Joss Winn – Page 7

Implementation Plan

Joss Winn / April 5, 2012March 27, 2013 / Announcements, Development, Policy, Project Management, Research

Introduction

The Orbital Implementation Plan (WP6) is intended to be a synthesis of our initial user requirements gathering (WP5), an assessment of Engineering research data (WP9), an evaluation of standards and technologies (WP10), informed by a literature review of previous work relevant to the Research Data Management (RDM) domain as it relates the discipline of Engineering (WP4).

Therefore, appended to this Implementation Plan is: i) a Technical Specification based on user requirements; ii) a Literature Review; iii) a summary of an institution-wide survey based on the Data Asset Framework; iv) and a draft Research Data Management Policy for the University of Lincoln (WP7), which is currently under-going internal review.

The Implementation Plan has been written at exactly a third of the way into the Orbital project (six months), allowing for a further year of development based on the work brought together in this document. It is worth repeating the objectives of the project, as stated in the Project Plan:

We intend to build on our previous work around the deposit, management and access to university research as well as further existing work in which we are building a platform for data-driven services at the university.

Throughout this undertaking, we aim to improve our understanding of the issues around research data management; develop the requisite skills among the university community to better manage research data; re-use and develop some of the underlying tools we have built to provide an institution-wide service for the ingest, description, preservation and dissemination of research data; improve the way we work on such projects, refining our use of agile methods; build capacity for the local development of academic technologies at the university; develop and implement appropriate institutional policy for the deposit, management and sharing of research data; and develop a Business Plan for the university for the long-term sustainability of our research data.

Our work to-date has pursued many of these objectives closely, reflecting continued effort over the last six months, both inside and outside the project, to build on previous work by using institutional data to drive application development; to improve our methods of access and identity management; and develop an environment that fosters and supports in-house innovation.

This planning document is primarily intended to support the technical implementation of the Orbital application to manage research data at the University of Lincoln. What it does not address is the training to support the use of the application (WP11), nor the Business Case for sustaining the pilot service (WP13), which we are implementing. However, some preliminary work is underway to consider appropriate business models for sustaining Orbital as open source software and we believe that the technical decisions laid out in this Implementation Plan will support the development of a sustainable Business Case for Orbital. This area of work continues and the outcomes are due to be delivered towards the end of the project.

What follows is a brief summary of the appended Technical Specification and Literature Review. I would like to thank Nick Jackson and Paul Stainthorp for their work on these documents, which have brought clarity to the Orbital project and contributed to a much better understanding of RDM at the University of Lincoln.

Joss Winn, Orbital Project Manager, 2nd April 2012.

Literature Review

The management of research data is recognised as one of the most pressing challenges facing the higher education and research sectors. Research data generated by publicly-funded research is seen as a public good and should be available for verification and re-use. In recognition of this principle, all UK Research Councils require their grant holders to manage and retain their research data for re-use, unless there are specific and valid reasons not to do so. (JISC Managing Research Data Programme 2011-13).

To gain a clearer understanding of the more complex and unfamiliar concepts in the emerging discipline of Research Data Management, the Orbital project conducted a review of published literature on the subject (mainly web sites, project reports and guidance documents), with particular reference to RDM in the discipline of engineering.

An online Research Data Management bibliography is being maintained at: http://lncn.eu/bcf6

The project team identified the following nine themes in the literature – for each theme, a recommendation is made which will support the development of RDM infrastructure at the University of Lincoln.

1. Fundamentals of research data and RDM

Researchers are not a homogeneous group, and their data needs are changing as the research landscape becomes more complex. Recommendation: the Orbital project continue its work to assess the storage and other requirements of Lincoln researchers using surveys and interviews.

2. Particular requirements of the discipline of engineering

The ERIM (Engineering Research Information Management) project at the University of Bath has specified the first ever set of RDM principles and terminology designed specifically for engineers. Recommendation: the Orbital continue to work with the Bath team on implementing ERIM’s findings.

3. The behaviour of researchers

What motivates researchers to invest in RDM is not the same as what motivates their institutions. Recommendation: Orbital to use surveys and interviews to understand researchers’ requirements and develop appropriate advocacy materials.

4. RDM policies and legal aspects

All UK Research Councils are introducing mandates for data curation, and in some cases data publication. Recommendation: the Orbital team to support the University’s response to the imminently required EPSRC data policy roadmap and to help develop institutional policies.

5. Data sharing

Research data are at their most useful when they are interoperable with other data. Sharing data leads to a range of real and measurable benefits, and researchers’ interests are protected by a principle of ‘proprietary period’ of privileged access. Recommendation: Orbital work with Research & Enterprise to formulate clear policies on data sharing and licensing.

