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

Deadline 31st August 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 here. or below:

 

Start Date: October 2017

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 September 2017. Any incomplete applications will not be considered.

 

PhD studentship in Extreme Citizen Science – development of data collection tools with non-literate participants

For more information on ExCiteS and ECSAnVis, please visit here and here.

Studentship Description

The ExCiteS research group has been developing Sapelli since 2012 – this is a platform that facilitates data collection across language or literacy barriers through highly configurable icon-driven user interfaces. The successful candidate will join a team of anthropologists, ecologists, geographers,  computer scientists and designers and focus on extending the undertaken research work. This will include the design, prototype and implementation of Sapelli components that answer the needs and wishes of participants in citizen science projects. The research will mainly include aspects of data collection such as data validation, user authentication and designing user interfaces for non-literate participants. We will specifically focus on engagement of non-literate people and we need to understand how the process, from data collection to analysis, can be made meaningful and useful for their everyday life. The project will include working with non-literate forest communities in central Africa and seek to enable these vulnerable communities to conduct their own environmental monitoring or mapping. For more information, see papers by Stevens et al. 2014 (http://discovery.ucl.ac.uk/1431647/) and Vitos et al. 2017 (http://dl.acm.org/citation.cfm?id=2998242).

Person Specification

The applicants should possess a good honours degree (1st Class or 2:1 minimum) in a relevant discipline such as Computer Science, Electronic Engineering or Human-Computer Interaction. The ideal candidate should also hold an MSc in Computer Science, Human Computer Interaction, Human Factors, or Geographical Information Science and must have proven experience and skills in Software Engineering or Human Computer Interaction.

The ideal candidate should have excellent verbal and written communication skills and should be able to work as part of a team to design, develop and deploy software prototypes. Preferably, the candidate should have experience in programming in Java and Android and have a good understanding of code management systems such as GitHub. Skills in using Python would also be desired. Finally, since the case studies might involve travelling to central Africa, the candidate should have a willingness to travel in remote forest locations and thus French language skills would be desirable. The ability to be patient and understanding with local people is essential in this regard.

Eligibility

Applications are invited from UK and EU citizenship holders.

Start Date

Between September and December 2017, at time that is suitable to the candidate. A successful candidate will be asked to work with the Extreme Citizen Science group for two weeks (expenses covered) before registering to the PhD programme.

Application Procedure

Applicants should send a cover letter, 2-4 page research proposal, examples of academic writing (e.g. BSc or MSc dissertation), code outputs from past work and CV to Michalis Vitos (michalis.vitos@ucl.ac.uk) and Judy Barrett (judy.barrett@ucl.ac.uk). The cover letter should include a personal statement explaining your interest in citizen science, why you are interested in our project and how you would see your work integrated into ExCiteS vision and activities. You are welcome to contact Prof Muki Haklay (m.haklay@ucl.ac.uk) or Michalis or any other ExCiteS members with queries about your proposal or for an informal conversation. Deadline for application – 31st August 2017.

Funding Notes

Duration – 3 years

Funding –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.

Funding Body: European Research Council (ERC) and UCL.

Cambridge Conference 2017 – The Willing Volunteer

wp-1499185668092The Cambridge Conference is an event that is held every 4 years, organised  by the Ordnance Survey, and it is a meeting of many heads of National Mapping Agencies who come together to discuss shared interests and learn from each other.

The history of the conference is available here. This year, I was asked to provide a talk about volunteered geographic information and the role of crowdsourced information in the service of national mapping bodies. As common in these conferences, I was given a title for the talk and request on the topic – this was “The Willing Volunteer –
Incorporating voluntary data into national databases” – and the description was: At present few mapping databases contain crowd sourced or voluntary data. Consider how, in the future, this will be a valuable source of data for national geospatial, cadastral and mapping agencies.

