4. Project reports
This was the first full year of the JISC ASK project that is building repository services for the university community. These services will enable learners, researchers, administrators and librarians to control the sharing of resources or academic artefacts within an internationally recognised interoperability framework.
The JISC-funded Community-Led Image Collections (CLIC) project is investigating how best to promote the use, and sharing of digital images within the educational community. The affordability of capture devices such as scanners and digital cameras has led to an increase in the creation of image collections by educators. CLIC is seeking to determine how these collections could be made more useful and accessible. Efforts by individuals to share their material are often thwarted by issues such as copyright, intellectual property and cataloguing information ? not to mention the technical difficulty of building a robust Internet resource. CLIC will be making recommendations for a proposed national network of community-led image collections that should help to overcome these obstacles.
- An overview of current community image collection activity
- Selecting community image collections to use as case studies
- Surveying image owners and users
- Reviewing technical needs and solutions for image collections
- Investigating viable models for a network to promote sharing of image collections
The Digital Certificate Operation in a Complex Environment (DCOCE) project was a two-year Joint Information Systems Committee (JISC) funded project that completed in February 2005. The main aim of the project was to look into the use of digital certificates by end users in higher and further education (HE/FE) for authentication to services and also to look at the methods of issuing certificates to users and how to manage the accounts. DCOCE developed code and rolled out to several sets of pilot users and evaluated feedback from over 80 people. Despite some current Linux and MacOS implementation issues, we believe that the use of cryptographic tokens is highly attractive for a scalable PKI for HE/FE institutions.
In general, the use of the client digital certificates was quite straightforward for most users. However, there were many minor problems regarding usability of the certificate issuing mechanism. There were no showstoppers that could have meant that the widespread use of client certificates was infeasible in a HE/FE context. Nevertheless, many of these minor problems have been in existence for some time and show no progress in being resolved. In short, the use of certificates should not be difficult, but for the time being, remains that way. The project final reports may be found at http://www.dcoce.ox.ac.uk/docs/.
This study looked at four projects funded under the JISC/NSF Digital Libraries in the Classroom scheme. Specifically it was asked to look at the e-Learning tools developed or used in the projects, their availability, functionality, accessibility, applicability, and adherence to the emerging JISC IE (Information Environment) guidelines.
The evaluation consisted of document analysis (including project plans, biannual reports, presentations, published articles, evaluation so far, etc.), on-line surveys, structured interviews, and benchmarking against the IE Standards and Guidelines and the JISC e-Learning Framework.
- Specific recommendations to the JISC concerning the specification and management of collaborative projects and the dissemination of tools developed
- Recommendations to the four projects concerning developed tools
- A detailed evaluation report (available at http://www.jisc.ac.uk/programme_dlitc.html)
The Oxford e-Science Centre (OeSC) is a collaboration between OUCS and the Oxford University Computing Laboratory providing the Grid infrastructure within the University and supporting over twenty five e-Science projects across departments. A good proportion of the original e-Science projects have now been completed successfully (e.g. e-Diamond, Remote Microscopy, VideoWorks), to be replaced by a number of new projects (e.g. Integrative Biology, CancerGRID and Evaluation of Shibboleth and PKI for GRIDs). OeSC continues to have representation both at the national level (through representation on various UK e-Science Programme committees and task forces), and at the international level (through representation on the Global Grid Forum Steering Group).
OeSC has recently purchased a number of portable machines which can be loaned out for using AccessGrid (an advanced video conferencing environment). The e-Science Centre is currently transitioning to the newly formed Interdisciplinary e-Research Centre (IeRC) under the direction of the newly appointed Executive Director, Anne Trefethen. One of the first outputs from the IeRC will be the developed of OxGRID, a campus Grid for using idle cycles of desktop machines.
The Evaluation of Shibboleth and PKI for Grids (ESP-GRID) project is funded by JISC to investigate whether and how Shibboleth offers solutions to issues of grid authentication, authorization and security. A further general aim is to investigate the requirements of grid security and whether (and how) the acceptance of Shibboleth assertions for access controls could be used within existing PKI based grids. Another focus is whether Shibboleth can contribute towards an interface between existing information environments and grid computing environments. The project began in August 2004 and is due to complete in March 2006.
