The Open Education Working Group: Bringing People, Projects and Data Together

Abstract

New areas of research and discovery benefit from a space in which a community can discuss key questions, interact and be allowed to grow. For the open education data community these questions are still crude and fundamental. For example: what exactly is open education data, why is open data relevant to education and how can we use open datasets to meet educational needs? The Open Education Working Group inaugurated by Open Knowledge, a community-based not-for-profit organisation that promotes open knowledge in the digital age, is beginning to explore the relationship between data and education by exposing projects and bringing together people working on related activities. This chapter will consider initial discussions and explorations related to open data in education that have taken place through recent online activity and workshops. It will also look at the creation of a working group established with the intention of taking these discussions further.

Keywords

1 Introduction

Open data in education is a relatively new area of interest with only dispersed pockets of exploration having taken place worldwide, these initial explorations will be covered later in this chapter. The phrase ‘open educational data’ remains loosely defined but might be used to refer to:

• all openly available data that could be used for educational purpose

• open data that is released by education institutions

Understood in the former sense, open educational data can be considered a subset of open educational resources (OERs) where datasets are made available for use in teaching and learning. These datasets might not be designed for use in education, but can be repurposed and used freely.

In the latter sense, the interest is primarily around the release of data from academic institutions about their performance and that of their students. This could include:

• Reference data such as the location of academic institutions

• Internal data such as staff names, resources available, personnel data, identity data, budgets

• Course data, curriculum data, learning objectives,

• User-generated data such as learning analytics, assessments, performance data, job placements

• Benchmarked open data in education that is released across institutions and can lead to change in public policy through transparency and raising awareness.

The World Economic Forum report on education and skills [5] sees there as being two types of education data: traditional and new. Traditional dataset include identity data and system-wide data, such as attendance information new datasets are those created as a result of user interaction, which may include web site statistics, and inferred content created by mining datasets using questions.

Whatever the classification it is clear that open education datasets are of interest to a wide variety of people including educators, learners, institutions, government, parents and the wider public. Some will have a passion for improving teaching and learning or have a vested interest in a particular individual’s education, while others will be mining data to influence policy decisions or exploring monetisation of datasets. Open education data holds huge potential for many and its exploration is both inevitable and necessary.

2 Establishment of an Open Education Working Group

One of the goals of the LinkedUp Project [3]¹, which focused on the exploitation and adoption of public, open data available on the Web by educational organisations, was to grow a community of linked data and open data practitioners that would continue to network with one another after the end of the project lifecycle. The expectation was that this community would continue to build applications, educate each other and others, and influence the future of linked and open data in an educational context.

In an effort to see discussions around open data in education pulled into the wider debate around open education, the LinkedUp Project dissemination partner, Open Knowledge², established the Open Education Working Group.³ Open Knowledge is a not-for-profit organisation that promotes open knowledge, including open content and open data. It provides open data services and is the creator and licensor of CKAN⁴, the worlds leading software for open data portals. Open Knowledge has been working in the open data space since 2004 and are world experts around open data, open content, principles, standards and practice in open. In 2006 Open Knowledge began work on the Open Definition⁵ which sets out principles that define “openness” in relation to data and content. The definition makes precise the meaning of “open” in the terms “open data” and “open content” and thereby ensures quality and encourages compatibility between different pools of open material. Its development and use has been key in the open movement. Open Knowledge and the Open Definition Advisory Council recently announced the release of version 2.0 of the Open Definition. The short version of the definition is given as: “Open data and content can be freely used, modified, and shared by anyone for any purpose”. The Definition “sets out principles that define openness in relation to data and content” and plays a key role in supporting the growing open data ecosystem.

Open Knowledge co-ordinates over 20 domain-specific Working Groups that focus on discussion and activity around a given area of open knowledge. The Open Education Working Group joins others looking at areas including open access, open science, open economics, open spending and open government data. The Open Education Working Groups goal is to initiate global cross-sector and cross-domain action that encompasses the various facets of open education, including open data. It brings together people and groups interested in the various facets of open education, from OERs and changing teaching practices, to licensing and emerging open policy. However it differs from other online discussion groups that deliberate on open education topics. This is because its goal is to seed and support trans-global, cross-sector and cross-domain activities and projects. The group also provides an opportunity for collaboration across organisations through engagement with existing groups.

3 Group Launch

The Open Education Working Group officially launched to a physical audience of over 100 in September 2013 at OKConference⁶ in Geneva at a LinkedUp Open Education Panel Session titled, The facets of open education (See Fig. 1).

Fig. 1.

OKCon Open Education Panel session. From left to right: Mathieu dAquin, Davide Storti, Jackie Carter and Doug Belshaw.

The panel session examined the different ‘faces’ of open education. The premise was that while many other open education groups exist their focus tends to be national and subject specific, considering one particular area of open education. This panel session, and in conclusion the working group itself, was an attempt to explore the synergies between different areas of open education. The panel was moderated by Doug Belshaw, Badges and Skills Lead, Mozilla Foundation. The panelists were:

• Jackie Carter, Senior Manager, MIMAS, Centre of Excellence, University of Manchester, who gave the Open Educational Resources perspective

• Davide Storti, Programme Specialist, Communication and Information Sector (CI), United Nations Educational, Scientific and Cultural Organization (UNESCO) who gave us the open practitioner perspective.