The talk itself covered 4 parts – since the conference as a whole looked at the future needs of mapping in the next 15 years, I’ve mentioned the trends that will influence crowdsourcing over this period. I’ve included both the technical and the social trends that will influence this area. I then covered few examples, and paid attention to the need to think differently about crowdsourced information (using the metaphor of scarcity/abundance as a way to explain that), then provided two insights from the “crowdsourcing geographic information in government” study that I’m currently leading. I’ve finished with few slides that demonstrate that engagement can reach out to everyone, regardless of their literacy.

Here are the slides:

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!

Into the night – training day on citizen science

dscn1936Last December, the Natural Environment Research Council (NERC) awarded funding to UCL Extreme Citizen Science group and Earthwatch as part of their investment in public engagement. The projects are all short – they start from January to March and included public engagement and training to early career researchers.

“Into the Night” highlights the importance of light pollution, a growing environmental stressor to both wildlife and humans, through collaborative and co-designed citizen science research. The project aims to increase awareness of this issue through two public workshops exploring the potential of two citizen science focal points – glow-worms and human wellbeing – explicitly linking ecological and human impacts. The project will culminate with a set of public activities (pilot data collection and educational) to coincide with Earth Hour (25.03.2017).

The project aims to build public engagement capacity through PhD internships with Earthwatch (Europe), CEH, Natural England and UCL, and forms a dedicated training day on the design and implementation of citizen science for 50 early-career researchers and PhD students.

The project is led by UCL (in collaboration with North Carolina State University – NCSU) and Earthwatch, bringing together leading research and practice in citizen science. It is the result of two co-design workshops, with over 30 participants from environmental science, social science, public health, National Parks, and NGOs. Based on this preparatory work, and with active training of early career researchers, we will run two focused workshops which will take place in dark sky reserves. These workshops will focus on two preliminary ideas for citizen science projects: a countrywide survey of glow-worms and the impact of artificial light on their activities, and the influence of lightscapes and dark green spaces on human wellbeing while balancing safety and concerns.

The two projects will generate public awareness and provide the public with opportunities to have debate and dialogue on the subject, as well as involvement in data collection and analysis. Results will be shared through social and traditional media. The outcome will advance ideas for a national citizen science project, which UCL and Earthwatch will take forward.

The training day run in Oxford on the 2nd February and during the day I gave two 45 minutes sessions. First, I provided an introduction to the field of citizen science, how to design a project, and how to evaluate such a project.

The session provided a brief overview of the types of citizen science that are relevant in addressing environmental challenges. We looked at classifications of citizen science projects, explore their potential goals, the process of recruitment and retention as well as the need to start project evaluation from an early stage. At the end the participants engage in a short exercise to consider how these elements can be used in the design of a citizen science project.

The second talk focused on technology.

The talk aim was described as follows: Current citizen science seems effortless…just download an app and start using it. However, there are many technical aspects that are necessary to make a citizen science project work. This session provided an overview of all the technical elements that are required – from the process of designing an app, to designing and managing a back-end system, to testing the system end to end before deployment. Again, at the end of the session, a short exercise considered the design of an app for a citizen science project that addresses light pollution.

 

Podcast – discussion with Liz Killen and Alice Sheppard on citizen science

Several weeks ago, Liz Killen, who is studying for an MSc in Science Communication at Imperial College organised an interview with me and Alice Sheppard about aspects of citizen science, for the I, Science the science magazine of Imperial College. This is the second time ExCiteS is covered in the magazine, after a report in 2013 by .

Crowdsourcing the Future?

About a month ago, on 7th December 2016, DR Kingsley Purdam (Manchester) organised a one day workshop on citizen science, and in particular on citizen science from a social science methodological perspective. The day organised with the support of the National Centre for Research Methods (NCRM).

The purpose of the workshop/conference was to explore the future of citizen science and citizen social science methods as research tools. In particular, understanding the different types of applications, methods, the data and the challenges posed. Because the point of view was based on methodologies in social sciences, issues about expertise, divisions of labour, different ways of seeing, data quality, questions about what might still be going undocumented and the ethical issues raised were all discussed.

