degree of uncertainty, all of which hinder exploration, hypothesis generation/ exploration, and decision making. To facilitate the exploration of such data, advanced tool sets are needed that allow the user to interact with their data in a visual environment that provides direct analytic capability for finding data aberrations or hotspots. In this talk, I will present a suite of tools designed to facilitate the exploration of spatiotemporal datasets. These tools allows users to search for hotspots in both space and time, combining linked views and interactive filtering to provide users with contextual information about their data and allow the user to develop and explore their hypotheses. Use cases will focus on syndromic surveillance emergency department records, while presenting details on how such tools and techniques can be applied to a wide array of problems.
15th November 2011
Tyng-Rong (Jenny) Roan1 (UCL) and Zhiwei Cao2 (UCL)
Modelling pedestrian evacuation
Reducing geo-processing errors by using a combinative geo-processing
Modelling pedestrian movement is a challenge task, because human behaviour is complex and every individual is unique. This research focuses on the situation of fire disasters in pubic buildings. To simulate a more realistic evacuation situation, this model demonstrates human evacuation behaviour based on fire investigation reports using agent-based models.
Traditional geo-processing is implemented sequentially, and their final results are influenced by various issues, such as data uncertainty and computation efficiency. To reduce the effect of these problems, this project describes a combinative method for geo-processing, thereby avoiding data processing errors, large storage of intermediate values, and massive computations.
22nd November 2011
Ross Purves (University of Zurich)
Exploring place through user generated content
Online resources such as Flickr and Gumtree provide us with new sources of data to explore the multiple ways in which locations are described and perceived by individuals. Together, these data can provide us with possibilities to develop more place-centred geographies, for example through exploration of how regions are named and perceived. In my talk I will describe methods developed to explore such data, and show how previous empirical studies on the description of place can be replicated with new media. Furthermore, I will show how such knowledge can be applied to the automatic annotation of georeferenced images with keywords, and comment on both the promise and limitations of work with user generated content.
29th November 2011
Ed Manley1 (UCL) and Berk Anbaroglu2 (UCL)
Understanding and modelling non-recurrent urban road congestion
Detection of emergent non-recurrent traffic congestion
Many urban phenomena are a result of the actions and interactions of many individuals, acting independently. This work seeks to better understand this relationship, specifically where individuals contribute towards the formation of non-recurrent road congestion. Recent research into understanding the behaviour of individuals is presented, with further description of how simulation can be used to better understand these occurrences.
Non-recurrent traffic congestion is one of the most important problems that cities currently face. This talk describes the adaptation of a statistical method which is proven in epidemiology to detect emerging disease outbreaks to the detection of non-recurrent congestion. The approach offers the prospect of mitigating the confounding effects of non-recurrent congestion could be mitigated.
6th December 2011
Professor Paul Longley (UCL)
Geocomputation and Geodemographics
Geocomputation has come to prominence at a time when the amount of available data pertaining to an ever-widening range of human activities is without precedent. What is less well known is the provenance of many such sources or, as a consequence, how depictions based upon rich but partial digital data may be triangulated with more analytically robust spatial data infrastructures. Recent research has also demonstrated some of the ways in which the outputs of data reduction and pattern seeking techniques are sensitive to initial analytical conditions and choice of particular technique. This presentation describes some recent progress in computationally intensive processing of a range of novel data sources that can be used to devise neighbourhood scale classifications of people and places. The results might be used to identify and discriminate between long and short term dynamics of change. The presentation concludes with speculation about the prospects of relating such classifications to emergent geographies of virtual interactions and activity patterns.
13th December 2011
Patrick Weber (UCL)
No Title or Abstract
10th January 2012
Muhammad Adnan (UCL)
Spatio-temporal linkage of real and virtual identity
A name often provides an indication of its bearer’s cultural, ethnic and religious affinity in the real world (Mateos et al 2011), as well as the place in which he or she probably lives (Cheshire and Longley 2012). This presentation begins to consider whether and how tokens of virtual identity can be linked to the probable characteristics of people and places.
We begin with a retrospective on two strands of work at UCL that has used computationally intensive analysis to classify populations into: (a) ‘naming networks’ based upon social similarities in naming conventions; and (b) ‘surname regions’, based upon locational proximity of people who share names with common geographic roots. Together, these two approaches offer the prospect of context-sensitive generalisation of the geography of naming conventions, as well as measurement of the long term effects of population change.
The second part of the presentation begins to consider the linkage between georeferenced email addresses and the probable names of their owners. We begin a preliminary investigation of data harvested from the worldnames.publicprofiler.org website, which are used to develop a spatio-temporal model of virtual identity formation. We suggest priorities and directions for future research that will link identities in real and virtual spaces.