• Mathieu dAquin, Research Fellow, Knowledge Media Institute, Open University, UK, who is a LinkedUp Project team member gave the open data perspective.

A write up of the launch is available from the LinkedUp Project blog⁷. Slides from the session are available on Slideshare.⁸

4 Group Structure

The Open Education Working Group has taken the approach of building its governance and defining its member structure in consultation with the wider community. This happens through bi-monthly working group calls⁹ which are open to all and through use of a charter which sets out the member structure¹⁰. The group also participated in an open consultation process and appointed an Advisory Board which contains high-profile Open Education advocates who are experts in the field¹¹. The Advisory Board provides thought leadership about the direction of the working group and helps to raise the profile of the working group by talking about the group and its work at conferences and events.

The current Advisory Board has 6 members:

• Karien Bezuidenhout, Chief Operating Officer at the Shuttleworth Foundation

• Lorna M. Campbell, Assistant Director of the Centre for Education, Technology and Interoperability Standards

• Dr. Cable Green, Director of Global Learning at Creative Commons

• Joonas Mkinen, Finnish maths teacher carrying out exciting open text book work

• Bernard Nkuyubwatsi, initiator of the Open Education Rwanda Network

• Rayna Stamboliyska, founder of RS Strategy and OpenMENA

The Open Education Working Group website includes a blog, information on how to get involved and details of the mailing list and activities. The blog features regular guest blog posts including a series of Open Education Around the World blog posts currently covering 15 countries from Europe, Asia, Africa and America¹². There is also an online interactive open education timeline which capture events, projects and activities related to open education all around the globe. The group has an active mailing list¹³ and a Twitter feed¹⁴.

The Open Education Working Group aims to be more than just a place for discussion, current actions include support for LRMI initiatives¹⁵, standards, supporting a platform for open standards work, promotion of multilingualism for OERs, support for member activities, connections with local Open Knowledge groups (Belgium, Finland, Brazil) etc.

5 Open Education Working Group Events

Concrete activities initiated by the group have so far comprised of online discussions relating to the definition of open data in education, community webinars highlighting research and real-world activities and the collaborative writing of a living web document entitled the Open Education Handbook. We explore these activities in more detail below.

5.1 Open Education Smörgåsbord

The Open Education Working Group delivered its first official group workshop at the Open Knowledge Festival in July 2014 in Berlin¹⁶. The Open Education Smörgåsbord session¹⁷ featured 6 different tables of activity, from open badges and open data to OERs for teachers and Open education policy, provided by 8 Open Education Working Group members. The tables were facilitated by Kristina Anderson from Creative Commons Sweden Miska Knapek, an information experience designer from Denmark involved with Open Knowledge Finland Irina Radchenko, Associate Professor at the Higher School of Economics in Moscow Tom Salmon, a teacher and open development researcher Darya Tarasowa Darya, maintainer of SlideWiki.org Alek Tarkowski, director of Centrum Cyfrowe, Polish NGO focusing on open issues and Public Lead of Creative Commons Poland and European Policy Fellow with Creative Commons and Marieke Guy, the Open Education Working Group Co-ordinator. One of the activities was Tom Salmons wall of open data case studies (See Fig. 2).

Fig. 2.

Wall of open data in education case studies (CC BY) by Gregor Fischer (http://www.gfphotography.de/).

These covered a variety of different areas:

• AREA 1 USING DATA IN EDUCATION: Looking at use of open data to enrich teaching and learning, and to learn about data and data analysis within education or lessons.

• AREA 2 METADATA and OPEN EDUCATION RESOURCES: Looking at how metadata and open data can support education, for example through the use of OERs in universities and schools and the role of initiatives such as the learning resource metadata initiative (LRMI).

• AREA 3 GOVERNMENT DATA for EDUCATION: Looking at ways that different governments are making contributions with open data to improve education in Brazil, Holland, New Zealand and the UKl.

• AREA 4 LEVERAGING OPEN BADGES IN EDUCATION: Looking and learning about how open badges (which leverage different kinds of metadata) can be used to support and extend formal education, and personalise it in different ways.

• AREA 5 MOBILE LEARNING with OPEN DATA: Looking at how free, open source (FOSS) app authoring tools can be used to build apps that use open datasets with all kinds of applications.

In area 1 particular attention was paid to previous LinkedUp Competition entries, for example there was an explortation of how we can use open data to learn about data and data analysis within education or lessons¹⁸ or with specific focus issues (e.g. sustainability with Globetown¹⁹ or politics with Polimedia²⁰).

5.2 ‘Making It Matter’ Workshop

The Making it Matter workshop²¹ which used the subtitle ‘Supporting education in the developing world through open and linked data’, was held in central London on 16$^{th}$ May 2014. The workshop was organised by the LinkedUp Project and the Open Education Working Group in collaboration with associate partner The Commonwealth of Learning²².

