Authors

• Francisco Bernad
• Roberto Catalán
• Joaquín Legammare

What is Spotify Most Wanted?

Spotify Most Wanted is a project based on a huge Spotify database which contains all the songs in Spotify's Daily Top 200 charts in 35+1 (global) countries around the world for a period of over 3 years (2017-2020). The idea behind the project is to showcase the most wanted songs, albums and artists of the last three years. The project exposes an API based on Express.js, and a frontend to visualize the information built on Angular. It makes use of two different databases: Neo4J and MongoDB, and takes advantage of each one for optimal queries. You can have a preview of the project here.

What can you do?

There are two ways of using the application:

1. Consume the Express API directly via your preferred way (the API docs can be found on the /api/docs path).
2. Use the Angular project we provide to visualize and query information.

MongoDB

Our MongoDB database exposes the following functionalities:

1. Search Spotify's most popular songs from the Top 200 daily chart of the last three years filtering by:

   • Artist.
   • Genre.
   • Country.

2. Search Spotify's most popular albums from the Top 200 daily chart of the last three years filtering by:

   • Artist.
   • Genre.
   • Country.

3. Search Spotify's most popular artists of the last three years filtering by:

   • Genre.
   • Country.

Neo4J

Pass in an artist and discover the collaborations he has made in the last three years. Visualize the information with a graph via our Angular application.

Local Execution Steps

Database Popualtion

1. Download Spotify HUGE database - daily charts over 3 years dataset from Kaggle.
2. Extract both csv files inside datests/raw-data folder.
3. Run csv_to_json_array.py script inside the population-scripts folder to transform the csv data to json array. This will generate a file named database.json.
4. Run pip install neo4j to install the neo4j driver.
5. Install the apoc library in your neo4j instance.
6. Run python3 neo_import.py dev inside the population-scripts folder to populate Neo4J database. Wait until the script end up running. You MUST have a Neo4J instance running on localhost with the default port and no authorization.
7. Run mongoimport --db spotify --collection songs --type json --file population-scripts/database.json --jsonArray on your desired shell to populate MongoDB database. You MUST have a MongoDB instance running on localhost with the default port and no authorization.

Project Execution

1. Run npm install inside the backend and frontend folders.
2. Start the backend by running npm run dev in the backend directory. This will start the Express server on http://localhost:3000.
3. Start the front end by running ng serve in the frontend directory. This will start the Angular application on http://localhost:4200.