6. Costs and benefits

The most significant RDM costs for the institution occur at the data acquisition/ingest stage. Institutions that invest in RDM can expect significant benefits including new, unforeseen research activities made possible through the re-use and aggregation of data. Recommendation: Orbital provide guidance to researchers on ensuring RDM is costed into future research funding bids.

7. Curation standards, metadata and citation

Without a system for assigning citations to research data, further curation and sharing is impossible. Recommendation: Orbital incorporate the functionality of DataCite to allow Lincoln researchers to secure a DOI (Digital Object Identifier) for their data objects.

8. Technical considerations

The range of file formats involved in engineering research is a significant area of complexity. Recommendation: Orbital continue to work with Siemens, the School of Engineering, the University of Bath and the DCC to develop expertise in handling engineering data formats.

9. Tools, support and training

A range of immediately re-purposable RDM training kits and planning tools already exists. Recommendation: Orbital review the available material, and use them to design a RDM training programme for the University of Lincoln – also incorporate Data Management Planning (DMP) tools within the Orbital application.

In light of this, the initial objectives of the Orbital project were on the mark, but indicate a broad area of institutional responsibility that goes beyond scholarly communication to affect strategic areas such as recruitment and training, business intelligence and continuity, IP and income generation, as well as future curriculum design and our corresponding investment in infrastructure and estate. No small task.

Technical Approach

Our Project Plan outlined the technical approach that we originally anticipated and six months later this has not fundamentally changed. As detailed in the Technical Specification, we remain convinced of the benefits of pursuing a data-driven, API-centric model of development, using storage and access control methods that support the creation of a modular and scaleable web application that is attractive to both Users and Developers.

As we have learned from our requirements gathering and literature review, Research practices both within and across subject disciplines are varied, suggesting that over the next 12 months, the Orbital project should concentrate on developing an application that remains open and attractive to further development, rather than seeking to design a single workflow for all users’ needs – an impossible task.

We believe this approach best supports the sustained development of Orbital beyond the life of the pilot project, allowing both Researchers and software Developers to create applications for Orbital to suit the requirements of specific research disciplines at a given point in time. Likewise, an API-centric approach will also ensure that our existing and related applications, such as institutional repository software and research information systems can equally be treated as ‘users’ (producers and consumers) of Orbital.

As we outlined in our Project Plan, this approach allows us to benefit from work which continues outside the Orbital project such as that around Access and Identity Management and academic profiles, and the development of data.lincoln.ac.uk. It is also a suitable approach for the development of Orbital as open source software, which should remain simple to develop for specific user’s needs if it is to receive interest and contributions from developers outside the university.

The Technical Specification contains five core functional requirements: Projects, Workspace, Archives, Working Dataset, and Publication. A Project may result in a specific Publication(s), while the Workspace, Archives and Working Dataset allow for three non-sequential methods of data storage, manipulation and analysis. These requirements are loosely coupled to one another, but do not represent a publication workflow. Orbital is not simply intended to be a data repository, but the basis of a flexible collaborative environment for working Researchers.

Each Project acts as a conceptual container for all data and represents the ‘space’ in which administrative, descriptive and contextual metadata is captured and stored, as well as the datasets themselves. It is at the level of a Project that Orbital will interface with other systems, such as an institutional repository or research information system by storing, exchanging and publishing information according to recognised standards, such as CERIF, SWORD2, DOI, etc.

Finally, a core requirement from Orbital is that data should be stored, accessed and transported securely. Being a native web application, we have opted to implement the OAuth 2 protocol to provide secured access to all API functions over HTTPS. As such, all user applications will be treated equally and will be required to access the core Orbital APIs via this popular and mature standard for application authentication on the web. OAuth is increasingly being deployed at the University of Lincoln and work continues outside of the Orbital project to implement it as part of an institution-wide Single Sign On (SSO) architecture.

News from Orbital: Policy, Implementation Plan and a first release date

Joss Winn / April 4, 2012April 4, 2012 / Development, News, Policy

It feels like there’s been quite a lot coming together recently. Here’s what we’re working on:

A draft RDM policy (WP7)
Our Implementation Plan (WP6), which includes our Literature Review (WP4), initial user requirements analysis (WP5), our technical evaluation (WP10), and assessment of data sources (WP9). All of this goes to the Steering Group tomorrow and all being well, will be posted on this project blog a day or two later.
A DAF-based survey (23 questions) of researchers’ data management practices and requirements. We’ve asked all Engineering academics to complete this via the Head of School; we’ve sent a request to all Research Directors in the university to encourage their academic colleagues to complete the survey, and have also advertised it to all staff on three occasions this week via the daily all staff alerts. So far, 28 people have completed it (about 5% of staff on academic contacts). We’ll continue to push this after the Easter break and publish a summary once we think we’ve exhausted our chances of staff filling it in on mass. Probably later in the month.
Harry Newton joined us as a second Developer, working with Nick Jackson. Harry graduated with an MSc in Computer Science from Lincoln last year. He’s been ‘bedding in’ this week and started working on adding ‘projects’ functionality to Orbital.
A release date for v0.1 of the Orbital software is May 1st. A couple of weeks ago, a Robotics researcher asked us if we could help him publish his datasets (20GB). We did so, offering him server space, guidance on his choice of license and a proxy URL to use for citation. It made us realise that there’s probably quite a few researchers like him that just need to get data on the web for citation purposes, so we thought we’d aim to have something permanently in place for the university by May 1st. Functionality for the v0.1 release will be: secure login, basic project creation and deletion, file upload, license picker, publish to permanent URL for citation. We think this is the bare minimum needed for a researcher to publish open research data so that it is permanently citable. From May 1st, we’ll maintain a working system on which to base discussions with users about additional functionality. For the time-being, researchers wishing to upload data will have to discuss it with us first.
We’re on the list of testers for the new DataCite API and have registered with ORCID, too, which has a mock API for testing against.
We’re helping organise the JISC MRD/DevCSI MRD Hackday on May 2-3rd, when we hope to be able to demo this work and talk about the implementation in detail. Fingers crossed.

Drafting a Research Data Management Policy

Joss Winn / April 2, 2012April 4, 2012 / Policy

This morning, four of us (Bev Jones and Paul Stainthorp, Library; Annalisa Jones, Research Office; and Joss Winn, Centre for Educational Research and Development) met for three hours to draft a Research Data Management Policy for the University.

We began by Paul and Bev summarising their experience attending the RDM Policy workshop in Leeds last month, and then went on to look at the requirements of UK funding bodies, as summarised by the DCC. We then reviewed the four university RDM policies linked to from the DCC’s institutional policy page and set about creating a draft policy for Lincoln, which will first be reviewed by the Orbital Steering Group later this week and then referred to the Academic Board and Research, Innovation and Enterprise Committee for approval.

Our draft policy is modelled on the Cross Council Policy Overview by the DCC, which broke down funders’ policies as follows:

Published outputs: a policy on published outputs e.g. journal articles and conference papers
Data: a datasets policy or statement on access to and maintenance of electronic resources
Time limits: set timeframes for making content accessible or preserving research outputs
Data plan: requirement to consider data creation, management or sharing in the grant application
Access/sharing: promotion of OA journals, deposit in repositories, data sharing or reuse
Long-term curation: stipulations on long-term maintenance and preservation of research outputs
Monitoring: whether compliance is monitored or action taken such as withholding funds
Guidance: provision of FAQs, best practice guides, toolkits, and support staff
Repository: provision of a repository to make published research outputs accessible
Data centre: provision of a data centre to curate unpublished electronic resources or data
Costs: a willingness to meet publication fees and data management / sharing costs

We then drew from Edinburgh’s policy to look at how it meets each of these points. Then, we began merging points and writing a policy response, again borrowing from Edinburgh at times.

You can read our draft policy online. If you’re interested in seeing in detail how it was written, go to the File menu, click on See revision history and then at the bottom of the page, click Show more detailed revisions. Amendments to the policy will continue to be made at that location, so we should see the full history of the policy development over time. Sorry, it appears that in read-only mode, Google docs doesn’t allow access to the document revision history. UPDATE: See the link to a version maintained on Github in the comments below.

I should note that this is intended to be a pithy policy statement, similar to what other institutions have written and will be supported by more detailed written guidance, which we’ll develop over the course of the Orbital project.

Understanding and participating in open source culture

Joss Winn / March 8, 2012March 8, 2012 / Events

There are direct and anticipated outcomes of running relatively big projects like Orbital – outcomes which are integral to the success of the project, such as those listed in our project plan: a technical infrastructure for research data, support and training, an institutional data management policy and a business plan for sustaining the work of the project. There are also outcomes which, to be honest, I didn’t entirely anticipate, such as Orbital becoming the pilot project for how the university tackles integration with the cloud; or the implementation of a new development tool-chain and associated working practices.

Yesterday wasn’t originally anticipated either, as the Orbital project hosted a meeting to raise awareness of ‘open source’ among staff at the university. It’s a term that we hear quite often these days and increasingly it’s being applied to non-software domains, such as hardware, data and education. In effect, it’s being used to refer to a method of participation and collaboration, as much as a legal statement about the ownership of property. In my day-to-day experience, more often than not, it’s a term that’s poorly understood and mis-used, so an open source software development project like Orbital seemed like a good opportunity to ask the question, “what is open source?” and see if anyone else was interested in learning more about what it means and how it relates to the work of a university. With that in mind, I arranged for Sander van der Waal, from OSSWatch, to lead a meeting where we discussed open source in general, but also began to address some specific issues that I think we need to work on as we continue to both re-use and produce more open source software.

