Consuming LRMI educational alignments to support the discovery of educational resources

[Renato]

Dear All,

I have recently used LRMI metadata to support the discovery of “similar” educational resources, adding a “more like this” custom functionality to existing search engines (such as Google). Phil suggested to share some ideas here.

My activities focused on the possible utilization of LRMI metadata, not on the technical aspects of their specification. Starting from initial requirements, a few prototypes were iteratively developed and tested with actual educators. The latest prototype exploits a similarity metric based on educational alignments, offering users full control on the degree of similarity of the resources identified. You might have a quick look at the screenshots at https://github.com/renatomario/ERD3.

The prototype builds on open metadata and services of the Learning Registry, as well as on many ideas discussed/prototyped in that context and here (by Phil Barker, Steve Midgley, Pat Lockley, Stuart Sutton – to mention only a few).

One of the findings was that many of my test users were quite reluctant to deal explicitly with standards such as the CCSS or NGSS. Yet, they were definitely happy to use standards transparently, in order to identify similar resources on the basis of learning objectives, rather than content matching. Educators appreciated, in particular, the possibility to control the degree of similarity of the resources identified. For example: resources with high degree of similarity (in terms of learning objectives) for remediation activities, resources with lower degree of similarity for in-depth activities.

Is anybody interested in implementing this particular “more like this” functionality on their search portal, or to discuss these ideas further? Please contact me here or privately. Thank you,

Renato