Up to July 2005, the project focused mainly on future requirements for grid computing with respect to security and access management. The project attempted to capture previously documented requirements and drew up a newer set, bearing in mind the need to expand the use or benefits of grid computing wider than highly technical computer scientists. The next phase of the project is to produce a working demonstrator of grid use mediated via Shibboleth.
The Guanxi project addresses the problem of controlling access to shared resources in inter-institutional collaborations, particularly in the e-learning context. Guanxi uses Shibboleth, which makes use of user attributes passed between the institutions participating in the agreement to determine access permissions. Guanxi project partners are University of Leeds and the UHI Millennium Institute, and because the Bodington VLE is in use at all three institutions, is an ideal candidate for demonstrating the application of Shibboleth.
Over the past year the project has added Shibboleth capability to Bodington, and Bodington can now be used out-of-the-box as a Shibboleth IdP (Identity Provider) and SP (Service Provider). The Guanxi component can also be used independently of Bodington. Other developments included enabling the VLE to use Webauth. This work was completed and went live successfully in April 2005, and there is now one-click access to the VLE after having logged into the Webauth system.
The Humbul Humanities Hub, based within the Research Technologies Service, is a national service which employs a distributed network of subject specialists to find and review web sites suitable for research and study within a wide range of humanities subjects. Humbul forms part of the Resource Discovery Network and receives funding from the JISC and the Arts and Humanities Research Council.
- Achieving 10,000 publicly available reviews of web sites in May 2005.
- Recruiting and training over 20 subject reviewers, many of whom are post-doctoral and practising academics in subjects we had identified as a priority for development.
- Publishing two booklets which highlight key web sites for English and History. Both had a print-run of 5,000 and both were written in collaboration with an institutional partner. On-line versions were also published. The subject booklets proved very popular amongst target audiences, all copies being distributed.
- Exposing Humbul metadata records through an Open Archives Initiative repository to be harvested by OAIster (Michigan) and subsequently fully indexed by Yahoo! (resulting in a significant impact on referrals to Humbul from the Yahoo! search engine).
- Highlighting freely available peer-reviewed online journals across the humanities and the web sites of AHRC-funded projects through the publication of two discrete collections.
The BNC is an internationally famous 100-million word language corpus, representing the state of the English language as of the end of the twentieth century. OUCS has been responsible for its maintenance and distribution since 1994.
Work on converting the corpus from its original SGML format to a more up to date XML format was carried out this year. To test this process, a new sampler CD, containing just 4 million words together with the newly developed Xaira software, was produced, and proved very popular. This CD, named BNC Baby, was adopted by the Open University for an introductory course on corpus linguistics.
The BNC is distributed on CD and during this year, about 470 copies were sold, rather less than half of these being copies of BNC Baby. This covered the costs of distributing the corpus, and also funded some further developments of the Xaira software.
- A major trial of the Learning Activity Management System (LAMS) in the post-compulsory education sector
- An ongoing investigation into the use of LAMS and Moodle to design and disseminate support for study skills in East London schools (jointly with the University of Greenwich)
- A study of the acceptability and re-usability of LAMS activity sequences in university English departments
- An investigation into lecturers? use of generic tools such as Word, PowerPoint and mind-mapping software in learning design, as well as the forms (graphical and/or textual) in which they represent their completed designs
The recommendations which we have made regarding future technological innovations have been enhanced by rich insights into the cognitive aspects of designing for learning, and our work has already attracted interest at two international conferences and within JISC.
OUCS is building federated search software that can be embedded into WebLearn. The software will allow learners, researchers and librarians to compile and share reading lists according to an internationally agreed interoperability framework.
- Release of prototype open source web-based reading-list management.
- Release of an open source reading-list management application for the desk top.
- Creation of a project community within Oxford to evaluate and train academics and learners in the use of the software.
