Eduserv Foundation Symposium "I before E: identity management, e-portfolios and personalised training" RSA, London 18 May 2006

Andy Powell, Eduserv/JISC opened The day is based on the premise that you can't roll out e-portfolios without a solid framework of identity management, and a national framework at that. So we're looking at the intersection between e- portfolio (EP) and identity management (IM).

Which areas should we be looking at regarding funding? Pilots, demonstrators, standards.

Alan Cranston from DfES

Head of information policy, DfES Technology Group

What's the DfES doing in this area?

They have a 5 year strategy. "Harnessing technology" is among that. Also looking heavily at eGovernment issues.

The Technology Group there has only just been formed (after the 5 year strategy established).

Four themes

Strategic technology provision

• A model for deployment
• Infrastructure, commonmethods of deliver etc.
• "A market-making kind of activity"

Personalised content

• "A new model of content provision for the individual"

E-maturity

• Supporting people and orgs & measuring system level impact

Knowledge architecture

• "A consistent system-wide framewor to guide delivery"

Knowledge architecture

System architecture, data standards, e-portfolio, identity management.

EP has a great impact upon the system architecture that is chosen and so is a big part of this theme.

Identity management always considered part, but has become a huge issue. "The DfES has not needed to know who is out there before." It's now at the core of what we do.

ID cards may have been a solution looking for a problem, but the DfES needs to be driven by business need. What is appropriate? IM needs to be shown as something that has a lot of benefits (for people to buy into it).

3 or 4 years ago the dept was "in the primeval soup" with regard to IM. Now Cranston describes the current state as 'emerging' "some bits recognisable" but without everything being joined up. A few years ago, people had been looking at using National Insurance numbers, but is really inappropriate for education (possibly approp. for tax and benefits).

Unique People Number - used a bit in schools. And Unique Learner Number (started out for post-16 sector). The UPN can't be used in its current form for everything they need to use it for. Need to link the outward-looking UPN (or whatever number used) with the 'child index' so that children are never lost to the education system. Will probably need to go to Parliament to get permission to do this.

(MN: There is a strong assumption that a unique number is absolutely necessary. I don't know why!?!)

e-Portfolio - where are we now?

Complex picture. Functions vary and benefits difficult to assess.

How can EPs be transformational?

What can be left to the e-learning market? What would happen if the DfES did nothing on this issue?

How do we pull out a route map?

A 'push' approach is unlikely to be successful. We need the pull from teachers and learners.

Work to date

Established strategic Knowledge Architecture Programme, starting to tackle issues such as id and data standards.

Launched Information Standards Board for info. data and pedagogic standards.

[More that I missed]

Next steps

Develop a roadmap for e-portfolio approach for Ministers by the summer. (Likely to have some open questions still there by then). Broadening out consultation and development work oce we have the steer from Gov.t.

In the Autumn or early next year, will probably be consulting much more widely.

Morning Parallel Session

Architectural/tech/business model issues around e-portfolios

Simon Grant, John Harrison, Edentity Ltd

What should an EP be uded for?

• assessment
  • pre-defined audience
  • pre-defined content type
  • (close to exam)
• reflection
  • free form content (e.g. blogs)
  • audience is self

• tracking/evidencing skills & achievement

  • content types clear
  • various audiences
  • close to CV.

EP may interact with lots of other systems, though (e.g. e-learning).

Need a system of identity management and sharing information between people.

Problems for the individual

• Too many usernames and passwords
• difficult to update multiple orgs with new contact details

Problems for the orgs

• un-met desire to share data
• expensive authentication costs

One approach - central repository

Unique Learner Number (ULN).

• (Some questioning of the need for a unique number and/or a central repository)

Another approach - permissions hub

Possibility for a "permissions hub" to go into the middle between the individual, the individual's store and the educ. establishments databases.

Where should a permissions hub be held

A logical place could be at an identity broker/information broker. Lets use info. broker.

• Need SSO
• Need authN

Who should host?

• Public sector, charity, commercial sector?
  • John Harrison (privately) thinks that at first PS, then as trust grows, it could move to charity and to commercial sector depending on the needs or interest of the user.

Really fascinating talk and follow-up question session. I'm very impressed by their architecture (via this permissions hub) as it:

• effectively removes the need for unique learner numbers (or national id)
• doesn't need a huge central repository
• maintains privacy - it should be easy for the user to only expose what s/he wants the 'reader' to see
• provides for many e-portfolios per users (they can impress who they like, how they like).

Possible problem: is there a perceived use case for EPs to provide a more central functionality? (I think there might be an assumption that that will be possible - but is it needed?). For example, a gov.t body may wish to contact or publish something to everyone who got an A at Physics A-level in 2006 and who has not proceeded to University. Would something like this be desired or possible? It would certainly fall short of the expectation implied by Alan Cranston (DfES) that someone would be 'keeping track' of schoolchildren to check that they do not 'disappear' from education.