The workshop had two real aims: firstly, it focused discussion around real-world requirements in the developing world that could be aided through the releasing of data and/or the building of relevant applications and prototypes. Secondly, it explored the opportunity to look at tools developed through the LinkedUp Challenge (the Veni and Vidi Competitions) and See how they could be used in the developing world.

The workshop was attended by approximately 30 delegates: teachers, educators, members of the open development movement, open data and linked data communities, developers and technologists. It combined talks, tool demonstrations and break-out group sessions. As the event focused on education in the developing countries, it was felt to be very important to make provisions for those who could not physically attend. Video for the entire day, including break out sessions, was streamed and shared afterwards. All breakout activities were carried out in online etherpads and a form was used to facilitate questions from remote participants.

Concrete outputs from the day were formulated around the answers to three questions:

• What real world problems are there related to education in the developing world that could potentially be solved by data and technology solutions?

• What data is out there and what data could be released to aid education in the developing world?

• Next Steps what are we going to do?

These discussion outputs fed into the requirements for a focused track for the LinkedUp Vici Competition that looked for educational applications that target developing countries.

5.3 Online Community Session

In May 2014 LinkedUp and the Open Education Working Group took part in an Open Knowledge Community Session entitled ‘What has open data got to do with education’? The session was facilitated by Heather Leson (community builder at Open Knowledge) with talks from Marieke Guy (Open Education Working Group, LinkedUp Project) and Otavio Ritter (Open education data researcher, Getulio Vargas Foundation, Brazil). Otavio shared his findings from work on paper involving a comparative analysis of school open data in England and Brazil and the availability (transparency) of government information related to primary/secondary education area. Video and slides are available online.²³

6 Open Data in Education Discussions

Through events and the mailing list the working group has supported dialogue and discussion around open data in education. These discussion topics have been explored using blog posts and in the Open Education Handbook. The main areas of dialogue have been: consideration of drivers for open data support for finding open data and analysis of use cases for open data. Elaboration on these discussion areas is given below.

6.1 Drivers for Open Data in Education

The current main drivers for open data use in education are principle, policy and practice. The charitable mission of education can be helped through a commitment to open data, enabling educators and institutions to engage with learners more effectively and in better ways. Data openness and exchange can drive quality research (collaboration, testing, replication) while promoting the social role and place of institutions themselves, helping maintain public and political commitment to the institution and making it more transparent. Education institutions are already subject to freedom of information, but new open research data policies (such as the UKs Higher Education Funding Council for England (HEFCE) consultation on inclusion as part of next Research Excellence Framework²⁴) may alter obligations. In the UK, for example, large amounts of institutional data (finance, student performance, etc.) are already collected by the Higher Education Statistics Agency (HESA) and the Universities and Colleges Admissions Service (UCAS) and made widely available, and this is a trend which can be observed in many countries. The next logical step is for more open data about institutions to be made available. With agreed frameworks and metrics in place it will be easier to substantiate comparisons and claims about widening participation, or student performance, for example.

Institutions can use their own data to inform decisions and management practices, and improve business and pedagogical intelligence. By linking across other open datasets and curating the most relevant information staff and students can be supported in teaching and learning.

6.2 Types of Open Data in Education

There are many different types of data that can be relevant to education and come from education. Relevant sources might include:

• Publications and literature: ACM, PubMed, DBLP (L3S), OpenLibrary, etc.

• Domain-specific knowledge and resources: Bioportal for Life Sciences, etc.

• Historic artefacts in Europeana, Geonames, etc.

• Cross-domain knowledge: DBpedia, Freebase, etc.

• (Social) media resource metadata: BBC, Flickr, etc.

Explicitly educational datasets and schemas include:

• University Linked Data: e.g. The Open University UK [1]²⁵

• Southampton University, University of Munster (DE), education.data.gov.uk, etc.

• OER Linked Data: mEducator Linked ER²⁶, OpenLearn, etc.

• Schemas: Learning Resource Metadata Initiative (LRMI²⁷, mEducator Educational Resources schema²⁸

• Learning Analytics and Knowledge (LAK) Dataset [2]²⁹

• Vast Open Educational Resource (OER) and MOOC metadata collections (e.g. OpenCourseware, OpenLearn, Merlot, ARIADNE)

• UK Key Information Set³⁰

• Education GPS is the OECD source for internationally³¹, as well as comparable data on education policies and practices, opportunities and outcomes. Accessible any time, in real time, the Education GPS provides you with the latest information on how countries are working to develop high-quality and equitable education systems.

There are also many different ways to categorise this data.

• Student data: attendance, grades, skills, exams, homework, etc.

• Course data: employability related to courses, curriculum, syllabus, VLE data, number of textbooks, skills, digital literacy, etc.

