Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Data Access in A Cloud World

Data Access in A Cloud World

Two key strategic trends are in progress: pedagogical use of learning analytics and the move to cloud infrastructure. This session explores the intersection of these two trends. As we seek to bring more learning analytics to drive student success, ensuring robust access to learner data is essential. We will discuss past practices and look to current options in the context of cloud infrastructure, and invite attendees to bring their practices and questions to our session.

Sean Yo

May 17, 2017
Tweet

More Decks by Sean Yo

Other Decks in Education

Transcript

  1. Sean Yo – D2L Sr. Product Manager Analytics & Data

    Platform In a Cloud World Data Access
  2. Hi! I’m Sean Yo • Sr. Product Manager @ D2L

    – Most of my 4 years: Analytics & Data Access • 12 Years in Higher Ed IT • Teach with Brightspace at the College level • Get the slides: http://sean.ly/LILI17
  3. Why are we here? • Two key trends in education

    – Learning Analytics – Cloud Computing • Both continue to have strong growth • How does the intersection of these trends affect data access in EdTech?
  4. What does this change mean? • I want to level

    set on Learning Analytics • This is an investment for our discussion about the cultural change, strategy and tactics that we’ll wrap up with.
  5. Learning Analytics • What is Learning Analytics? – For individual

    learners to reflect on their progress; – As predictors of students requiring help; – To help instructors plan interventions; – To drive instructional design improvement; and – For institutions to make business decisions. Powell, Stephen, and Sheila MacNeill. Institutional Readiness for Analytics A Briefing Paper. CETIS Analytics Serie s. JISC CETIS, December 2012. http://publications.cetis.ac.uk/wp-content/uploads/2012/12/Institutional-Readiness -for-Analytics-Vol1-No8.pdf.
  6. Learning as a Digital Experience • Today’s notion of Learning

    Analytics is built on the data capture capabilities of digital platforms – If it doesn’t happen online – we can’t see it – If we can’t see it we can’t count it – If we can’t count it…there is no data
  7. Emerging Value • Shifting learner intervention to an increasingly evidence-based

    process • Detailed tracking of learner and instructor behaviour • Large, well-formed data to analyze – Analytics, Machine Learning, Prediction, Adaptation • Easy to automate and low cost to distribute
  8. Trade Offs • Digital course design requires explicit development to

    drive data collection • Pedagogical methods need to be adapted • LMS systems can only measure what is delivered via the LMS • Data Literacy increasingly important
  9. What does this change mean? • I want to level

    set on Cloud Computing • This is an investment for our discussion about the cultural change, strategy and tactics that we’ll wrap up with.
  10. What is Cloud Computing? • Co-location? • 3rd Party Hosting?

    • External Data Centre? • Outsourced Computing Infrastructure?
  11. Computing as Utility • On-demand – computing resources, over the

    internet on a pay-for-use basis. • Elastic resources – Scale up or down quickly and easily to meet demand • Metered service – Only pay for what you use • Self service – All the IT resources you need with self-service access https://www.ibm.com/cloud-computing/learn-more/what-is-cloud-computing/
  12. What does this change mean? • We’ve level set on

    Learning Analytics and on Cloud Computing • Now we get the payoff for our discussion about the cultural change, strategy and tactics that we’ll wrap up with.
  13. Data Access • What is it? – Exports of data

    from the LMS to use outside of the L MS – ODBC, SQL Export, CSV, JSON/API • What isn’t it? – Reporting, Visualization • Why is it important? – Custom/in-house analysis, data discovery, integration
  14. Data Access – On Prem • On-Prem Data Access was

    essentially transparent • Infrastructure is on-site – data storage & transfer are effectively $0 • Direct access via SQL or File System access = fast, simple & convenient • Risk: Capacity Planning, Cost, No data contract
  15. What have you done? • How have you used on-prem

    data access? • Are you still doing this? • What are the benefits for your org? • What are the pain points for your org?
  16. Data Access - Cloud • Cloud Data Access – Typically

    CSV extract or JSON API – Could be a SQL export, but this is inefficient • Benefit: Data Contracts, built for integration • Risk: Storage and Transfer are metered, New skills (scripting/API) may be required
  17. What have you done? • How have you used cloud

    data access? • How has integrations worked? • What are the benefits for your org? • What are the pain points for your org?
  18. The Hard Problem • Cultural Change is the hard problem

    with Cloud-based Data Access – Shifting away from infrastructure being “under the desk” and accessible by “sneaker net” – More value by more usage of LMS – More success with clear requirements in advance
  19. Strategies • Support data standards! Put them in your purchase

    requirements! • Identify what you’re doing today around data access • Validate the “We Need All The Data” request from internal stakeholders • Articulate measurable outcomes and the data needed to achieve those outcomes • Lean in the strengths of Cloud delivery services
  20. Tactics • Prepare for a heterogeneous data pipeline • Start/Grow

    in-house data ecosystem – ETL/DW, Data Lake etc. – How can you make your data ecosystem Future Friendly? • Assess if you have sufficient data expertise – Data Analysis is a mature discipline