Association Rules to suggest new courses in e-Learning



Hi all,

I am new in the Data Science domain. Currently, we want to improve our e-Learning system and we are thinking about how to suggest new course to our learners based on the courses that they have been studying. I am trying to research the Association Rules and apply them in e-Learning. However, I am not sure what input data will need to be run Association Rules to meet our expectation in e-Learning.

Highly appreciate if any suggestion.

Best regards,
Dung Dinh


Hi @dungdinh,

Correct me if I am wrong, can your problem explained by the statement “People who have done course A, have also done course B” ?

If yes, you should use a recommendation system to solve your problem. As described here

Recommender Systems aim to provide the most relevant and accurate items to the user by filtering useful stuff from of a huge pool of information base. Recommendation engines discovers data patterns in the data set by learning consumers choices and produces the outcomes that co-relates to their needs and interests.

The input data to the system would be the historical data of all the learners along with the courses which they have done. You can then train your system on this data, and ask it to give the top courses which similar users prefer.

PS: Please read the quora answer on “How is association rule compared with collaborative filtering in recommender systems?”


Hi @jalFaizy,

Yes, it is my problem. How do you think Association Rules will solve this problem. Actually, I am new so I am not sure there are other algorithms to help me on this?


As I pointed out earlier,

To get a background reading on this topic, you should go through these papers

What I would suggest you is,

  • Do an extensive survey of the techniques used now to solve similar problems (viz. Recommender systems) (this link could be a good starting point)
  • Search for practical implementations of these techniques.
  • Use a prebuilt library for your problem, or build one from the scratch


Many thanks for your attached links @jalFaizy . From now, I can capture some idea to solve my problem. Very interesting!

Thanks again,