Recommendation Systems

Discussion and resources on Recommendation Systems using cnvrg.io Blueprints

1 Like

The AI Blueprints Recommendation System Workshop is starting tomorrow!

Here’s what you need to know.

If you want to run it on your own data, please upload your data to [here] (Data for recommendations workshop Mar 2022 - Google Drive) which we will move to the S3 bucket
*Make sure is it up to 10MB.


Reminder about data requirements:

Recommendation engines learn from previous interactions between users and items. It can include rating of the item (Explicit), reflecting the level of satisfaction from the item or just the interaction that means that the user purchased/clicked/viewed the item (Implicit).

For example:

User Id: 22345 viewed Item Id: 3454

Or

User Id: 22332 gave Item id: 3342 Score: 4

What you’ll need for the workshop:

If you want the recommendation engine to give recommendations based on your items, you should bring a CSV file containing either: user-item interactions, or user-item-rating records.

Here is the exact format:

Header row

user_id, item_id

Or

user_id, item_id, rating

Interactions (user_id, item_id) pairs, or (user_id, item_id, rating) triplets. Each interaction in a separate row.

user_id, item_id can be alpha-numeric. Rating should be numeric between 0-5.

For example:

user_id, item_id

23233, 3431

23233, 4332

23233, 4543

21111,114

21111, 3431

Or:

user_id, item_id, rating

23233, 3431,4

23233, 4332, 2

23233, 4543, 1

21111,114, 0

21111, 3431, 2

*** Explicit dataset (includes rating) will bring better results in terms of accuracy.

Please feel free to ask any questions here by simply replying to this post

1 Like