The meeting ran all morning, from 9.30-12, and could have gone on for longer. I kicked things off with the slides below, which were intended to provide a brief overview of the work we’ve been doing over the last four years where the use of open licenses was central, and in particular, give a brief summary of why we undertake the work we do and some of the benefits of ‘openness’. I finished up with a list of things I think we need to address and take forward for further discussion. I was pleased that Dr. James Murray, the IP Manager for the university was attending and keen to engage in this discussion, too.

Having set the scene, I handed over to Sander, who led the rest of the meeting. As you can see from his slides below, he covered a lot of ground, which we were grateful for, and we intend to draw from them in our next follow-on meeting. I hope that the Orbital project will now act as a catalyst to the development of guidelines on the use and creation of open source software, as well as a clearer understanding of the business case and business models for open source.

Open Source & Open Development

View more PowerPoint from Sander van der Waal

On a more personal note, having joined the university in 2007 as Project Officer on the JISC-funded LIROLEM institutional repository project, yesterday felt like a bit of a milestone, when I was able to draw together a lot of our work under the banner of ‘open’, and impress upon colleagues what we’ve learned and achieved and the direction I think we need to go in.

Before too long, I’d like there to be a greater appreciation across the institution of how the open source movement is changing the way some of us think about (intellectual) property and the nature of work and how this is reflected by the environment we work in. Open source (and its open * derivatives), is not a panacea to society’s problems, by any means, but its impact on our lives in just twenty years or so has been quite profound and it’s impact on the nature of research, teaching and learning is increasingly apparent. Since the development of time-sharing systems fifty years ago, programmers have been building tools with each other that allow them to share their knowledge and their productive capacity across divides in space and time that once presented significant barriers to collaboration. Variations on these tools (hardware, software, legal), are now available to researchers, teachers and students outside Computer Science programmes and present challenges as well as new ways to conceive the organising principles of property and work.

In the future, I’d like institutional projects (not just discreet research projects), such as Orbital, to somehow be tied into curricula for courses where we turn classrooms into hackerspaces, project work into apprenticeships, award degrees on the basis of participation in and learning from open source projects, and help students form start-ups by creating an intensive but supportive learning environment along the same lines that Y Combinator has done. None of this is beyond the capability of our institutions, nor in conflict with the idea of the university. From where I stand, it is the only direction available to us if we wish to remain relevant to young people’s lives and aspirations: on an every day level, technology is a determining force in society and is determining how we undertake research, teaching and learning, but in response, it’s the hackers who are changing technology and therefore have a role in the future of the university.

If you’re interested in further reading about open source, I recommend the following books:

Benkler, Y. (2006) The wealth of networks: how social production transforms markets and freedom.

Fogel, K. (2006) Producing open source software: how to run a successful free software project.

Lindberg, V. (2008) Intellectual property and open source.

Weber, S. (2005) The success of open source.

Notes from team meeting 23 February

Joss Winn / February 23, 2012February 23, 2012 / Meetings

Development

Work on authentication complete. Blog post to follow. User authentication working over OAuth. API framework in place. Recruited Harry Newton as Developer on Orbital, starts end of March. Jenkins Continuous Integration quality assurance environment set up. Staff profiles producing RDF dump. EPrints integration awaiting work from EPrints Services.

Work packages

WP3 complete
WP4 complete end of Feb
WP6 complete end of Feb
WP9 complete end of Feb
WP10 complete end of Feb

Events

Dev8D/CodeIgniter conferences (14-19/02). Helped organise both conferences. Nick presented on API-driven development, Lightning Talk on OAuth, part of data repositories panel, both presentations featured on Slideshare home page. Joss Chaired data.*.ac.uk panel.

Business case for open source (Lincoln, 07/03 am)

MRD Policy meeting (Leeds, 12-13/03)

RDM Policy drafting workshop (Lincoln, 02/04 am)

OR12 paper/workshop (05/03 deadline). API-driven development.

MRD HackDay (Manchester, 3-4/05)

MongoDB conference (London, 20/06). Lee/Harry.

Actions

Nick: Workshops on source control, testing, CI and deployment. 3rd week of March.

Nick: Documentation, write up tech review, data review.

Nick/Lee: Review choice of database.

Nick/Mark: Review security.

Nick/Bev: Meet to look at metadata requirements

Joss: Write Implementation Plan

Paul: Arrange meeting with Nottingham/ADMIRE project.

Paul: DAF survey

Paul: Lit Review by end of week.

Author: Joss Winn

Implementation Plan

Introduction

Literature Review

Technical Approach

Related project blog posts

Chosen Methodology

Literature Review

Initial User Requirements

Assessment of Data Sources

Evaluation of standards and technologies

News from Orbital: Policy, Implementation Plan and a first release date

Drafting a Research Data Management Policy

Understanding and participating in open source culture

Notes from team meeting 23 February

Development

Work packages

Events

Actions