Step 3: Upload to S3 Bucket

Before you can upload the datasets to the S3 bucket, you need to structure it according to our API’s requirements.

You will need to create four directories in the root of your S3 bucket:

  • item_meta_data
  • user_meta_data
  • user_item_interaction
  • schemas

The following directories are where you will upload your CSV datasets:

  • Item_meta_data
  • user_meta_data
  • user_item_interaction

Please note that you can upload a singular CSV file or multiple files. Our API will read and process all the CSV’s from each mentioned directory.

The result should look something like this:

  • aws-bucket-name
    • item_meta_data
      • item_meta_data_batch_1.csv
      • item_meta_data_batch_2.csv
      • ...
    • user_meta_data
      • user_meta_data_batch_1.csv
      • user_meta_data_batch_2.csv
      • ...
    • user_item_interaction
      • user_item_interaction_data_01.csv
      • ...
    • schemas
      • user_item_interaction.json
      • item_meta_data.json
      • user_meta_data.json