PhD studentship in collaboration with the Ordnance Survey – identifying systematic biases in crowdsourced geographic information

Deadline 10th November 2017

UCL Department of Geography and the Ordnance Survey are inviting applications for a PhD studentship to explore the internal systematic biases in crowd-sourced geographic information datasets (also known as Volunteered Geographic Information – VGI).

The studentship provides an exciting opportunity for a student to work with Ordnance Survey on understanding the use of crowd-sourced geographic information MH DSCN0571and potentially contributing to the use of such data sources by national mapping agencies. Ordnance Survey is an active partner in its sponsored research and offers students opportunities to work on-site and to contribute to workshops and innovation within the business. In addition, the student will be part of the Extreme Citizen Science group at UCL, which is one of the leading research groups in the area of crowdsourced geographic information and the study thereof.

For more information about the project, the studentship and details how to apply, please see below:

 

Start Date: January 2018

Funding status: Applications are invited from UK and EU citizenship holders.

Funding Body: EPSRC and Ordnance Survey

Funding Details: The scholarship covers UCL student fees at the Home/EU rate and provides a stipend of £16,553 per annum tax free. Travel expenses and research equipment will also be provided to the successful candidate.

Project Description:

UCL Department of Geography and the Ordnance Survey are inviting applications for a PhD studentship to explore the internal systematic biases in crowd-sourced geographic information datasets (also known as Volunteered Geographic Information – VGI).

There has been a rapid increase in information gathered by people from all walks of life who are using connected devices with an ability to collect and share geographic information, such as GPS tracks, photographs with location information, or observations of the natural environment in citizen science projects. There is now a vast array of projects and activities that use this type of information, and each project has its own characteristics. Yet, it can be hypothesised that some of the characteristics of crowd-sourced geographic information will be systematically biased, and these biases differ between projects and data sources.

Crowd-sourced datasets will have some systematic biases that repeat across crowd-sourcing platforms. For example the impact of population density, business activity, and tourism on the places where data is available, or a weekend or seasonal bias of the temporal period of data collection. Others biases are project-specific – for example, some projects manage to attract more young men, and therefore places that are of interest to this demographic will be over-represented. One of the major obstacles that limit the use of such data sources is understanding and separating systematic and project-level biases and then developing statistical methods to evaluate their impact. In order to use such datasets to identify hidden features and patterns, there is a need to identify what are the relationships between a dataset and the world.

The aim of this research project, therefore, is to create a large collection of crowd-sourced GPS tracks and pedestrian trajectories, and use conflation techniques and advanced analytics to develop methodologies to identify and estimate the biases. Once this is done, the aim will be to identify hidden characteristics to be more confident about the patterns that are being observed.

Studentship Description

The studentship provides an exciting opportunity for a student to work with Ordnance Survey on understanding the use of crowd-sourced geographic information, and potentially contributing to the use of such data sources by national mapping agencies. Ordnance Survey is an active partner in its sponsored research and offers students opportunities to work on-site and to contribute to workshops and innovation within the business. In addition, the student will be part of the Extreme Citizen Science group at UCL, which is one of the leading research groups in the area of crowdsourced geographic information and the study thereof.

The project will run for four years and will be supervised by Prof Muki Haklay from UCL and Jeremy Morley from Ordnance Survey. Professor Muki Haklay, who is a professor in the UCL Department of Geography and who has a track record of research and publication relating to crowdsourced data management and quality. Jeremy Morley is the Chief Geospatial Scientist at Ordnance Survey, leading the long-term business research programme, and has research experience in crowd-sourced geographic information.

 Person Specification

Applicants should possess a strong bachelor’s degree (1st Class or 2:1 minimum) or Masters degree in Computer Science, Spatial statistics, Ecology, Geomatics, Geographic Information Science or a related discipline. The skills required to build the required database of case studies and the programming and analytical skills to assess biases and develop algorithms for their identification, are highly desirable. Candidates will ideally have some relevant previous research experience and should also have excellent communication and presentation skills.

The funding is provided for 4 years, and will involve spending time at the Ordnance Survey in Southampton.

