-
New Feature
-
Resolution: Fixed
-
Critical
-
3.6, 3.6.2
-
MOODLE_36_STABLE
-
MOODLE_37_STABLE
-
MDL-64636-master -
Context
The Moodle Analytics API allows Moodle site managers to define prediction models that combine indicators and a target.
Targets are the key element that defines the model. Represent the event the model is attempting to predict (the dependent variable in supervised learning, also known as labels).
Since Moodle 3.4, Moodle Analytics ships with two built-in targets:
- students at risk of dropping out;
- no teaching
There was no way to define new models through the UI.
After MDL-60944, users can create new predictive models from scratch, combining existing indicators and targets.
We are increasing the number of available targets out of the box, to offer more possibilities for the definition of new models.
Goal
This issue is about adding a new target for course completion, which indicates whether a student is at risk of not completing a course, according to the course completion criteria of that course; therefore, Completion tracking must be enabled for the course and some course completion criteria have been set.
The code for this target is located at moodlesite/lib/classes/analytics/target/course_completion.php where moodlesite is the root directory of your Moodle site.
It can be similar to the existing course_dropout target in core.
There is an example of who it was implemented in tool_inspire. It was later discarded in favour of a solution limited to lack of activity logs during the last quarter of the course.
The requirements shared between course-level targets like course dropouts, course completion or course competencies achievement should be abstracted from the current course_dropouts target.