Afternoon Parallel Session

Shibboleth

Richard Dunning, Karin Maslen, MATU, Eduserv Graham Mason, Kidderminster College

MATU

Karin started by describing MATU (Middleware Assisted Take-Up Service for JISC Funded Early Adopters)

Described Eduserv (not for profit, 10 years etc.)

MATU - central repository of information, advice, training, and to scope future requirements for institutions adopting Shibboleth.

28 Shibboleth projects in UK

• 11 early adopter projects, 17 institutions
• [More data about new projects]

Karin gave a quick spin around the MATU web site. Then 'what is Shibboleth'?

• She also said that "Shibboleth is also used to refer to the project and the code package that implements the architecture" - worth pointing out!

Also mentioned the SWITCH web pages as an excellent introduction (like I do!).

What does Shib provide?

• rationalises usernames, passwords etc.
• SSO solution, greater security, secure anonymous access to resources, intitution and individual users can control ARPs.
• Location independent
• Helps with authZ

KC-ROLO Project

Then Graham Mason began to talk about the KC-Rolo project partnership led by him at Kidderminster College and in partnership with University of Birmingham.

Kidderminster Coll. - about 2000 full time students and x part time (so very small for a FE College).

Their needs were to:

• share internal VLEs
• share VLEs with partner institutions
• share learning objects with partner institutions

They use Moodle for their VLE and intranet (about 50% of institutions in the midlands seem to).

• Latest version of Moodle shibbed out of the box

They have a Repository at the City College Coventry

• Search front end protected by shib
• Beginning to Shibbolize the repository submitter

Users like it because:

• fewer logins
• access to more (remote) sites

Created their own federation.

Graham is a good evangelist for Shib. Especially for FE - he showed lots of examples of a Kidderminster user accessing Moodle etc. at Coventry and at Wolverhampton etc.

Second plenary session

Seminar feedback, panel discussion etc.

Nick Johnson - managing information about people

(Seminar feedback). Talked about LDAP, and AuthN and AuthZ (and Shibboleth). How do you identify people as they move from institution to institution. e.g. someone leaves the school sector and moves to HE or FE. They need a unique identifier to take from school onwards and keep for life.

Also need more work on attribute stores for different groups on campuses. A lot of identity management to do. IM isn't usually on the institutional roadmap, but it's a huge undertaking in front of us all.

EPs need not necessarily associate with the institution that you are currently with (they're more for who you AIM to be with). The scope of any one person's EP is going to be greater than the scope of the federation. The EP needs to work with many federations.

Richard Dunning - Shibboleth

Feedback from session that I was in. Richard described what MATU had been doing and talked a bit about IM. He highlighted the problem of someone having an e-portfolio with elements that were outside the federation - how can you access it all?

Helen Richardson

A wide variety of folks in the room. How to get people to work with their EP? In a NE study, about 1500 user case study of EP. Why can't people use something that's already out there (such as MySpace or eportfolio.org)? Problem that this is not supported and asserted by the University system to see transcripts and exam results etc.

SSO very important. Systems need to be customised for different learners and different abilities. Should EPs be private or public? Should be private unless the learner wants to advertise.

Storage: talked about this (no details).

George Roberts - Managing Cultural Change

Had a discussion about what is an EP. Lots of different use expectation. People tend to think of one subset of activity at a time for what is an EP. Therefore, we may need names for these subsets. e.g. reflective learning vs the evidencing of having passed qualifications. We need a body of evidence and case studies - there are some emerging (e.g. Royal College of Nursing in Wales).

Need to define EPs fairly narrow, but for many different user groups. We don't want to produce a white elephant when the users have gone away and used something else like MySpace for themselves.

Questions

I asked a question about the assumed need for a Unique Learner Number - do we need this for EPs to be useful for users? Gave rise to a fairly good discussion, but at least the assumption (and it appears to be a jumped-to assumption) has been questioned. There is some valid questioning as to whether the statistical returns need unique numbers for good quality data, but this might also be worked around.

Also talked about motivation of users to decide what to use EPs for - and that would describe the requirements.

What is IM needed for?

• So that I can get access and no-one else
• So that course results etc. are rightfully applied to me.

Is the gov.t going to come along and mandate another ID management system (e.g. ID cards). The gov.t could, but we could also use (e.g.) mobile phone companies etc. to do this.

Personalised learning: the DfES wants it, so do we (in this room), but what are we going to do about it. One respondent: it is formalised knowledge that is respected and it won't be easy to change that.

Came back to the problem of EP meaning so many different things. It's a problem!

Andy Powell summed up at the end - good that there was people from lots of different groups, HE, FE, local and central gov.t, libraries etc. etc.