FSRS Swift Dataset

The original dataset is hosted on Networks-Learning/spaced-selection.

This repository contains the code to extract reviews from the dataset, group them by user, and save them to parquet files.

The parquet files can be used for srs-benchmark.

Usage

1. Download the dataset from here.
2. Unzip the file and move the stats-20191220-20200731 folder to the root of this repository.
3. Run python group_user_reviews.py to group reviews by user and save them to csv files.
4. Run python build_parquet.py to build parquet files.