• Institution data: location data, success/failure rates, results, infrastructure, power consumption, location, student enrolment, textbook budget, teacher names and contracts, drop out rates, total cost of ownership, sponsorship, cost per pupil, graduation rates, male vs female, years in education, ratio of students to teaching staff, etc.

• User-generated data: learning analytics, assessments, performance data, job placements, laptop data, time on tasks, use of different programmes/apps, web site data, etc.

• Policy/Government data: equity, budgets, spending, UNESCO literacy data, deprivation and marginalisation in education, participation, etc.

Other approaches to categorisation have been suggested. Louis Coiffait, then at Pearson, offered the following categories in an exploratory presentation given at the Education Innovation conference in Manchester: type (similar to listed above), purpose (intentional use), level (regional, national, international etc.)³². Pearson Blue Skies were responsible for a report entitled “How Open Data, data literacy and Linked Data will revolutionise higher education”³³ written in 2011. Octavio Ritter, Open education data researcher, Getulio Vargas Foundation, Rio de Janeiro, Brazil, offered other categories: macro (education policy), meso (school management) and micro (student level).

In addition to information about open licensing, a more detailed description of an open dataset may include:

• Provenance

  • Reference (government data, geo-data, etc.) – e.g. national curriculum data

    • $*$ Location of schools, universities, etc.

  • Core/Internal (course catalogue, course resources, staff data, buildings, etc.)

  • User-generated/contributed (user activities, assessments, etc.)

• Granularity

  • individual/personal

  • aggregated/analytics

  • report

• Descriptiveness

  • data streams (multimedia resources)

  • data content (textual content, database)

  • resource metadata

  • content metadata

  • paradata (as in metadata about data collection)

• Content

  • Usage/activity data (paradata as in the learning analytics definition)

  • student personal information

  • student profiles (interest, demographics, etc.)

  • student trajectories

  • curriculum / learning objectives / learning outcomes

  • educational resources (multimedia or not)

  • resources metadata (including library collections, reading lists, Talis Aspire)

  • assessment/grades

  • institutional performance (e.g., ofsted, Key Information Set)

  • resource outputs (publication repositories, etc.), research management data (projects and funding, etc.), research data

  • cost and student funding data, budgets and finances

  • classifications/disciplines/topics (e.g. JACS)

6.3 Finding Open Data in Education

One good source of open data is governments, who increasingly make data about their citizens available online. Examples from the UK include school performance data³⁴, data on the location of educational establishments³⁵ and pupil absenteeism³⁶. There is also data from individual institutions such as that collated on linked universities³⁷ and on data.ac.uk ³⁸ and from research into education, such as the Open Public Services Network report into Empowering Parents, Improving Accountability.³⁹

Previously much of the release and use of open educational datasets has been driven by the need for accountability and transparency. A well-cited global example has been the situation in Uganda where the Ugandan government allocated funding for schools, but corruption at various levels meant much of the money never reached its intended destination. Between 1995 and 2001, the proportion of funding allocated which actually reached the schools rose from 24 % to 82 %. In the interim, they initiated a programme of openly publishing data on how much was allocated to each school. There were other factors but Reinikke and Svenssons analysis [4] showed that data publication played a significant part in the funding increase.

However recent developments, such as the current upsurge of open data challenges (see the ODI Education: Open Data Challenge⁴⁰ the LAK data challenge⁴¹ and Open Education Challenge⁴², an EU funded initiative to support projects who receive mentoring and seed funding through the European Incubator for Innovation in Education, have meant that there is an increasing innovation in data use, and opportunities for efficiency and improvements to education more generally. Their potential use is broad. Datasets can support students through creation of tools that enable new ways to analyse and access data, for example maps of disabled access and by enriching resources, making it easier to share and find them, and personalize the way they are presented. Open data can also support those who need to make informed choices on education, for example by comparing scores, and support schools and institutions by enabling efficiencies in practice, for example library data can help support book purchasing.

As part of the Open Data Challenge Education, the Open Data Institute has compiled a set of interesting resources⁴³, including a list of potentially interesting datasets⁴⁴.

Education technology providers are also starting to see the potential of data-mining and app development. For example open education data is a high priority area for Pearson Think tank⁴⁵. Back in 2011 they published their blue skies paper “How Open Data, data literacy and Linked Data will revolutionise higher education”⁴⁶. Ideas around how money, or savings, can be made from these datasets are slowly starting to surface.

6.4 Using Open Data in Education

Schools Sector. Much of the innovative activity around open education data use has focused on the schools sector. Tools highlighted in the aforementioned Open Data Challenge include Locrating⁴⁷, defined as ‘to locate by rating: they locrated the school using locrating.com’ combines data on schools, area and commuting times Schools Atlas⁴⁸, creates an interactive online map providing a comprehensive picture of London schools, current patterns of attendance and potential future demand for school places. Data behind the atlas is available from the data store. RM Schoolfinder⁴⁹ which allows you to compare and contrast different schools, find out about what they excel at and how well children do academically. Most of the information comes from official statistical releases published by the Department for Education and Ofsted including School Performance Tables, GCSE Subject Results, school information from Edubase and summaries of the Ofsted school inspection report. Guardian GCSE schools guide⁵⁰ designed to help parents find and research local schools in England. Search by postcode to find which schools offer individual subjects, and compare how they have performed in GCSE results. Data is supplied by the Department of Education. School impact measures are based upon FFT contextual value-added scores by permission of FFT Education Ltd. Ofstead School Data Dashboard⁵¹ provides a snapshot of school performance at Key Stages 1,2 and 4. The dashboard can be used by school governors and by members of the public to check the performance of the school in which they are interested. The data is available in RAISEonline – you will need to login to access the data and not all is openly available.

