Quantifies the MOOC learner behavior with MOOC-Learner-Quantified (MLQ)
Entry point is
autorun.py. Configuration is done with
config/*yml, see e.g.
Two steps of adding a new feature extraction script to MLQ
- Add an entry to the
- Add a MySQL script to
Each feature table is describing one or multiple objects, where objects include but are not limited to user, video, problem, forum threads. There are two types of feature tables, longitudinal and non-longitudinal ones. If we split a feature by the number of week it belongs to in a course, we get longitudinal features. Only user longitudinal feature table is useful for dropout prediction. But visualization can work on all feature tables and non-longitudinal features may provide more meaningful plots than longitudinal ones.
Scripts for extracting features are in
feature_populate/scripts. Features are described in docs/README.md