Modified on June 23.
The Learning Analytics team at Moodle HQ is reportedly working on a visual interface for Moodle’s prediction engine, according to a feature charter submitted at the end of March to the Moodle Tracker by leader Elizabeth Dalton.
Since Moodle 3.4, the open source LMS comes with a powerful predictions engine. It is also based on open technologies and is fully customizable, but it needs programming skills to modify its components and obtain relevant results. It is an entry barrier for educators who want to enhance the learning experience for students. Currently, without any coding the interface only offers a binary (yes-or-no) predictor about the chances that a given student drops out.
With the new proposed feature (currently known as Issue
MDL-61806 MDL-62302) a visual interface would allow users to manipulate and link the different “reusable parts” available on the engine to “create many models.”
In the current version of the Issue, a no-coding-needed “Edit Model” page would enable users to build and refine their own predictions. The available data items can be used as Indicators (inputs) or Targets (outputs), and other model settings could be edited as well.
In the second stage of the feature, the user would be able to use the visual interface to set notifications depending on the model’s outcomes. These notifications could be associated with “actions” the user can take, whether they involve reaching out to a specific student or modifying parts of the course itself. (The Analytics API already offers this functionality for PHP developers.)
Models could be saved, “registered,” and exported using a “shareable format” across Moodle sites. The project is in the early stages, requiring work on a number of Moodle fronts and making it difficult to offer a dependable estimate for the date of availability. Dalton encourages users to visit the Moodle Tracker and show interest for ongoing work by Moodle HQ, primarily by voting and commenting for the Issue, as well as “Watching” it and “Saving” it as a favorite.
Modified on June 23. Moodle HQ’s Learning Analytics team leader Elizabeth Dalton offered the following clarifications:
- Issue MDL-62302 (not MDL-618016) in the Moodle Tracker offers the best reference for the current work on the visual analytics interface.
- Notifications and actions based on insights (model predictions) are already supported in the Moodle Learning Analytics API, and the existing models make use of them. The proposal is to add a web GUI interface to edit these, rather than requiring PHP coding.
- Community support for this work, helps set development priorities. Please comment in the issue MDL-62302 and vote if you want to support this proposal.
This Moodle Technology related post is made possible by: Moonami a company that provides a full range of Moodle services that combine the flexibility, scalability, and power of Amazon’s world-leading cloud platform (AWS) with fanatical Moodle support. Click here to learn more.