The UK is not alone in seeing the benefit of open education data, in Holland, for example, the education department of the city of Amsterdam commissioned an app challenge similar to the current ODI one mentioned earlier. The goal of the challenge was to provide parents with tools that help them to make well-informed choices about their children. A variety of tools were built, such as schooltip.net, 10000scholen.nl, scholenvinden.nl, and scholenkeuze.nl. The various apps have now been displayed on an education portal focused on finding the ‘right school’. RomaScuola⁵², developed under the umbrella of the Italian Open Data Initiative, allows visitors to obtain valuable information about all schools in the Rome region. Distinguishing it from some of the previous examples is the ability to compare schools depending on such facets as frequency of teacher absence, internet connectivity, use of IT equipment for teaching, frequency of students’ transfer to other schools and quality of education in accordance with the percentage of issued diplomas.

Also in Europe E-school Estonia⁵³ provides an easy way for education stakeholders to collaborate and organize teaching/learning information. The system has a range of different functions for its various users. Teachers enter grades and attendance information in the system, post homework assignments, and evaluate students’ behavior. Parents use it to stay closely involved in their children’s education. With the help of round-the-clock access via the internet, they can see their children’s homework assignments, grades, attendance information and teachers notes, as well as communicate directly with teachers via the system. Students can read their own grades and keep track of what homework has been assigned each day. They also have an option to save their best work in their own, personal e-portfolios. District administrators have access the latest statistical reports on demand, making it easy to consolidate data across the district’s schools.

Another interesting project is Social Accountability for the Education Reform in Moldova, a website for enabling the public to monitor the schools performance⁵⁴. The site includes the planned expenditures for all the schools in Moldova (2014)⁵⁵. The School Portal⁵⁶, developed under the Moldova Open Data Initiative⁵⁷, uses data made public by the Ministry of Education of Moldova to offer comprehensive information about 1529 educational institutions in the Republic of Moldova. Users of the portal can access information about schools yearly budgets, budget implementation, expenditures, school rating, students’ grades, schools’ infrastructure and communications. The School Portal has a tool which allows visitors to compare schools based on different criteria infrastructure, students’ performance or annual budgets. The additional value of the portal is the fact that it serves as a platform for private sector entities which sell school supplies to advertise their products. The School Portal also allows parents to virtually interact with the Ministry of Education of Moldova or with a psychologist in case they need additional information or have concerns regarding the education of their children.

Further afield in Tanzania Shule.info allows comparison of exam results across different regions of Tanzania and for users to follow trends over time, or to see the effect of the adjustments made to yearly exam results. The site was developed by young Tanzanian developers who approached Twaweza, an Open Development Consultant, for advice, rather than for funding. The result is beneficial to anyone interested in education in Tanzania. In Kenya the Open Institute used data collected from the Kenya National Examinations Council (KNEC) and the Kenya Open Data Portal to release KCPE Trends⁵⁸ a simple tool designed to visualise Kenya Certificate of Primary Education (KCPE) performance records of primary schools in Kenya from 2006 to 2011.

In Burkina Faso they have opened their open data portal⁵⁹. The open data team of the government have worked with civil society and some start-up to realise a pilot project that consist on visualizing on a map the primary schools of a municipality. In addition, some important indicators for Burkina were present. Those indicators (proximity of canteen, latrine, or potable water point) can help parents choose the best school for their children, investors to choose the better place to build a school, or the government itself to measure the impact of its actions. They also have information on success rates in examinations, the number of classes, the number of teachers, the number of girls and boys, the geo-localisation of the school, and also display a picture of the school. In Brazil the school census collects data about violence in school area (like drug traffic or other risks to pupils). Based on an open data platform developed to navigate through the census, it was possible to see that, in a specific Brazilian state, 35 % of public schools had drug traffic near the schools. This fact created a pressure in the local government to create a public policy and a campaign to prevent drug use among students⁶⁰. Further information is provided in Open Data for Education in Brazil⁶¹.

Similar activity is happening in North America, Canada, Australia and New Zealand. Discover Your School, developed under the Province of British Columbia of Canada Open Data Initiative⁶², is a platform for parents who are interested in finding a school for their kids, learning about the school districts or comparing schools in the same area. The application provides comprehensive information, such as the number of students enrolled in schools each year, class sizes, teaching language, disaster readiness, results of skills assessment, and student and parent satisfaction. Information and data can be viewed in interactive formats, including maps. On top of that, Discover Your School engages parents in policy making and initiatives such as Erase Bullying or British Columbia Education Plan. Education.data.gov⁶³ provides a wealth of information about education in the USA. The Open Data inventory⁶⁴ provides more data reported to the Department of Education. In New Zealand open government data on schools in an app⁶⁵ to help you find schools in the local area.

Bahawalpur Service Delivery Unit (BSDU)⁶⁶, an initiative by the Government of Punjab province in Pakistan, aims to engage citizens in the governance of service delivery. Led by Imran Sikandar Baloch, District Coordination Officer of Bahawalpur district in Punjab, this initiative is built on open data and has already delivered increased attendance of teachers and students over the past year. Technology and design partner for this initiative is Technology for People Initiative based at the Lahore University of Management Sciences. It features a mobile app that allows officials and citizens to monitor attendance by teachers and students at school. The information is aggregated online and made publicly accessible. The aim is to enable and motivate citizens to collect, analyze and disseminate service delivery data in order to drive performance and help effective decision making. The initiative has led to improved teacher attendance, which in turn has led to improved pupil grades. By showing how open data can help in the developing world, BDSU won the Making Voices Count global innovation competition.

Check My School is a social accountability initiative designed and instituted by the Affiliated Network for Social Accountability in East Asia and the Pacific (ANSA-EAP), and uses a blended approach through on the ground mobilization effort and community monitoring, tapping modern technology as a key tool. The CMS project is supported by the Open Society Institute and the World Bank Institute⁶⁷.

Other activities worth nothing are Education GPS⁶⁸, the OECD source for internationally comparable data on education policies and practices, opportunities and outcomes. Accessible any time, in real time, the Education GPS provides the latest information on how countries are working to develop high-quality and equitable education systems. The Pearson Learning Curve Index⁶⁹ combines national data and a number of international rankings – including PISA, TIMSS and PIRLS – to provide an interpretation of how countries systems are performing relative to each other.

Higher Education. In the UK the Open Data Challenge identified applications of open education datasets in services like Which? University⁷⁰ which builds on the National Student Survey (NSS) annual survey held in Unistats, the Key information sets and other related datasets to allow aid students to select a university In Higher Education the development of equipment.data⁷¹ has been funded by EPSRC in response to the need to improve visibility and utilisation of UK research equipment. This relatively simple technology enables searching across all published UK research equipment databases through one aggregation “portal”, allowing greater accessibility with the aim to improve efficiency and stimulate greater collaboration in the sector. The data used is available to download from the site.

A more recent activity has seen Universities UK, a membership organisation for university leaders, run seminar series entitled Creating value from open data. The series has now led into a Jisc funded project with partners from Universities UK, the Open Data Institute, the National Union of Students and the Leadership Foundation. The project has 8 universities signed up: Edinburgh, Oxford, Cambridge, Newcastle, Aberdeen, The Open University, Southampton, Greenwich. Together they will work to develop a web based application and strive towards release of appropriate datasets release. The web app will focus on student recruitment, business processes, research management and the Research Excellence Framework, student experience (e.g., use of labour market information). The project will also include a data capability study and an exploratory look at the data skills curriculum⁷².

The School of Data, through their data expeditions⁷³, are starting to do some important work in the area of education data in the developing world. And in January the World Bank released a new open data tool called SABER (The Systems Approach for Better Education Results), which enables comparison of countries education policies. The web tool helps countries collect and analyze information on their education policies, benchmark themselves against other countries, and prioritize areas for reform, with the goal of ensuring that in those countries all children and youth go to school and learn.

All over the world prototypes and apps are been developed that use and build on open education data.

There are still challenges that those keen to develop applications using open education data face. Privacy and data protection laws can often prevent access to some potentially useful datasets, yet many datasets that are not personal or controversial remain unavailable, or only available under a closed licence or inappropriate format. This may be for many reasons: trust, concerns around quality and cost being the biggest issues. Naturally there is a cost to releasing data but in many cases this can be far out-weighed by cost-savings later down the line, so for example a proactive approach is likely to save time and effort should Freedom of Information (FOI) requests be made.

Well-defined use cases are starting to emerge but can be still hard to find. The EU funded Open Discovery Space (ODS)⁷⁴ project aims to create a platform for teachers across Europe for sharing and repurposing of open educational resources.

However, ODS, also deals with mining data and usage for further improving the value chain of educational resources and open education. It creates a social data layer around education resources that crowd sources appreciation and usage data. Social data in this context is appreciation metadata that further describes a resource. It comprises intentional user inputs such as likert scale star ratings, comments, free or guided tags, shares, etc. From these datasets aggregations can be used in an infinite number of mashups to provide e.g. resource recommendations or karma measures. In addition, ODS also uses tracking data (called paradata) which collects users’ activities in the ODS portal (e.g. looking at a resource, downloading, etc.). This allows for other statistical analytics such as most looked at, or most downloaded resource. In more sophisticated ways it also permits to draw conclusions about the similarity of users that looked at or downloaded the same resources or that follow similar type users. Analogous methods are well known from social networks (Facebook: “friends you may know”, Twitter: “people who you may want to follow”), sales sites (Amazon: “people who looked at this also looked at”), or review portals (Tripadvisor: “most popular or most highly rated hotel”).

