skip-prediction-challenge-starter-kit

Installation

git clone https://github.com/crowdAI/skip-prediction-challenge-starter-kit
cd skip-prediction-challenge-starter-kit
pip install -r requirements.txt

Dataset

Please download the dataset from https://www.crowdai.org/challenges/spotify-sequential-skip-prediction-challenge/dataset_files, and extract the files to the data/ folder. Untar them (this might take some time) to have the following directory structure:

|-- data/
|   |-- training_set/ (training sessions)
|   |-- test_set/ (leaderboard partial sessions)
|   |-- track_features/ (track metadata and audio features)
|   |-- submissions/ (submissions folder - contains sample submissions)

Usage

Now you can refer to the list of Jupyter Notebooks for different aspects of the challenge and the datasets. You can access all of them by :

jupyter notebook

Available Notebooks

• Baseline Submissions

• Locally test the evaluation function

Acknowledgements

We would like to thank our co-organizers from WSDM and CrowdAI for making this challenge possible.