17th January 2012
Graham Clarke (University of Leeds)
GIS and spatial modelling for retail forecasting
This seminar will examine how major UK retailers use GIS and spatial modelling to identify new sites for development. It will be argued that although GIS is useful by itself, these organisations need the greater predictive power of spatial models, such as gravity or spatial interaction models. However, it will also be argued, through many case studies, that the models themselves need customisation and disaggregation in order to work effectively.
24th January 2012
Wayne Marsh (Crossrail)
GIS solutions for London’s Crossrail
Crossrail is the largest civil engineering project in Europe and the largest single addition to the London transport network in over 50 years. It has been designed to provide a new railway network for London and the South East and carry 200 million passengers a year. Within Crossrail, GIS is being used through the entire lifecycle of the project, including design, construction and maintenance, integrating and joining up data such as BIM and Asset Registries. At the heart of the GIS solution is an Oracle Spatial 11g server acting as the master repository and spatial analysis tool, glueing the information together. This talk will discuss how Crossrail arrived at this solution, how it is currently being used and how we plan to enhance it in the future.
31st January 2012
Keith Dugmore (Demographic Decisions Ltd.)
The developing agenda for retailing research
This presentation will set out the ways in which the retail industry makes use of GIS as a tool for operational and strategic decision making, drawing upon the experience of the 15 major Business-2-Consumer companies that are members of the Demographics User Group (DUG) www.demographicsusergroup.co.uk. It will also set out the ways in which current government ‘open data’ initiatives are of interest to retailers, and speculate upon ways in which retailers may seek to reuse government data. The final part of the presentation will describe ways in which the Economic and Social Research Council are seeking to incentivise MSc and PhD students to conduct research into retailing issues, and the resources that are becoming available to encourage this initiative.
7th February 2012
Chris Gale (UCL)
Creating a new Output Area Classification
The current Output Area Classification (OAC) uses only 2001 Census data in its construction; meaning it has become increasingly outdated since its original release. With data from the 2011 Census due to be made available by the end of 2012 plans for a new OAC are well underway. The new OAC will take advantage of not only 2011 Census data but other open source data sources that are now freely available. A new OAC will be a free, open source classification with a fully published methodology.
21st February 2012
Richard Goodman (Intergraph)
Richard will be talking about the Geomedia Professional Suite of GIS Software, highlighting its features and giving examples and case studies of where it has been deployed.
5th March 2012
Robert Brown (Remote Sensing Analyst, BMT Argoss, Atyrau, Kazakhstan); Angharad Stone (Web GIS Officer, Environment Agency); David Tong (GIS Consultant/ Technical Solutions lead for Central Government, ESRI [UK]); Kay Pallaris (Programme Manager - Systems & Technology Department, Corporate Services at the Olympic Delivery Authority) and Chris Holcroft (Director & CEO, AGI)
AGI Environmental Special Interest Group Seminar
Members of the Environmental Special Interest Group of the AGI (Association for Geographic Information) will form a panel to answer your questions on career options related to Geographic Information, based on their own experiences. From a variety of backgrounds, panel members will give a brief summary of their career, the route that brought them into GI-related work, and the skills which helped them get there. The panel will then try to answer your questions on pathways to GI related work.
13th March 2012
James Haworth (UCL)
A kernel based approach for spatio-temporal modelling and forecasting
Traditionally, statistical models have been used for spatio-temporal forecasting due to their strong theoretical foundation and interpretability. However, many large scale spatio-temporal datasets display complex, nonlinear, nonstationary properties that violate the iid assumptions of classical statistical models. Increasingly, practitioners are borrowing techniques from the machine learning community because of their innate ability to deal with this type of data. In particular, kernel based approaches such as the support vector machine have been successful because they use the so called “kernel trick” to allow linear algorithms to model nonlinear data. In this session, a kernel based approach to spatio-temporal forecasting is introduced. The model is tested using travel time data collected by automatic number plate recognition cameras on London’s road network.
20th March 2012
Garavig Tanaksaranond (UCL) and Adel Bolbol (UCL)
Visualization of traffic in space-time
Traffic congestion has many negative effects on people in the city. Although large amounts of road traffic data have been collected, we still have not been able to fully understand traffic congestion. The task is becoming still more difficult because traffic data are in effect too abundant and are highly disaggregate with respect to space and time. The research reported in this presentation seeks to use visualization techniques to understand road network performance in space-time, using data gathered from monitoring devices in London.
Page last modified on 16 nov 15 16:16