Systems and Journeys are being combined with Memosets. (Perfect).
At the lowest level, I could denormalize and add descriptive columns but it would be nice if I could then tag (add attributes ) at the set and group level.
I still want to be able to differentiate systems, journeys, and memosets but when doing analysis on them all 3 are learned and the learning could be measured and tuned.
Systems,Journeys,Other are a Type of Memoset
Memosets use Systems/Journeys/Other
(linked list // self-referential association )
underlying assumption -
If I/you/someone could get training data out of your system (both in the software and in the memory league training it could be analyzed.
If you could categorize and find features within the larger dataset of all students then you could hook this into sci-learn or azure machine learning or TensorFlow. Even basic multivariate analysis and old fashioned regression is likely to be interesting.
Both on the training and on the software side your data could be immensely valuable to improving learning techniques and then be used for feedback into the software and/or published as someone’s PhD/Masters thesis if they were bent that way.
There is going to be some extremely worthwhile data to be mined here.
Or I could be completely wrong. ( I suspect if there aren’t diamonds here that it would only be because there aren’t enough data points yet).