Archive

Archive for September, 2010

Portal prototype

September 28, 2010 6 comments

Peter and I met on Friday – joined virtually by Helen in the afternoon – to discuss how we might implement the “ALPS portal” and what our prototype should look like based on input from the workshop in August. Peter did a quick sketch which I’ve reproduced (almost faithfully) below and we tried to anticipate some of the technical issues that we will face; Peter and I plan to meet with Gareth Waller to discuss possible solutions to some of these issues via mooted integration with Jorum and, in the meantime, Peter will work on some mock-ups while we get feedback from the rest of the group via the blog (hint!).

This outline is with a view to developing a portal whereby users can both search for (ALPS) resources AND deposit into their respective institutional repositories; these two criteria present technical challenges that are quite separate in some respects though more closely related in others.

Deposit:

  • Authentication:

It is essential that users are able to authenticate in some way in order to deposit.  At the very least we would want to know Name; email and Institutional affiliation and we considered LDAP authentication via the existing institutional system(s) – not sure how easy or difficult this would be and will require us to liaise with systems folk at our respective institutions.  Peter suggested that it might be quicker and easier to set up our own LDAP server so that users need to register before they can deposit – sustainability beyond the life of the project is perhaps a drawback with this scanario.

  • Deposit form

What metadata do we want to capture – what is our Application Profile?  What metadata, if any, can be automatically generated?

An AP based on UKOER used at Leeds Met would comprise:

Metadata field Comments
Title
Description
Keyword(s)
Classification We classify by HEA subject centres and JACS; Jorum classifies by JACS only – we may wish to classify ALPS resources differently?
Contributor:  Role of Contributor
Contributor
Contribution Date
Technical Information:  Technical Format MIME media type – 70 technical formats; intraLibrary identifies file type at upload and field is auto-completed
Educational Properties:  Type of Resource Terms from LOMv1.0: Diagram/ Exam/ Exercise/ Experiment/ Figure/ Graph/ Index/ Narrative Text/ Problem Statement/ Questionnaire/ Self Assessment/ Simulation/ Slide/ Table;

Terms we have added to our vocabulary: Podcast/ Not Applicable/ Presentation/ Photograph/ Quiz/ Spreadsheet/ Tutorial/ Video/ Lecture/ Game/ Animation/ Assessment/ Audio/ Case Study/ Database/ Workbook

Subject to Copyright
Statement of Copyright and Restriction* Our template includes URL for Creative Commons Licence – http://creativecommons.org/licenses/by-nc-sa/2.0/uk/ (required for Jorum Open)

Perhaps a deposit form could be based on that used by Jorum?

Jorum deposit interface

Jorum deposit interface

  • Potential issues
    • Different software uses different metadata standards / Application Profiles -> will need to map between them.
    • ALPS may require different metadata than ukoer – e.g. explicit priority is “Resources presented in the context of specific learning/assessment outcomes” – can Jorum accommodate this?

Selective deposit depending on user affiliation

Jorum has a SWORD endpoint that accepts METS (not IMSCP) so, in theory, all resources could be deposited to JorumOpen by default; user-affiliation – derived from the authentication process – could then determine which of the other Institutional Repositories the file is submitted to.

(N.B. Whether this will be via SWORD is still a moot point – Archivalware doesn’t yet have a SWORD endpoint / uncertain whether Leeds will test with DigiTool or EPrints / intraLibrary accepts IMSCP but not METS; could just be a semi-automated process for the prototype?)

In any case, as far as the user-journey is concerned, they log-in (preferably with institutional LDAP), upload file and add basic metadata – the resource is then deposited in their own Institutional Repository AND Jorum Open and is discoverable from the portal (+ JO; + IR)  – can’t see this being immediate and there is likely to be some sort of lag I think – just how much of a lag will depend on technical implementation; obviously want it to be as quick as possible (and email when available?)

Discover:

Via Jorum?

Peter and I discussed in more detail the possibility of somehow integrating with Jorum – with limited time and resources it seems unrealistic to develop the infrastructure to harvest / cross-search the three repositories ourselves and preferable, in any case, to utilise the national UKOER infrastructure…

If all resources are deposited to Jorum by default, can we liaise with them to this end?  As I understand, Jorum’s new search interface – http://www.jorum.ac.uk/searchOptions.html – searches indexed metadata from both JorumOpen and JorumUK so is there any way we could access that index (API, web-service) to build a bespoke facility to search from our own environment?

This approach would have the added benefit that all resources will also be discoverable from Jorum itself (assuming they are released under an appropriate license – Peter, there was a previous discussion I think around the option to deposit to an open OR closed collection which we’ve neglected to consider.

Search/Browse

One of the main requirements that came from the workshop was:

Resources presented in the context of specific learning/assessment outcomes (will guide discovery)

I’m not certain what these specific learning/assessment outcomes will be and we will need further input but if we anticipate a browsable hierarchy then it is clear that, however we search across the ALPS resources, the underlying metadata/classification will need to include this hierarchy…which might be an issue depending on how Jorum (DSpace) deals with metadata that it “doesn’t recognise” as we are likely to require quite specific classification schema (rather than JACS).

Miscellaneous issues

A couple of issues that probably should be thought about sooner rather than later:

  • What will we call the thing – “ALPS CETL repository portal”? Might seem trivial but could hold things up…
  • Where will The Portal be hosted – on an institutional server…or http://www.alps-cetl.ac.uk/…who is responsible for that site?