Tuesday, June 19, 2012

#mlearncon Aaron Silver and Friends on Tin Can API

Designing learning experiences – over lots of digital modes

Going point a to b in cognition

Tin Can
-         New API – not necessarily a replacement for SCORM
-         Language / tech that allows you to track activities
-         Allows you to track structured data
-         Apply to learning / courseware / apps etc
-         Very generalized
-         Different than designing courseware

Experience design – multiple lenses into an experience
-         Empathy with the notion that we all have different lenses.
-         We are all part of shared experience of being there.  But the individual view is different

Tin Can provides common data format
-         No matter what doing with Twitter, Tappestry, Rapid Intake etc – your activity can be shared with everyone
-         Robust picture of what learning looks like in your org with any event

Tappestry (Chad Udell) – hasn’t been an app built ground up
-         Most learning technologies tied to LMS
-         Most LMSs tied to organization
-         How do you go cross-LMS.  Cross-org
-         Social network that is open, focused on learning, talks to other social networks
-         People use social networks to track mundane stuff (foursquares, fitness apps, etc) 
o       All going out to separate areas
o       No safe repository for life as a whole
-         Every day you learn things.  Constant stream of learning activities
-         Closest – Path. Share thoughts and actions.  Still not made for learning.
-         Want to be able to send activities within systems and between systems.
-         November – Using Location –based services to fill our brains.
-         Want folks to create own learning, tags etc
-         No more digital hoarders – we can use this activity to create our own personal learning system

How TinCan augment and change this?
-         Gave Tappestry a blueprint for the “plumbing”
-         (Intro video)
-         Single safe data store
-         Could use Amazon Cloud DB – but wouldn’t give compatability with other systems
-         Learning activities are YOURS – not your employers.
o       Right now – housed with employers
-         Make your learning part of your digital identity

Tin Can 2.5 months old

Robert Gadd – Onpoint Digital
-         Learning objects on mobile phone.
-         Requirement to connect mobile device to system of record (LMS)
-         Figured out how to do this in an app with no online connection (no API)
o       Store within the app then synch
o       The download course model
-         SCORM is a specification – not a standard
o       Simple statements
o       Point back to LMS
-         Had to marry Tin Can support with SCORM support
o       Balancing act between the two
-         Not as ambitious as what Reuben and Chad have done.
-         Changing the tires while the car is moving
-         Trying to trap stuff in the app itself. Extend the player
-         Mobile not just about taking a course
o       Not replicating the course
o       Social components / communication
-         How do you use the Actor / Verb / Action structure to report
o       Into various systems
o       May help connect to other things
-         May ultimately cut development time and integration time

Rustici software – the implementation
(I really need to talk to Aaron 1-on-1 tomorrow)

LRS – a learner record store
-         Can be bolted onto LMS.  Can also be stand-alone
-         A repository
-         Not once CENTRAL cloud. – want to be able to “roll your own”

Did not want to directly tie to learner management system
-         big complex beast

Tin Can for designing learning in real day job
-         Really didn’t change your job dramatically
-         May remove constraints
-         How can we get these enterprise systems to share with each other?
-         What are we doing and how does this get tied to performance?
-         Maybe more closely marry “learning” activity with business activity (will need to see how this implements)

(OK – still don’t get how this gets implemented.  I may not be technical enough)

Enabling “Big Data” (and this is where Reuben comes in)

Cory – Tin Can could be very disruptive and force us to look at our entire architecture
-         learning in micro-systems and feed all of that into a datamart

Learning Management vs Learning Record Store
-         Learning Management – different
-         Learning Record – how it moves around (maybe looking at it like the electronic medical record?)
-         You decide what matters for you.

This is not a REPLACEMENT for LMS. 
-         (This could be all of a piece…..it’s a spec.  So might be a tool to help gather info on separate activity)

Right now – Tin Can in beta.
-         Query language – still in process
-         Extending verbs – still in process

Vision: Learner owns own data
-         Learner has a “data locker” – web service / app/ something.
-         You ALLOW orgs access to that data
o       Not necessarily all of it
-         Not there yet – federated system

Reuben – if people using Tin Can properly, people are creating unique experiences within orgs.  May wind up being difficult to put it in a way that is meaningful.
-         Tim martin – records moved over are individual statements.

Designing with Tin Can – also have to design for how the data will be used.

(So maybe like the “standards” for health care systems?)

Tim Martin – how do you ID someone
-         Spec references “friend of a friend”
-         If I can figure out what this person does, over many contexts, over extended period of time – what does this mean?  What can we do?

Tin Can – one use is to look for patterns
-         Activities (where subscribing, what reading)
-         Context
-         Other……
-         (setting off paranoia alarms – so what are the ethical implications of this type of data collection?)
-         Instead of after- the-fact, can look at time.  Visible to rest of organization. 
-         Potentially easier to model?  Catching some of the less “visible” behaviors?

Not just about social.
It is plumbing about how to communicate that.

What types of reports can we run?
-         Tin Can doesn’t really deal with that
-         Would need to create response engine that looks for activity patterns that describe a certain type of experience.

(So this is essentially an attempt to gather “Big Data” and capture it for later mining)

Is there a roadmap about how Tin Can will be applied and use cases?
-         “Yes and…..”
-         Learning / Performance Management etc all part of the same ecosystem
-         They are starting with learning because this is what we know

(OK – at this point I am seeing a “Google Analytics” type of thing.  So I need to take a look at what Rustici built.  Plus Articulate – I saw a “Tin Can” reporting component. Sort of a back-end process……The trick will be – how does this get combined and reportable.  And how does this leverage enterprise system reporting ?)

Tin Can – communicating from one end to another. Tin Can and string.
-         Fundamentally – getting systems to talk to each other.

No comments: