PreferredAI
[FEATURE] Improve Cornac Documentation
Description
Current status
Currently in readthedocs, we split our sections to 4 broad sections:
- Installation
- Quickstart
- For developers
- For researchers
Explanation of proposed improvement
More could be done to further break these sections into smaller tutorials. For example, sections such as the following could be their own tutorials with a broader scope.
- https://cornac.readthedocs.io/en/v2.2.2/user/iamadeveloper.html#data-loading
- https://cornac.readthedocs.io/en/v2.2.2/user/iamadeveloper.html#obtaining-recommendations
- https://cornac.readthedocs.io/en/v2.2.2/user/iamadeveloper.html#model-persistence
- https://cornac.readthedocs.io/en/v2.2.2/user/iamadeveloper.html#deploying-in-production
Further linkage and integration with existing tutorials on these 2 repository directories could be done for completeness as well.
- https://github.com/PreferredAI/cornac/tree/master/tutorials
- https://github.com/PreferredAI/tutorials/tree/master/recommender-systems
Proposed Tutorial Structure
|- Installation |- Getting Started (Renamed from current Quickstart) |- Further Experiments (Focus on finetuning hyperparameters) |- Loading your Own Data |- Obtaining Recommendations |- Persisting a Trained Model |- Deploying as an API |-Model Types | |- Matrix factorization (Link to tutorial at https://github.com/PreferredAI/tutorials/blob/master/recommender-systems/03_matrix_factorization.ipynb) | |- Multimodality (Link to tutorial at https://github.com/PreferredAI/tutorials/blob/master/recommender-systems/05_multimodality.ipynb) | |- Explainability | |- Deep Learning | |- Ensemble Training |- Contributing | |- Link to some contribution tutorials at repo