Eligibility

Applications are invited from UK and EU citizens residing in UK. In particular, applicants must meet EPSRC eligibility and residency requirements found here:

http://www.epsrc.ac.uk/skills/studentships/help/eligibility/

Application Procedure

Applicants should send the following by e-mail to Judy Barrett (judy.barrett@ucl.ac.uk) and Prof Haklay (m.haklay@ucl.ac.uk):

  1. Cover letter, including a personal statement explaining your interest in the project.
  2. Examples of academic writing and outputs from past work (e.g. a dissertation or assignment)
  3. Academic transcripts
  4. A CV

Shortlisted applicants will be invited to interview during November 2017. Any incomplete applications will not be considered.

 

Advertisements

Citizen Science 2017 – workshops day and opening panel

The Citizen Science Association conference is held at the River Center in St Paul, Minnesota on 17th to 20th May. This post and the following ones are notes that were taken during the meeting in the sessions that I’ve attended.

Wednesday was dedicated to workshops, and I joined the Citizen Science at College level workshop. Organised by Thomas Tisue (Muskegon Community College); John R. Jungck (University of Delaware); Aerin W. Benavides (University of North Carolina); Julie Feldt (Adler Planetarium); Colleen Hitchcock (Brandeis University); Leslie Ries (Georgetown University); and Terry A. Gates (North Carolina State University). The workshop aim was to bring together academics who work with undergraduate students to discuss best practices for developing citizen science research within their university classes.

Some ideas about citizen science at undergraduate level – it can be about enculturating students with the concepts of democratisation of science, the value of open science and education. It also brings up issues of data quality, and understand connections beyond their discipline. There is plenty of opportunities for experiential learning. The issue is that move from a closed process that it is evaluated by the instructor to one that they are evaluated by peer and even participants.  The students, of course, have different motivation to participate in citizen science.

John Jungck – philosophically, citizen science can be thought differently including different levels of engagement and about how it fit with societal goals. Think that each assignment is for a social good. Learning skills for the 21st century require developing citizenship skills, and investigation in the field is assisting in the development of issues in physics, biology, and mathematics.

Colleen Hitchcock is seeing the view of citizen science as an integral to much of the studies, and doing things like phenology on campus can increase bio-literacy and understand changes. In every class that she teaches there is an element of citizen science. The way to allow students to engage in research is to join an existing project as a way to save resources. The assignments include – what is citizen science – reflect on the experience, explore SciStarter, learn through a contributory project. Citizen Science can assist in enhancing classes such as about Climate Change. It also provides an opportunity for professional development.

Leslie Ries – she looks at research in the classroom, and instead of running a programme, joining an existing one. Looking at butterfly at continental scale as her research area, and this allows for well-curated data sets – butterfly informatics that provides data. She integrated the module into existing course to introduce students to this is to teach informatics in ecology and where the data come from. She now got a module that starts in a lecture that explains history, needs for large scale data and then citizen science as a source for that. The question development about being able to develop a question and focusing on butterfly ecology and develop question and acquire data

Aerin W. Benavides talked about the value of citizen science for project-based learning – it provides an opportunity for exploration that is missing in the previous schooling. This leads to teaching teachers about the potential of citizen science.

Thomas Tisue – in community colleges there is a need to help STEM students who are looking for research opportunity and linking that to citizen science group that want to learn to education and outreach, and considering the limited resources of the college open up an opportunity. In community colleges, students are many time first generations and lack context of study, and sometime need financial support to allow them to participate in an internship programme with local environmental monitoring. The faculty need to be involved and assure integrity. College can offer credit through independent study time.

Julie Feldt – at Adler and work on zooniverse. Different opportunities – educators are sharing material to use projects in teaching, with examples from middle school, high school, and college introduction. They have done access to Galaxy Data on Google Drive to allow students to use these data to examine information with Google spreadsheet

Bucky Gates – the students go on SciStarter, doing a project, and then fill in a review form that was designed with SciStarter, and got over 500 forms completed, and that helps in assessing which projects work and how people who are volunteered to a project react to different projects. Projects need to be simple – and creative. Can make data collection simple so it is malleable to different areas and allow an opening for creativity.

In different breakup sessions, participants explored 6 teams: analysing citizen science data; supporting pre-service and in-service educators (teachers training); independent research – supporting students; learning within the semester; using citizen science project in the classroom; and an open one.

The using citizen science in the classroom group highlighted the need to simplify and focus on what possible to get. Challenges of teaching in one semester – reaching out to mailing lists, creating more collaborative/co-created than just contributory and who to partner around college on a specific citizen science. Supporting educators – citizen science is not visible in museums, science centres etc. Analysing data – issues of developing different ways. Independent learning – learning from UK, Chile, USA. Social engagement is an important part of citizen science and is it suitable to expect students to come up with a question or join someone’s question.

There is a growing recognition of the need to have introductory material on where to start and which system, project, and platform to use. There are resources such as Studentsdiscover.org that provide information for teacher to get into citizen science – mostly to middle school

The first plenary event of the conference was “Meet the Authors with Darlene Cavalier (ASU): The Rightful Place of Science: Citizen Science; Caren Cooper (NCSU): Citizen Science: Changing the Face of Discovery; and Mary Ellen Hannibal: Citizen Scientist: Searching for Heroes and Hope in an Age of Extinction. The moderator was Heather McElhatton, MN Public Radio.

Darlene Cavalier talked about her “The Rightful Place of Science: Citizen Science”. Darlene defined citizen science as a way to advance science without a formal degree, or simply science. Darlene described her journey into citizen science: from journalism to science communication, and technology assessment. Darlene also explains the link to Science Cheerleaders and the way it works together and allows to promote citizen science project. The name if the book came from a series by ASU. The selection of chapters that were included in this book was done in order to keep the book cheap and to ensure that the process is manageable while being written by passionate experts. The different chapters are exploring the link to policy; definitions of citizen science by Booney and Irwin and the tension in them; How citizen science can be linked to teaching in class situations; media aspects of citizen science – but there are situations where citizen science and citizen journalism is getting close. Darlene also explained the role of SciStarter – different ways for getting involved in citizen science and giving multiple routes that will allow people to join in. The chapter on citizen science in community citizen science – and how it is linked to concerns of the community. The final chapter is demonstrating how citizen science can engage with people in exploring microbiome in the international space station, with people also analysing the data, and the story of how the project came about.  The book ends with “now it’s time for you to explore citizen science”. The challenge of the book is to open citizen science to many audiences – truly everybody that is curious can participate in a project about their concerns.

Caren Cooper combined her curiosity, and the need to become scientist in order to engage with nature in a serious way. Once she had kids, doing field work wa less possible, so she started collaborating with volunteers. It’s important to acknowledge citizen science, as it contributed to science but also pointing to the limitation of main science – of things that scientists just cannot achieve alone. Covering the history of science. Caren identify the smartphone as very handy tool in influencing ability to collect and share the data. The purpose of her book is to demistify science and make it accessible to people – it’s collective activity where everyone is giving a little bit, and collectively, it’s a feast. Caren was surprised of the range of disciplines and fields that involving people and the different ways in which it is happening. There are some great stories of community transformation in the book about community action of plastic bags following turtle monitoring, to engagement of prison inmates in dealing with entomology research. The take away – citizen science should become the new norm in science and life.

Mary Ellen Hanibal, brought a new concept to audiences. Her journey began 10 years ago from a book about evolution. In the California Academy she explored  with taxonomies, and she understood the concept of sixth extinction that is happening and she started to look at conservation biology, and that led to understand big data, and citizen science within this. The word “hope” in the title is not what she wanted, and want to see action and having a heroe’s journey. She also explored the need to act, to think about concepts like half earth and be aware of the emergency of saving species. Citizen science is a platform to bring people together and make people come togehter. She want people to reconsider the plae of humans in the circle of lifes – it’s part of a journay of life and we need to support other life form and find a way to do it all together.

More on Darlene book here and Mary Ellen book here. I’m in the middle of Caren’s book and hope to write about it soon!

The Potential of Volunteered Geographic Information (VGI) in Future Transport Systems

dsc01541An aspect of collaborative projects is that they start slowly, and as they become effective and productive, they reached their end! The COST Energic (European Network for Research into Geographic Information Crowdsourcing) led to many useful activities, with some of them leading to academic papers. From COST Energic, we’ve got the European Handbook on Crowdsourced Geographic Information, a paper on VGI quality assessment methods, and more.