- Building and maintaining a growing community of interested users: developers, learners, researchers and librarians.
The Oxford e-Science Centre (OeSC) is funded by JISC to host one of the core resources of the National Grid Service (NGS). The compute cluster was delivered in December 2003 and entered early adopter production phase in April 2004. In September 2004, the NGS was launched as a production service. Since production launch, the Oxford NGS cluster has been highly utilised (accounting statistics indicate average utilisation of 75%). The NGS user community includes researchers from UK academia, as well as international collaborators of various e-Science projects within the UK. There has been steady growth of the NGS user community. A table of the number of NGS users is given below.
The National Grid Service has grown with other institutions donating resources to the Grid. In the past year, 3 new partner sites have joined the original 4 core nodes. The joining of new partner sites has been handled by OeSC on behalf of the NGS. There are currently 2 other sites being taken through the joining process and a number of other sites that have expressed interest. The rate of partner growth is currently limited by lack of staff effort.
OSS Watch is a JISC-funded national advisory service on free and open source software. In 2004, JISC requested that OSS Watch substantially increase its offering as a result of user demand. The staff complement increased from 1.25 FTE to 3.2 FTE with a corresponding increase in materials published, events organised and presentations given across the UK, Europe and further afield.
OSS Watch organised two national conferences during this period: Open Source: National Frameworks and Building Open Source Communities. A series of open source roadshows was organised specifically for further education colleges. OSS Watch also participated in numerous conferences and workshops, notably the AURIL spring conference, the EUNIS conference, and the Coimbra Group workshop on eLearning Quality and Open source/Open standards. In May OSS Watch made a series of presentations at Australia universities.
More than 20 briefing notes were published on the OSS Watch website during the year. Of special note is Open Source Development - An Introduction to Ownership and Licensing Issues, which was featured by JISC and received notable mention in the open source media. During the spring OSS Watch investigated wikis (reader-editable websites), eventually settling upon the MoinMoin wiki software launched later in the year.
The Oxford Text Archive (OTA) hosts the Literature, Languages and Linguistics subject centre of the Arts and Humanities Data Service which is funded by the Arts and Humanities Research Council and JISC. In this capacity, the OTA plays a key role in advising academics in literary and linguistics subject areas on the technologies and methods involved in creating electronic resources.
Successful workshops and symposia were held in alliance with major associations and conferences including the British Association of Applied Linguists, the Poetics and Linguistics Association, EUROCALL, Digital Medievalist, the International Medieval Conference, and the Corpus Linguistics 2005 conference. The OTA is now also represented in the Text Encoding Initiative, the Open Language Archives Community, and EUROCALL, an association focusing on computer-assisted language learning.
The OTA embarked on a new European project to develop corpora of spoken language data for language teaching purposes, and a project with Lancaster University funded by the Leverhulme Trust to create a corpus of texts from the first decade of the twentieth century. The latter is to be compared with similar corpora from the 1930s, 1960s and 1990s. Researchers in Amsterdam are working with the OTA to identify and categorise metaphors in language corpora. Agreements have also been made to archive materials collected by the Endangered Languages Archiving Program at SOAS, and by TRACTOR, an archive of Central and Eastern European Language Resources.
As part of JISC's E-Learning Tools programme, the RAMBLE project investigated mobile weblogs and their integration into Learning Environments. The gathering of requirements was conducted with small groups of undergraduates in Medical Sciences and Chemistry for the off-line authoring of weblog entries on PDAs and subsequent upload to a weblog server. This informed the development of a new blog/newsfeed component for the Bodington VLE that allows Weblog content to be incorporated through the reading and configurable display of syndicated newsfeeds (RSS and Atom). WebLearn has been upgraded with the new version, which in turn has helped drive investigations into a central newsfeed system.
RAMBLE has boosted the understanding of mobile learning - both in the practicalities of working with hand-held devices, and the potential offered for personalised learning. Findings have been disseminated in various publications, including issue 44 of Ariadne, published by UKOLN.
