Materials Data Centre

Our project review meeting today was great in pinning down the final tasks to complete the project. We're finishing off the MDC back-end and refining the front-end with end-user reviews. On the final stretch with the finish line in sight!!!

For Ep2DC and Materials Data Centre we are using Codeplex and the Ms-Pl open source license.

Our projects are managed through the Microsoft Institute for High Performance Computing at Southampton. we like to make sure that our projects have some longevity, and possibility for extension throughout their life. As we work closely with Microsoft, we like to use their open source licenses to ensure that any output from us may be used by them in future initiatives. If you want to find out more about Ms-Pl and Ms-RL licenses check out - http://www.microsoft.com/resources/sharedsource/communitysourcelicensing.mspx

Actual end users have been identified as Materials researchers (first year PhD students) working within the Engineering Materials Research Group at Southampton University.  Two have just produced their first publication (ready to submit to ePrints) and accompanying tensile test data.  We are using this for our exemplar case and have chosen a standardised test as an example data set.  Both end users attended a recent project review with David Flanders.  The disproportionate feedback suggested by David on uploading data (e.g. Other alloys tested are......or here's how that data looks) were both thought to be extremely interesting by the users.  Data format issues are also being discussed with the EP2DC team.

We were delighted to welcome David Flanders from JISC to Southampton on 29 Sept for a project meeting. The discussions were very wide-ranging, largley focussing on user requirements. We ran the first ever end-to-end EP2DC demo, a major milestone for the project!

A major milestone reached today by EPrints developer Seb. We now have EPrints talking to our materials data centre services for the first time via the REST interface, including data file upload using a PUT method.

After an anticipated delay and much preliminary discussion, the EP2DC team finally got together for a series of face-to-face meetings end August - the most productive of which was the curry at Kuti's in Southampton city centre ;-)

EP2DC aims to add a module to EPrints that allows experimental data to be captured together with the publication to which it corresponds, making use of tailored remote data centres to manage the experimental data.  While the EP2DC module will be designed to be data centre agnostic, the Materials Data Centre (another recent JISC-funded initiative at Southampton) provides the remote data centre against which the EP2DC module will be tested and refined.

The EP2DC project comprises three development sprints (data entry, data retrieval, and validation + packaging).  Although entrirely a University of Southampton efforts, the project is very much a cross-discipline, collective effort, involving EPrints Services, the Engineering Materials Department, and the MSIHPC.  As a result of the face-to-face meetings, three clearly defined stages to the first sprint have been identified—deployment and authentication, EPrints UX design and development, and integration.

Deployment and authentication—objective here is to deploy both EPrints and the MDC to a Shibboleth SSO infrastructure, the intention here being to integrate the EP2DC module into the UK HE repository infrastructure as seamlessly as possible.

EPrints UX design and development—the work concerns the reconfiguration of EPrints for an additional stage in the default workflow.  The additional stage allows XML-formatted data sets to be uploaded independently of the manuscripts (or more accurately, the unit of work) to which they belong.  This work is well advanced, and related documentation (design artifacts and prototype user interface) will be posted at the IE Demontrator site, with a corresponding posting at http://blog.iedemonstrator.org.

Integration—focus here is on delivering a range EPrints to data centre integration options that anticipate differing data transfer scenarios.  Presently HTTP POST and RESTful solutions are under development.

We've completed the REST endpoints for the MDC services. The EPrints team are now getting their end to talk to these services. REST for now, with SOAP/XML to follow shortly.

Check out Paul Walk's report on the JISC RI workshop held in Manchester last week.

Our EP2DC team were there and met a lot of great like-minded people. Thanks JISC and UKOLN for organising this!

We've now created our test data centre services for EP2DC Use Case 1.

We've created two endpoints:
(1) SOAP
(2) TEST-POX (Plain Old XML)

Next step is to get EPrints to talk to this!

Here's our second use-case for automagically retrieving data used in an  EPrints publication from a remote data centre.

UC TITLE   
UC 02-accessing test data via EPrints

CONTEXT   
Accessing test data associated with a publication retrieved from EPrints.

BRIEF UC   
A publication is retrieved from EPrints that has associated test data stored at a remote data centre.  A request is issued to the owner of the data to allow access.  This request is approved, and the data set is made available as an URL to the data set at the data centre.  On accessing the data centre, the user is authenticated, and the data made available.

FULLY DRESSED UC
Scope    
EPrints data inject module.

Level    
User goal.

Primary Actor   
Scientist-interested in the test data associated with a publication that is relevant to their own work.

Stakeholders     
- Analyst-needs access to data.
- Data reuser-interested to access the data in anticipation of using the data in a different context.
 - Funder-needs to be able to audit the work performed.
 - Time-any delay conserving the data has a direct impact on data quality.
- Machine-machine operating system and application suite determines data format.

Preconditions     
A publication is retrieved from EPrints that has associated test data stored at a remote data centre.

Postconditions    
The data set is made available after a demand and authorization process.

Main Success Scenario
1. Scientist navigates to EPrints article that has corresponding data sets.
2. Scientist clicks a control (link, multi-select list, etc) to request access to data.
3. The owner of the data authorises data access.
4. An URL is made available to the external facility where the data is located.
5. The scientist navigates from EPrints to the target URL, is authenticated at the data centre, and retrieves the data.

Extensions
2a. Data access is unrestricted and approval is not required.
3a. The owner of the data refuses access.
5a. Data set at external facility cannot be located.