Skip to end of metadata
Go to start of metadata

Present: Cindy, Colin, Justin, Michelle, Ned, Philip, Ted

Google Doc


  1. ML Tool to help learners to find their learning needs
    1. This learner should be able to use machine learning tools to better understand their own needs and learning approaches, and to find new material, connections, and techniques to support their personal (and community) learning journey

  2. OERs
    1. “how some content is used or what kinds of content it’s related to”

      ML to scan thru OERs to learn about what metadata should be attached with what kind of content. The human intervention should be involved to review/approve auto-assigned metadata.

  3. The match between above 2
  4. Feedback to what works or doesn’t work

Technical starting points:

  1. What “metadata” of OERs that can be auto detected 
    1. and
  1. Can ML systems generate metadata such as “this text contains images” about OERs?
  2. Can ML systems help learners find OERs that are interesting to them by forming “fuzzy associations” between different OERs beyond “these have the same tag”, e.g. “many learners use these together”.
  3. Can an ML system help learners self-reflect with less effort, e.g. by applying tags created by a learner in a MyL3-like diary to untagged posts? 
    1. Inspired by the Coda qualitative data analysis interface 
  4. Three design features of our tools, to start with: 
    1. Reflect on your learning process (e.g. journalling, note taking, guided self-assessment, goal-setting, etc.)
    2. Collect data about how you learn (e.g. tracking how long you read for, how many breaks you took, etc.)
    3. Find or adapt content to fit you best (e.g. using metadata, find a version of the content that is better suited to your needs; personalization infrastructure that can adapt the presentation of content to suit you, etc.)
  • No labels