ODS goes beyond collecting data from users of the portal alone, but also harvests social data from other OER portals. This is to say that if a user star-rates a resource in a sister portal to ODS, this rating will enter the ODS ratings data through a data harvesting cycle. In this way, opinion mining is not restricted to a single portal alone and enhances the value of the resource descriptor no matter where the users tag it. Harvesting social metadata from other portals encounters no legal obstacles, even if this data is not linked open data, because: (1) it is anonymous data and cannot be connected to a user’s identity, (2) there is no copyright associated with protecting user expressions like star ratings, bookmarks or keyword tags. This is because it does not constitute an act of (substantial) creativity on behalf of the author of such social metadata. ODS not only re-uses social data from associated repositories, it also aims at exposing its own data as open linked data to other third party service providers. It has to be said, though, that paradata (recording user activities in the portal) is not going to be exposed due to ethical and privacy reasons.

6.5 Open Data in Education and Learning Analytics

Online education is producing vast amounts of data about students. Much of these online courses are openly available and the data from them should be too. The data will enable academic institutions and course providers to deliver their courses more efficiently and more appropriately to their students. It will also allow students to personalize their educational experience to best suit their needs. Data collected can include administrative data, demographic information, grade information, attendance and activity data, engagement metrics, course selection etc.

Learning analytics is defined as the measurement, collection, analysis and reporting of data about learners and their contexts, for purposes of understanding and optimising learning and the environments in which it occurs.

Data from online courses can enable grade prediction and student success, measure student performance, improve student retention and determine what learners know and what they currently do not know. It can also monitor learner engagement, personalize learning which in turn can ensure relevant content is delivered. Other potential uses are the reduction of classroom administrative work. For further information on learning analytics see the Learning Analytics Community Exchange (LACE) Project⁷⁵ and the Society for Learning Analytics Research (SoLAR).⁷⁶

6.6 Open Data in Education Challenges and Opportunities

The main benefits of using open data are around transparency, releasing social and commercial value, and participation and engagement. By opening up data, citizens are enabled to be much more directly informed and involved in decision-making. Open education data holds huge potential for students, schools and institutions and governments and policy makers. However there are challenges that need to be addressed.

One clear double edged sword relates to the monetary value of data. The Omidyar Network believe open data, including open access research, could contribute as much as $13tn to the economies of the G20 nations cumulatively over the next five years. This contribution is primarily in transparency and improved efficiencies. The Open Data Institute (ODI), a private limited company established as a not-for-profit organisation set up by the UK government to catalyse the evolution of an open data culture to create economic, environmental, and social value is a good example here. The ODI aims to unlock supply, generate demand, create and disseminate knowledge to address local and global issues. They have carried out work looking at business models in the open data space. Their guide How to make a business case for open data⁷⁷ offers three general business models:

• freemium: you provide an “added value” data product or service, for which you charge

• cross subsidy: you reach more customers, or provide enhanced services to existing customers, through wider sharing and use of your data

• network effects: by collaborating with other organisations, you reduce your costs in maintaining data which you use in your work or extend the possible audience for your products and services

However monetary gain also has issues and trading in data has shown a clear upward trend. Discussions in a recent Twitter chat facilitated by Marketplace⁷⁸ noted that “Schools are often handing over datasets to private industry with few checks and balances on whether it can be warehoused, bought, sold, and traded – all to develop product to sell back at great expense to public education. The key question is the product really furthering public education or is it just lining the pockets of the rich?

Opening anything up makes organisations more vulnerable, especially if they have something to hide, or if their data is inaccurate or incomplete. There is also a cost to releasing and building on data. Often this cost is outweighed by the social or economic benefit generated, but this benefit can develop over time so can be hard to demonstrate.

Other matters of contention include the possible misinterpretion or misrepresention of data, privacy and ownership and the measuring and monitoring of individuals.

6.7 Open Education Handbook

The LinkedUp Project description of work initially described the ‘LinkedUp Handbook on Open Data in Education’ as a “resource for both educators and Web data providers as well as adopters... The LinkedUp Handbook will be created as a living document to reflect project learnings and findings, which will help others, both during the project and beyond it”. To fulfill this brief over time the handbook⁷⁹ has evolved to consider the broader scope of open education resulting in it being renamed as ‘the Open Education Handbook’. During its evolution the handbook has received contributions from organisations and individuals that span sectors and countries. The writing of the handbook has been very much embedded within the Open Education Working Group and will continue to remain an important part of working group work. Embedding the writing of handbook in such a group has ensured that it is part of a committed community made up of practitioners working in open education and those interested in its broader implications.

