New recommendation models

[andrewtavis]

Please use this issue to make suggestions for new models that could be added to wikirec. Suggestions would ideally include some of the following:

• A blogpost or other source where the method is applied for related NLP tasks
• A research paper that details the method and its potential applications
• Source code for the method in Python or another language

Estimates of the model's efficacy would also be appreciated so that a new good first issue can be made and prioritized.

Thanks for any suggestions!

[andrewtavis]

#36 has been created for adding an embeddings neural network to wikirec.

[zedeus]

Have you taken a look at this research paper? Curious to know how it would compare to TF-IDF and your other methods.
https://arxiv.org/abs/2009.11771
https://meta.wikimedia.org/wiki/Research:Convolutional_Graph_Embeddings_for_article_recommendation_in_Wikipedia
https://github.com/digitalTranshumant/WikiRecNet-ComplexRec2020

[andrewtavis]

Thanks for brining it up! I hadn't seen it yet :) Will check the paper out!

[zedeus]

The purpose of recommending articles to edit is obviously different than what to read, but I assume you'd get good results by simply swapping out the list of edits with a list of reads/likes.

[zedeus]

On another note, did you try BM25? From my reading it seems to be TF-IDF but better, by simply adding document length normalisation.

[andrewtavis]

Also haven't tried this one :) We could even do a specific issue for that?