https://github.com/guidasneves/spotify_recsys
End-to-end Spotify recommendation system
https://github.com/guidasneves/spotify_recsys
ai artificial-intelligence content-based-filtering data-science deep-learning machine-learning recommender-system recsys tensorflow
Last synced: about 2 months ago
JSON representation
End-to-end Spotify recommendation system
- Host: GitHub
- URL: https://github.com/guidasneves/spotify_recsys
- Owner: guidasneves
- License: mit
- Created: 2025-01-31T03:49:07.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-03-05T16:16:12.000Z (over 1 year ago)
- Last Synced: 2025-03-27T06:16:14.827Z (over 1 year ago)
- Topics: ai, artificial-intelligence, content-based-filtering, data-science, deep-learning, machine-learning, recommender-system, recsys, tensorflow
- Language: Jupyter Notebook
- Homepage: https://medium.com/@guineves.py/spotify-tracks-recommender-system-with-deep-learning-45d1bb3d3788
- Size: 4.94 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# End-to-End Spotify Recommender System (Sistema de Recomendação Spotify de Ponta a Ponta)
**[EN-US]**
* **Medium article** about this project in English: [Spotify Tracks Recommender System with Deep Learning](https://medium.com/@guineves.py/45d1bb3d3788);
* Complete code for this project, with all detailed explanations in English and Portuguese: [notebooks](./notebooks);
* Complete scripts for this project in English and Portuguese: [src](./src/).
**[PT-BR]**
* **Artigo no Medium** desse projeto em inglês: [Spotify Tracks Recommender System with Deep Learning](https://medium.com/@guineves.py/45d1bb3d3788);
* Códigos completos desse projeto, com todas as explicações detalhas em inglês e em português: [notebooks](./notebooks);
* Scripts completos desse projeto em inglês e em português: [src](./src/).
This repository contains the implementation of a Spotify music recommendation system with the Content-Based Filtering algorithm using Deep Learning.
The system was created end-to-end, that is, from the connection and authorization with the [Spotify API](https://developer.spotify.com/), extraction of data from the API, performing exploratory analysis and pre-processing of the data, definition of the model and inference, given the recommended songs, they are added to a Spotify playlist.
The Content-Based Filtering algorithm was created using the [tensorflow](https://www.tensorflow.org/?hl=pt-br) framework.
**[PT-BR]**
Esse repositório contêm a implementação de um sistema de recomendação de músicas do Spotify com o algoritmo Content-Based Filtering utilizando Deep Learning.
O sistema foi criado de ponta a ponta, ou seja, desde a conexão e autorização com a [API do Spotify](https://developer.spotify.com/), extração dos dados da API, realização da análise exploratória e do pré-processamento dos dados, definição do modelo e inferência, dada as músicas recomendadas, elas são adicionadas em uma playlist do Spotify.
O algoritmo do Content-Based Filtering foi criado utilizando o framework [tensorflow](https://www.tensorflow.org/?hl=pt-br).
## Table of Contents
* [End to End Spotify Recommender System](#1)
* [About](#2)
* [Table fo Contents](#3)
* [Setup](#4)
* [Dependenceis](#4.1)
* [Install](#4.2)
* [Environment Variables](#4.3)
* [Run](#5)
## Setup
### Dependencies (Dependências)
* [python](https://www.python.org/)
### Install
```terminal
git clone https://github.com/guidasneves/spotify_recsys.git
python -m pip install -r requirements.txt
```
### Environment Variables (Variáveis de Ambiente)
**[EN-US]**
Definition of environment variables in the `.env` file, necessary for the system to run.
```text
CLIENT_ID_SPOTIFY:"the client ID generated after registering your application"
CLIENT_SECRET_SPOTIFY:"the client secret generated after registering your application"
```
**[PT-BR]**
Definição das variáveis de ambiente no arquivo `.env`, necessárias para a execução do sistema.
```text
CLIENT_ID_SPOTIFY:"o cliente ID gerado após registrar seu aplicativo"
CLIENT_SECRET_SPOTIFY:"o client secret gerado após o registro do seu aplicativo"
```
All scripts are inside the [src](./src/), `./scr/` directory. Use Jupiter Notebooks in the [notebooks](./notebooks/), `./notebooks/`, directory with the complete system explanation of each step and the complete code for each step.
Notes:
* Spotify has removed the feature to extract the features of each track from the API.
**[PT-BR]**
Todos os scripts estão dentro do diretório [src](./src/), `./scr/`. Use os Jupiter Notebooks no diretório [notebooks](./notebooks/), `./notebooks/`, com a explicação completa do sistema de cada etapa e o código completo de cada etapa.
Notas:
* O Spotify removeu o feature para extrair as features de cada track da API.
