Federating the Next Generation: Day 2 Closing Keynote

Posted 3 months, 2 weeks ago at 1:06 pm. 0 comments

Nate Klingenstein from Internet2 is going to bring us to a close with a ‘cynical’ keynote.

Today…

  • Concept has been proven

  • Many successful federations covering a variety of sectors and applications

  • Note the rise of consumer federated identity e.g. Google, Facebook

  • Scaling internationally and across sectors is a big challenge.

  • Getting the Discovery, Trust and User Experience problems right is only going to get harder as these federations scale.

  • Ongoing protocol wars – these will end … someday but not anytime soon… Goal of the Shibboleth project has been to insulate users from these protocols.

  • Levels of Assurance and Attribute support are another long running problem.  Need to reconcile tension between enterprise and consumer identities.

  • The anecdote of the Cardiff Giant .. fakes can be popular and travel faster than the truth.  Fake identities and identity theft is a growing problem.

    The Consumer Factor

    Why have consumer organisations jumped into federated identity … because it is lucrative.  Double click and personalised advertising is still most of Google’s income.  Second generation is email companies like Gmail and hotmail and now Facebbok and Twitter are becoming the largest repositories of personal information and have been very successful at monetising this.  They have also done a great job of raising user expectations that everything should be free and easy and powered by ads.

    Universities and Identities

    Universities house both applications and identities and are the natural home for much user data.  We may outsource the systems that run these but we are not going to outsource the business capability – it is too core.  We also host a number of applications but increasingly not all of them will be locally hosted.

    The Important Players in Academic Identity

    • Government

    • Faculty

    • Applications

    • Users

    “These groups will collectively shape identity in learning over the next 10 years”.

    Nate is now unpicking the different perspectives and goals these groups might have.

    Governments want strong data protection, assertion and protection of digital identity.

    Faculty want good learning resources by the easiest possible route, but they also want undivided attention – do they want social networking in the classroom?  Tension between both stronger IPR and freely circulated intellectual property.  Functional IPR is essential to the cretaion of knowledge – probably not in current form though.  The incentives for creating knowledge are less than previously.

    Commercial applications want a user base to make money from, licensing fees and advertising is a nice plus.  Other applications aren’t really sure what they want but would be happy to be helped with the username/password problem.  They talk vaguely about security and usability.  However Identity Services are critical – indeed, foundational – for “cloud computing”.  Whatever the cloud is one thing your organisation needs to do is have good identity management and federating of that identity.

    Users want CONSISTENCY. There is agreement that users want consistency, but huge disagreement as to what that consistency should be.  There more screaming there is hopefully the faster this will get resolved.  Users get confused.  They like buttons.  They do get the concept of work and personal personas and are able to switch between them – they may not want these to be converged.  Privacy and security are very important to users – particularly in countries where privacy laws are weaker.

    Consumer Identity Today… Facebook Connect or Facebook/Twitter.  Facebook Connect is the most successful consumer identity and is built on a proprietary protocol via a single identity provider. Their inducements for applications are sweet.  The key component of both is the news stream.  Type pad have been collecting stats on the consumer identities used to login to their service – and they offer a lot!  73% are still using their legacy Typepad ID.  However in the last month over 62% of new signups have used one of the consumer federated identities on offer.  MAny consumer applications seem to be promoting three 1. Their own profile 2. Facebook Connect 3. Twitter with other offerings shoved under more…

    Convergence between Academic Identity and Consumer Identity?

    Google Apps is an example of this.  Every Google Apps domain is an OpenID provider.  Shibboleth access into Google Apps.  Users are being ‘trained’ in their consumer habits … to click on Facebook.  Others will be pushing for their buttons to be more prominent.  We don’t have a BUTTON we can put on this landscape (seems like this comes back to Rhy Smith’s work on the Publisher Interface Study and the need for an academic identity ‘brand’).

    Assurance is gravitating to the lowest common denominator.  Non-bouncing email address is the lowest consumer identity provider level of trust.  We need to have modernisation of these systems and their Level of Assurances (LoA).  We aren’t sure what peer validation of identity like in Facebook provides in terms of assurance.

