Machine Learning

1 readers

1 users here now

Community Rules:

Be nice. No offensive behavior, insults or attacks: we encourage a diverse community in which members feel safe and have a voice.
Make your post clear and comprehensive: posts that lack insight or effort will be removed. (ex: questions which are easily googled)
Beginner or career related questions go elsewhere. This community is focused in discussion of research and new projects that advance the state-of-the-art.
Limit self-promotion. Comments and posts should be first and foremost about topics of interest to ML observers and practitioners. Limited self-promotion is tolerated, but the sub is not here as merely a source for free advertisement. Such posts will be removed at the discretion of the mods.

founded 11 months ago

MODERATORS

communick@academy.garden

[P] I built SRec.ai, a Steam video game recommender (details on comment) (srec.ai)

submitted 10 months ago by ilos-vigil@alien.top to c/machinelearning@academy.garden

3 comments fedilink hide all child comments

you are viewing a single comment's thread
view the rest of the comments

[–] goodreads-rebot@alien.top 1 points 10 months ago (1 children)

This is really nice. I am working on a book recommendation system myself as a hobby project too, and your website is really cool. May I ask you what stack do you use? Also, for the ML part, why did you choose this model? What I dislike with NN is that they provide black box recommendations while users would like to understand WHY results are recommended, which can be done with simpler heuristics (that are explainable). Let’s keep in touch !

[–] ilos-vigil@alien.top 1 points 10 months ago

May I ask you what stack do you use?

Stack for which part (model training, front-end, back-end or something else)? Anyway, i find Torch, TorchServe, PostgreSQL and NextJS are vital part of the project.

Also, for the ML part, why did you choose this model?

Do you mean GCE-GNN as smart recommender? If so, these are primary reasons (copied from SRec blog page),

User-free model, but utilize other user's sequence during prediction. By user-free model, I mean the model doesn't use user ID/representation on prediction. It's a required feature for SRec recommender system since all visitor is anonymous.
Authors of GCE-GNN provide source code of the model. This saves some time to re-implement GCE-GNN or verify different GCE-GNN implementations (such as RecBole-GNN) by myself.
GCE-GNN model architecture/framework is relatively easy to understand.

What I dislike with NN is that they provide black box recommendations while users would like to understand WHY results are recommended

As user, sometimes i also feel this. That's why i also create recommendation by game tags which show some explanation.