The workshop was structured around two blocks of discussion – the morning around methods, data and ethics, while the after looked at issues of participation and working in the area of policy, as well as a discussion of the specific issues that need to be discussed for a citizen social science project.

As an introduction, both natural science and social science projects were presented. You can find a summary on twitter of some of the points that came up during the day with the hashtag #crowdfutures.

Some of the important tweets are captured here with comments (bit storify style).

Chris Lintott started the day with a discussion of large-scale, online citizen science projects, with the story of Zooniverse.

People participate in Zooniverse because they want to do something useful, and he pointed to the complexities of combining machine learning with citizen science effort while maintaining motivation and interest.

While I presented after Chris, and mostly talked about a more social theory explanation of what Extreme Citizen Science is – in particular, the creation of technologies that are embedded with a social participatory process. Many of the processes that I described were small-scale, and local. I have also pointed to the growth of citizen science and the Doing It Together Science project that we currently run.

However, in the discussion that followed we agreed that the nature of participation and many of the issues that come in these projects are similar across the scales even if the mechanisms for engagement are different.

Ben Rich (BBC), covered issues of engagement in weather observation that the BBC implemented successfully, with million observations and report in the first year

Hilary Geoghegan (Reading) & Alison Dyke (SEI)  talked about the UK EOF study on the motivation of participants and the ethics of participation, as well as the tensions between contributory and co-created citizen science in environmental research.

Will Dixon (Manchester) described the Cloudy with Pain project which engaged 12,000 participants and receives substantial information. The project also experiments with some access to data and opportunity for analysis by the participants themselves.

Kingsley Purdam (Manchester) talked about the complexity of citizen social science about begging, when the beggars are involved in data collection. Another Manchester-based project looked at linguistic diversity in street signs

The next set of talks raised some important point, including by Erinma Ochu on the process of creating the Robot Orchestra as a participatory DIY electronic and creative process, raising issues about expertise and success (the orchestra is in very high demand); Monika Buscher (Lancaster) emphasising that citizen social science is not about bigger torch to understand reality, but critique science & social science; and Alex Albert (Manchester), who run  project to encourage citizen reporting of empty houses and consider what should be done with them, highlighted the challenges of starting a project and recruiting participants. Liz Richardson (Manchester), talked about the interface between participatory action research and citizen science, and described her work with a community who collected data and asked for guidance on how to analyse it. The three talks by Monika, Alex, and Liz raised many issues about the participation of people in different stages of the research process, and the role of established researchers in such projects.

The last set of talks focused back on ecological and medical projects: Rachel Webster of Manchester Museum explained the museum digitising effort, and how they are making progress one MSc in computing student at a time – the integration of citizen science with small museum activities is a resource challenge, so the work with students require some compromises. There was also a demonstration of setting systems for citizen science and then discovering how they are used:

Lamiece pointed that a challenge with such approach is to get the app downloaded and to see continued use, although so far there are 1500 participants, 800,000 observations. There is also Data challenge of presence/absence reporting to make sense of what the data means.

Ian Thornhill fro mEarthWatch who coordinates the FreshWater Watch project demonstrate how simple data collection tools open up space for participant’s innovations in tools and in data collection. He also provided different models of how projects are run – corporate sponsorship, or by payment from interested communities.

Some of the points in the discussion include the need to balance scientific data collection and activism (especially for projects such as those that Liz Richardson described). Also balancing small scale, deep engagement or large datasets, wide engagement – e.g. for 3 years as researchers on projects that got limited funding and a goal. The need to consider what participation is doing to citizen science, and what science is doing to it? How to balance between the two? and in general, the wider societal impacts of projects cannot be ignored. There are also people that coming from a policy perspective, and try to push for procedural aspects, not interested in engagement issues.

There are also ethical issues such as those that relate to volunteer management – what should be done with contributors that are not doing good work? exclude them? train? ignore them? There is a constant need to think of useful roles and how making people valued for their contribution.

Another set of questions explored what citizen social science does to science? How are issues about ownership,  responsibilities to ensuring data quality integrated into project planning and management?