Over the past year, SPIE has deployed a pilot Shibboleth environment at Oxford in a proof-of-concept project providing access to local and remote protected resources belonging to different federations, namely the Athens UK and the SDSS Development federations. The SPIE team liaised closely with Eduserv, EDINA, and the OUCS Registration services to formalise this joining process.
SPIE has developed web-based applications which, in conjunction with Shibboleth, enable users (typically, staff and students) to authenticate against Webauth and access a protected resource, without the need to re-authenticate when accessing other protected resources within the same federation. These applications enable users to decide which information about themselves they are happy to release, and allow resource owners to decide which information they want to know about the users before they are given access to local resources.
uPortal is being used as the front door for this environment. The integration of Webauth with uPortal, and the employment of institutional LDAP user attributes for authorisation, are being undertaken in conjunction with the OUCS system development team.
Our main role in the last year has been maintenance and development of the TEI Guidelines, development of workflow tools, provision of a UK web site, and teaching courses. The following staff from the RTS have worked on the TEI:
- Lou Burnard (European editor of the TEI): Guidelines development and teaching
- Sebastian Rahtz: infrastructure tools, publication stylesheets, Roma schema
http://www.tei-c.org.uk/Roma/), and teaching
- James Cummings: teaching and the TEI wiki
TEI development takes place on SourceForge (http://tei.sf.net/), and the first public release of a major new revision was made there in February 2005. Part of this release also forms part of an ISO standard on feature structures.
The goal of the project was to build interoperability points between WebLearn, the TOIA assessment system and the LAMS activity management software. This is a first step towards providing an integrated suite of software that supports core learning and teaching activities.
- Prototype environment for exploring the WebLearn integrated suite of software;
- Enhanced LAMS software that now includes improved interoperability and functionality;
- Enhanced TOIA software that also includes improved interoperability and enhanced; functionality;
- Resources for the teaching and learning of statistics that utilise the three applications (WebLearn, LAMS and TOIA).
The VLE team are responsible for maintaining and developing WebLearn. By attracting considerable external funding, 3 new full-time staff were employed in November 2004, bringing the total to 5 FTE. Funding was also secured to implement the ASPIRE PDP (Personal Development Plan) system, the pilot due to go live at the start of Hilary 2006.
Development has been far-reaching. Numerous bugs were fixed and new features added, including RSS/Atom display, visual access rights display, text and image resources, embedded WYSIWYG HTML editor, and a resource creation wizard (EasyBuilder). In addition, WebLearn was the first University-wide service to incorporate the Webauth single sign-on service.
The VLE team has managed two open source communities, organised four Bodington Buzz seminars, and presented at a number of conferences including JA-SIG UK, CETIS Enterprise SIG, and e-portfolios 2005; staff also teach on the MSc course in e-Learning.
In the summer of 2005 the RTS, in collaboration with the OeSC, the Oxford University Computing Laboratory, and the Humanities Division started work on a number of Virtual Research Environment (VRE) projects under the JISC VRE Programme. The Building VREs for Humanities project is establishing how VREs can aid research in the Humanities. Led by the Humanities Division, with technical assistance from the RTS, the project is currently interviewing researchers to collate their requirements, and building upon a Humanities ICT survey conducted in 2003. The Integrative Biology VRE project is working to embed in silico experiments for heart disease and cancer into the research process of computational and mathematical biologists. The SAKAI VRE project (in collaboration with CCLRC, Portsmouth and Lancaster) is developing a demonstrator VRE and low level tools for authentication and data retrieval. All three projects are working closely together to share the results of their investigations and technologies, within the general aim of providing VREs within other disciplines.
XAIRA (XML Aware Indexing and Retrieval Architecture) is a software system developed at OUCS, initially for use with the British National Corpus project. In Sept 2004, one year's funding was received from from the Andrew W Mellon foundation to rewrite the system as an open source library, supporting a range of APIs for the indexing and retrieval of very large amounts of XML-encoded textual data.
The first open source version of the package (release 1.12) appeared on SourceForge (http://tei.sf.net/) in April. There were three further incremental releases during the year, with more planned as the software undergoes further testing.