/tecton-book-recsys-apply-workshop

tecton as a feature store with recsys

Primary LanguageJupyter Notebook

[Public] Book recommendation system on Tecton

Data source

  • Inspired by https://www.kaggle.com/datasets/ruchi798/bookcrossing-dataset
    • note: rating is from 0 to 10
  • Add plausible timestamps to all three tables
    • book timestamps: e.g. base on yearOfPublication + x months
    • ratings: 2020 - 2023
    • user signups: 2020 - 2023 (min of random date + min rating for a user)
  • See data_cleaning.ipynb for cleaning steps

Cleaned data source

See files at s3://tecton-demo-data/apply-book-recsys/

Files are tracked with DVC too:

  • If you have DVC, you can run dvc pull to get the data files

(optional) DVC

  • dvc init
  • dvc add books_data/
  • dvc remote add -d storage s3://tecton-demo-data/apply-book-recsys/
  • dvc push