The handbook is living web document targeting educational practitioners and the education community at large and it been crowd-sourced and drafted over a series of online and offline events. The initial booksprint held in London and was attended by education experts from different sectors (commercial, academic, government, not-for-profit). A second booksprint took place in Berlin on Friday 22nd November 2013 and was organised in collaboration with Wikimedia Deutschland⁸⁰. During this event the handbook was ‘chunked up’ into a number of question areas and discussion took place over the direction of the handbook. On January 20 th 2014, as an activity for Education Freedom Day, the Open Education Handbook was translated and adapted into Portuguese. This process highlighted some interesting possibilities and challenges for the handbook such as the requirements of a global audience⁸¹. A timeline event also took place at which a group physically mapped important open education events, which were then added to an online timemap.

Throughout 2014 the handbook has been further developed through a series of Friday Chats that have taken place on the Open Education Working Group⁸². These discussions have provided the handbook with well-thought out objective content that is not available elsewhere on the web. In late September 2014 in preparation for the delivery of the ‘final version’ of the handbook an external editor was employed to proof read the handbook. The editor was asked to look at areas including overall structure, typos and poor writing, universal style, fact checking, citations and links, glossary and definitions.

Content is key within the handbook and it has a broad coverage considering both practical and factual areas and more discursive topics. Some of the questions it intents to help answer are: What is open? What is education? What is open education? Is traditional education not open? What affect does open education have on education? Who is meant to benefit from open education? What are open educational resources (OERs)? What are Open Licences? What is Open Learning and Practice? What is Open Policy? What is open education data? How does open data relate to open education? When possible references are given to examples and related projects and initiatives. It includes a section on open education data that considers the drivers behind data release and use, available technologies, data use in the developing world, open data competitions and current case studies.

The handbook outline was created using three Google documents. In late 2013 the handbook was moved from Google Docs to Booktype⁸³, an open source platform for writing and publishing print and digital books developed by SourceFabric. It has continued to be written in Booktype and the software has been found to be a suitable platform in which to house a collaboratively written handbook. As a resource the handbook offers an introduction to various topic areas but is also a springboard from which users can connect with other relevant resources. These connections are achieved by links and so the handbook is by nature hyperlink heavy. While it is possible to create a downloadable version of the handbook it is clear that PDF or Word are not the optimum mediums in which to view it. Prior to the delivery date for deliverables LinkedUp requested that the handbook could be delivered in two versions: Firstly an online version that is optimised for those viewing on the web Secondly an open ebook format that can be viewed on a computer while online or offline, and can also be printed. These formats differ from the usual EU deliverable format but it is hoped that the end-result is a useable, user-friendly output that can be reused. The current version of the handbook is now available online, as an ePub book and as a PDF. The Handbook is available under a Creative Commons Attribution 4.0 International (CC BY 4.0)⁸⁴.

The handbook is now a comprehensive and intelligent overview of the current situation with regard to Open Education and Open Education data. Supporting a large and disparate community to collaboratively produce an open resource has posed many interesting challenges: for example how do you resolve differences of opinion? Is targeting for audiences possible? What processes need to be in place to verify content? To realise its full potential such a resource needs to be allowed to continue to evolve and be built upon. As explained previously, the writing of the handbook has been very much embedded within the Open Education Working Group throughout the LinkedUp Project lifecycle, and it is here that it will continue to stay until a more appropriate place is found. Discussions have already taken place around the future of the handbook and possible ideas include moving it to Wiki books, embedding it within Wikipedia and building a front-end for it to use with Booktype. It is hoped that these ideas can be developed further in discussion with the community. An online community session is planned for early December that will explore the future development of the handbook and appropriate delivery mechanisms.

7 Connections with Other Groups

The Open Education Working Group is not working alone in the open education and open data space. Although it may be the only group to span the data and resources space there are many other more focused groups with which it has a lot in common. It will continue to connect with these groups and support their mission statements. Many of these groups are identified in the Open Education Handbook⁸⁵.

A formal connection has been made with the W3C Open Linked Education community group, a focus point for the community to collect, capture and adopt the practices that are going to be the foundation of the web of educational data. The group brings together existing to gather initiatives the practices currently employed to sharing education-related data on the web including vocabularies and best practices. The LinkedUp consortium will lead the community group from autumn 2014.

8 Conclusions

The Open Education Working Group is an engaged, multi-faceted, global community with broad interests related to open data in education. Through online and offline events it has supported a two-way process in which interesting and valuable dialogues have taken place. The long-term impact of the Open Education Working Group and its new and growing community is difficult to measure at this stage. However, issues around data creation and use in education (such as privacy, measurement of learning, online learning, data-driven decision making) are only likely to come closer to the fore over the next few years. Currently the US and European countries are leading the way in exploration of the potential of open data in education, but the next few years may well see other countries, such as Brazil, or those in the global south, picking up the baton. The Open Education Working Group want to support these groups around the world interested in applying the power of open data in education and using it for social good. Its intention is to bring many more people into the conversations that are starting to take place. The challenges at this stage are to ensure the sustainability of the group and that it strives for global coverage rather than the current US, UK and Euro-centric focus. With a growing, vibrant community it is a challenge it is well-equipped to take on.