The Altair Community is migrating to a new platform to provide a better experience for you. In preparation for the migration, the Altair Community is on read-only mode from October 28 - November 6, 2024. Technical support via cases will continue to work as is. For any urgent requests from Students/Faculty members, please submit the form linked here
recommendation - anyone using it? user knn works but item knn does not seem to
dctech1000
Member Posts: 2 
Contributor I
I am not an expert in this area so please correct my understanding:
1. User CF provides recommended items for a user hence the query in the apply model would be the list of users. This seemed to work well.
2. Item CF using the item knn recommends items for an item hence the query inputs to the model would be items. However, the model expects user list as the query input.
Have I gotten it all wrong? Or is there a an issue in the model?
1. User CF provides recommended items for a user hence the query in the apply model would be the list of users. This seemed to work well.
2. Item CF using the item knn recommends items for an item hence the query inputs to the model would be items. However, the model expects user list as the query input.
Have I gotten it all wrong? Or is there a an issue in the model?
0
Answers
This should take in the following triples (user, product,rating) for building the model.
For applying the model, the typical approach is to find out similar users, get their purchases and then take out the purchases of this user
This part seems to work fine but I cannot figure out to pass in the user's purchases but I guess can do that manually.
2. The Item-Item recommendation (Item Knn) also takes in (user,product,rating) triples for building the models.
However, for making recommendations, it has to get a list of products (e.g. my previous purchases) and it then finds the most similar products to that set.
This is the part which I am not being able to understand as the Apply Model does not require a list of items - and I cannot understand how it is then determing what items to recommend.