Which new hybrid filtering recommendation techniques is suitable for e-learning that can address cold-start problem? I need a technique especially that can be hybridized with collaborative filtering or content based while putting into consideration the heterogeneous nature of e-learning content which includes text, images, audio and video as well as the sequence of acquiring knowledge in learning.
John, look at this filtering techniques for what fits your purpose.. Combining collaborative filtering and K-Neighborhood scheme, content Filtering and Collaborative feature, association rule and Clustering rule. Most of this techniques you can combine to achieve your required result.
John, I do not understand the question well. Please could you expand
Thanks Salina Jeptoo, that is a good suggestion that may solve the problem in question. Combining content based and collaborative is a bit old technique but the other two can work. Was also thinking something like Sequential Pattern Mining which may take care of the learners nature of acquiring knowledge in sequence.
Alfredo & Douglas, what I am looking for is a new technique for automatic recommendation of learning resources to learners in an e-learning environment using an hybrid approach. So the question is which technique can be hybridized with collaborative filtering to give better performance and accuracy while solving cold start problem at the same time.
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India