    Consumer world is rapidly realising that attributes are the key.  We are going to have to solve the attribute aggregation problem.

    There are multiple convergence options

    • separate identities, applications, personas

    • side by side with extended discovery

    • attribute plumbing from campus to consumer providers (Google keen here)

    • outsource entirely

    Nate is whizzing through this now as he runs out of time.  There is a lot of big ideas and jey concerns in here so this presentation is definitely worth revisiting and reflecting on.

    How to prepare for the future?

    • Be protocol agnostic.

    • Expectations and functionality are driven by commercial and consumer identities.

    • Users and Governments are unlikely to influence change

    • Faculty will use best tools available

    • Applications like money

    If we want something more benign we have to consider the motivations of these key players and push them towards an outcome we think will be better.  Nate is not sure what this better outcome is yet but does say that whatever happens DISCOVERY is the key control point.  We need some sort of eduID although opinion divided here, but we must proactively consider partnerships with other identity sources.

    Our current course is excellent.  Our infrastrucutre will be key to most possible convergence routes and will be useful.  Hence why this is a paranoid/cycnical presentation but not a downbeat one.

    Phew interesting/exciting stuff but that was rapid and brain is definitely full now.  Need a lie down/sleep on the train.

    So question how does all this relate to Microsoft’s Forefront/Generva.  Basically not sure lies behind the huge Forefront Marketing wall that Microsoft have built up.  Internet2 is doing some interoperability testing with Microsoft’s stack.  Microsoft wants to get into doing this attribute plumbing.

    This was a really interesting event with some good keynotes and mix of parallel sessions, opportunities for discussion and a lot of fun.  Thanks JISC … now to go back and wonder what to think and do about it all.

    Personas

    Posted 6 months, 2 weeks ago at 12:54 pm. 0 comments

    My current research interests (if I ever get any time for research) are to do with the dichotomy in surveillance society (the fear of identity being public) and expose culture (the fear of identity not being public). At the same time my professional interests will soon turn back to managing identities within my institution and across federations of institutions where access to resources is shared or sold (if I ever finish our new institutional repository).

    So I was interested in coming across Personas from the Sociable Media Group at the MIT Media Lab via Shirley Williams.

    this tool is part of the Metro(pathologies) exhibit so it is an art project “about living in a world overflowing with information and non-stop communication”. It allows visitors to become part of the exhibit by contributing their identities. It’s a neat intersection of both my academic and professional interests.

    Which personas to we make and which personas are made? Many of us will create carefully crafted personas that operate in different contexts: family, friends, professionally, publically. Can there ever be a sense of an authentic sense amidst the kaleidoscope of carefully calibrated masks that we put up in front of ourselves as social contexts demand or encourage. This behaviour is multiplied as we create profiles and personas for every web service and social application we use. We may think we control these views of ourselves but we don’t because we will never really have any idea what other people see.

    Using data mining techniques this tool attempts to show us what the internet sees when it’s given our name. It is notable that this is what the internet sees of our name not of us. The name as identifier is pretty poor. Algorithms without names authority cannot easily distinguish one individual from another with the same name; but the whole point of the Personas tool is to illustrate this and reflect on how “digital histories are as important if not more important than oral histories, and computational methods of condensing our digital traces are opaque and socially ignorant”. Perhaps we might want them to be socially ignorant, because the other fear – of having a unique identifier always attached to our name element as an attribute in an xml rich semantic web would leave us with little place to hide and expose all data related to us whether controlled by us or not. This is where are need to be may be undermined by a much greater fear of being made to be if our every interaction is sharply rather than fuzzily exposed.

    As it runs there is a lot of stuff that definitely isn’t me, although I know there are other Alison Popes working in the same sector so they ‘could’ could have been me. I know enough to know that they weren’t, but would you?

    This page though is definitely me!

    persona_alisonpope

    Eduserv Symposium 2009

    Posted 9 months, 3 weeks ago at 11:24 am. 0 comments

    I’m at the Royal College of Physicians today attending the Eduserv Symposium 2009 on Researcher Identities.

    Rough notes are in my Events/Conferences notebook on Evernote. The event is being videoed and streamed live and there is a Ning Social network and chatroom if you want to participate remotely.