AL Workshop 2018 Meeting Notes

Owned by Nimisha Asthagiri (Deactivated)
Last updated: Apr 15, 2018 by Julia Mullen
3 min read

<Please enter your notes here>


Julie Mullen Notes:

Day 1: Presentations - posted elsewhere

Day 2: 

I arrived late - so I was trying to catch up with the discussion - see bullets below from morning session - images uploaded to other page.

Learning Path


Adaptive Applications

  • Adaptive engine interacts with user
  • UX elements
    • Sequence

Verbs:

  • On/off/elements ux elements
  • Content
  • Scope/filter
  • State


Basic central design – content (Pearson adaptive)

Learning model availability used to adapt UX

Adaptive exp runtime query


Adaptive Engine

  • Hinting – problem level
  • Mastery – problem level
  • Recommendation – consider authority level (what happens when we give user choices -
  • Gamification – adaptive ux elements
  • Need fallback path
  • Should lms inform AE re: platform (mobile, desk) (yes, UI etc)
  • Includes
    • Content map (? Not same meaning)
    • Forums/discussion
    • Tools/schedule
    • Learner mode

xAPI

  • Event is dot – who did what when with what result
  • But we need the sequence – student stopped video example – of dots
  • Want to record events from pedagogic perspective


Afternoon Session:

Questions from the morning:

    1. Stakeholders
      1. Instructor – can override
      2. Course team – configures
        1. Student can modulate? (y/n)
        2. Required
        3. recommended
  • Adaptive engine - suggests
  1. Platform – enforces
  2. Learner – creates own path for learning goal cross platform
  1. Course teams configure AE
  2. How to use AE result
  3. Level 1 – Define interfaces for AE at Adaptive UX component level
  4. Level 2– assessment level
  5. Level 3 – adapting sequence/structure
  1. Concept map (Adaptive Sequences)
    1. Visual representation
    2. Mathematical or CS operation of mapping
    3. Structure of course
  2. Review Adaptive Usages list
  3. Specific use case walkthrough (water through the pipes)
    1. Walking through Harvard Engine
  4. UI experience – needs
    1. Pluggable (Adaptive) UX components (cross course potential)
      1. Can use Adaptive Exp. Service
    2. In-course x blocks
      1. Can also use Adaptive Exp. Service


Defining UI responsibility

Propose: not expecting adaptive engine to create UI (i.e html) but can give feedback re: how to configure (mobile/desktop), what adaptive options/block to add

key: not part of Adaptive Engine


What’s next API

  • Set of options
  • Could be defined by LM and LMS would render


Update diagram:

Replace LMS with “front end” and “back end”

What is Adaptive Ex – extended xblock with more flexibility and functionality

Custom engine <=> custom xblock

Braden: xblock = leaf content, UI adds the adaptivity

Agree on “pluggable UX components”


Clear Use Case

  • Adaptive hints
  • Adaptive sequence


Research suggests that maps overwhelm students

Autogeneration of path through content/course/multiple courses for “degree/cert”

Andrew Ang – walkthrough of VPAL updates

Zach – Berkeley use case

  • Initial training
    • Gets event
    • Uses items from event store
    • Compares versions (remove vertical/ add vertical)
    • Builds a set of recommendations
  • Turn on engine week 2
    • Student accesses page
    • Every unit (vertical) has content that comms with AE for recommendation (work around)
    • Comm student at page through events, xAPI
    • Engine not reply to Adaptive Exp until get retuls (to avoid race condidtion)
    • Update history
    • Add to model
    • Model gives as output à comms with Adapt. Exp
    • Adapt Exp talks to block store to get name of unit/vertical and that gets inserted in UI code
    • UI code div ==visible to display to learner
    • When user goes to new vertical – update time spent in old vertical
  • Question – how would this look if Zach had xUnit/adaptive unit

Immediate Deep Dives that need to be done later

  • Content tagging
  • Grading (tabled for future discussion)
  • Reuse some blocks for some users
  • API specification
    • Adaptive experience service API
    • Egnine’s – what’s next

Summary:

Projects:

  1. Blockstore (edX + Opencraft + Harvard
    1. Summer time release
    2. Built into master branch
    3. Tagable at every level
    4. Constructs for tagging
    5. Dissociates content from navigation
  2. Harvard Adaptive courses (ALOSI evolution)
    1. Adaptive assessments
    2. Adaptive units
    3. Fall time frame
  3. CET
    1. Have Campus.il (Open edX)
    2. Adaptive course using CET
    3. Need to integrate xAPI into edX insights
    4. Demo May, phase 2 – integrate to edX
  4. Berkeley
    1. 4 courses in May (2 Microsoft, 2 Delft)
    2. running with current edX workaround
    3. need Open edX partner
  5. Content Library (MIT Campus)
    1. What is the state
  6. Need tagging service (what does Branden see this as?)
    1. Taxonomy 1st class service
    2. How do you define tazonomy
    3. Way to associate content with taxonomy
    4. Query taxonomy
    5. Auto generated
    6. Used across serves
    7. Tags for
      1. Units,
      2. Sections
  • Components
  1. Courses


General Needs

  1. LMS + API from 3rd party
    1. xAPI translator for LMS clickstream
    2. LMS provides who à event store
    3. 3rd party provides did, what à event store
  2. event store
    1. for analysis
  3. plugable adaptive UI elements
    1. for adaptive sequences
    2. broader implications requires working with edX team
  4. block store + other new primitives
  5. metadata



Logistics:

  1. summary of workshop
  2. will make wiki available
  3. project definition open for contribution
    1. time line
    2. description
    3. interested parties
  4. Left for discovery
    1. Metadata/content tagging
    2. grading