One outcome came out from the close collaboration around the summer schools that were organised by the network. Prof Cristina Capineri was the chair of the COST network, and also the organiser of summer schools in Fiesole, near Florence. Prof Maria Attard organised the other summer school of the action, at the University of Malta. Based on our close working relationships (though Maria and I know each other since our PhD studies in CASA) we started working on a joint paper. Maria specialises in transport geography, so the support from COST Energic was a reason to consider how VGI will play out in future transport systems. The paper was published in the journal Urban Planning and the abstract reads:

“As transport systems are pushed to the limits in many cities, governments have tried to resolve problems of traffic and congestion by increasing capacity. Miller (2013) contends the need to identify new capabilities (instead of capacity) of the transport infrastructure in order to increase efficiency without extending the physical infrastructure. Kenyon and Lyons (2003) identified integrated traveller information as a facilitator for better transport decisions. Today, with further developments in the use of geographic information systems (GIS) and a greater disposition by the public to provide volunteered geographic information (VGI), the potential of information is not only integrated across modes but also user-generated, real-time and available on smartphones anywhere. This geographic information plays today an important role in sectors such as politics, businesses and entertainment, and presumably this would extend to transport in revealing people’s preferences for mobility and therefore be useful for decision-making. The widespread availability of networks and smartphones offer new opportunities supported by apps and crowdsourcing through social media such as the successful traffic and navigation app Waze, car sharing programmes such as Zipcar, and ride sharing systems such as Uber. This study aims to develop insights into the potential of governments to use voluntary (crowdsourced) geographic information effectively to achieve sustainable mobility. A review of the literature and existing technology informs this article. Further research into this area is identified and presented at the end of the paper.”

The paper is open, and can be found here

Published: Why is Participation Inequality Important?

bookcoverI’ve mentioned the European Handbook for Crowdsourced Geographic Information in the last post, and explained how it came about. My contribution to the book is a chapter titled ‘Why is Participation Inequality Important?. The issue of participation inequality, also known as the 90:9:1 rule, or skewed contribution, has captured my interest for a while now. I have also explored it in my talk at the ECSA conference on ‘participatory [citizen] science‘ and elsewhere.

In this fairly short chapter what I am trying to communicate is that while we know that participation inequality is happening and part of crowdsourced information, we need to consider how it influences issues such as data quality, and think how it come about. I am trying to make suggest how we ended with skewed contributions – after all, at the beginnings of most projects, everyone are at the same level – zero contribution, and then participation inequality emerge.

I have used the iconic graph of contribution to OpenStreetMap that Harry Wood created, but the chapter is discussing other projects and activities where you can come across this phenomena.

Here is a direct link to the chapter, and I’ll be very happy to hear comments about it!

 

New book: European Handbook of Crowdsourced Geographic Information

COST EnergicCOST ENERGIC is a network of researchers across Europe (and beyond) that are interested in research crowdsourced geographic information, also known as Volunteered Geographic Information (VGI). The acronym stands for ‘Co-Operation in Science & Technology’ (COST) through ‘European Network Researching Geographic Information Crowdsourcing’ (ENREGIC). I have written about this programme before, through events such as twitter chats, meetings, summer schools and publications. We started our activities in December 2012, and now, 4 years later, the funding is coming to an end.

bookcoverOne of the major outcomes of the COST ENERGIC network is an edited book that is dedicated to the research on VGI, and we have decided that following the openness of the field, in which many researchers use open sources to analyse locations, places, and movement, we should have the publication as open access – free to download and reuse. To achieve that, we’ve approached Ubiquity Press, who specialise in open access academic publishing, and set a process of organising the writing of short and accessible chapters from across the spectrum of research interests and topics that are covered by members of the network. Dr Haosheng Huang (TU Wien) volunteered to assist with the editing and management of the process. The chapters then went through internal peer review, and another cycle of peer review following Ubiquity Press own process, so it is thoroughly checked!

The book includes 31 chapters with relevant information about application of VGI and citizen science, management of data, examples of projects, and high level concepts in this area.

The book is now available for download hereHere is the description of the book:

This book focuses on the study of the remarkable new source of geographic information that has become available in the form of user-generated content accessible over the Internet through mobile and Web applications. The exploitation, integration and application of these sources, termed volunteered geographic information (VGI) or crowdsourced geographic information (CGI), offer scientists an unprecedented opportunity to conduct research on a variety of topics at multiple scales and for diversified objectives.
The Handbook is organized in five parts, addressing the fundamental questions:

  • What motivates citizens to provide such information in the public domain, and what factors govern/predict its validity?
  • What methods might be used to validate such information?
  • Can VGI be framed within the larger domain of sensor networks, in which inert and static sensors are replaced or combined by intelligent and mobile humans equipped with sensing devices?
  • What limitations are imposed on VGI by differential access to broadband Internet, mobile phones, and other communication technologies, and by concerns over privacy?
  • How do VGI and crowdsourcing enable innovation applications to benefit human society?

Chapters examine how crowdsourcing techniques and methods, and the VGI phenomenon, have motivated a multidisciplinary research community to identify both fields of applications and quality criteria depending on the use of VGI. Besides harvesting tools and storage of these data, research has paid remarkable attention to these information resources, in an age when information and participation is one of the most important drivers of development.
The collection opens questions and points to new research directions in addition to the findings that each of the authors demonstrates. Despite rapid progress in VGI research, this Handbook also shows that there are technical, social, political and methodological challenges that require further studies and research

 

Esri survey123 tool – rapid prototyping geographical citizen science tool

There are several applications that allow creating forms rapidly – such as Open Data Kit (ODK) or EpiCollect. Now, there is another offering from Esri, in the form of Survey123 app – which is explained in the video below.

Survey123 is integrated into ArcGIS Online, so you need an ArcGIS account to use it (you can have a short experiment if you register for a trial account, but for a longer project you’ll have to pay). The forms are configured in XForms, like ODK . The forms can be designed in Excel fairly quickly, and the desktop connection package make it easy to link to the Survey123 site, as well as testing forms.  I tried creating a form for local data collection, including recording a location and taking an image with the phone. It was fairly easy to create forms with textual, numerical, image and location information, and the software also supports the use of images to items in the form, so they can be illustrated visually. The desktop connector application also allow use to render the form, so they can be tested before they are uploaded to ArcGIS Online. Then it is possible to distribute the form to mobile devices and use them to collect the information.

The app works well offline, and it is possible to collect multiple forms and then upload them all together. While the application still showing rough edges in terms of interaction design, meaningful messages and bug clearing, it can be useful for developing prototypes and forms when the geographic aspect of the data collection is central. For example, during data collection the application supports both capturing the location from GPS and pointing on a map to the location where the data was collected. You can only use GPS when you are offline, as for now it doesn’t let you cache a map of a study area.

As might be expected, the advantage of Survey123 is coming once you’ve got the information and want to analyse it, since ArcGIS Online provide the tools for detailed GIS analysis, or you can link to it from a desktop GIS and analyse and visualise the information.

Luckily for us, Esri is a partner of the Extreme Citizen Science group and UCL also holds an institutional licence for ArcGIS Online, so we have access to these tools. However, through Esri conservation programme can also apply to have access to ArcGIS Online and use this tool.

Citizen Science and Policy – possible frameworks

Back in February, my report ‘Citizen Science & Policy: a European Perspective‘ was published by the Wilson Centre in the US. As I was trying to make sense of the relevance of citizen science to policy making, I used a framework that included the level of geography, area of policy making and the type of citizen science activity. This helped in noticing that citizen science is working well at the neighbourhood, city and national scales, while not so well at regional and international level. The reasons for it are mostly jurisdiction, funding and organisational structure and scale of operation.

Later on, at a workshop that was organised by Prof Aletta Bonn on Citizen Science and Policy impact, the idea of paying attention to the role of citizen science within the policy cycle was offered as another dimension of analysis.

Last week, at a workshop that was organised by the European Environment Agency (EEA) as part of their work on coordinating the European Protection Agencies (EPA) Network, I was asked to provide an introduction to these frameworks.

The presentation below is starting with noting that citizen science in an EPA is a specific case of using crowdsourced geographic information in government and some of the common issues that we have identified in the report on how governments use crowdsourced information are relevant to citizen science, too. Of particular interest are the information flows between the public and government, and the multiple flows of environmental information that the 3rd era of environmental information brought.

After noticing the individual, organisational, business and conceptual issues that influence use in general, I turn to the potential framing that are available – geography, stage in policy formation and mode of engagement, and after covering those I’m providing few examples of case to illustrate how specific cases fit into this analysis.

It was quite appropriate to present this framework in the EEA, considering that the image that was used to illustrate the page of the report on the Wilson Center site, is of the NoiseWatch app which was developed by the EEA…