MADE Winter 2023/24 Project: Sentiment-Driven Spotify Music Recommendation
This project is one of the MADE Winter 2023/24 Projects implemented by our students. We run these projects every semester, please be in touch if you are interested in participating!
Author: Md Badiuzzaman Pranto
Project Description: The primary objective of this project is to build a song recommendation system based on the sentiment extracted from users’ social media posts and also by considering the users’ musical preference from their playlist. The initial phase involved training a Logistic Regression model using Twitter posts, followed by an evaluation of its performance in sentiment detection from social media content. The model demonstrated a commendable 76% test accuracy. Subsequently, leveraging the trained model, the sentiment of individual users is detected. The recommendation system then integrates both the user’s sentiment and their existing music playlist to provide personalized music suggestions. This dual consideration aims to enhance the relevance and emotional resonance of the recommended songs, contributing to a more tailored and engaging music recommendation experience.
Further Project Details:
- Report: https://github.com/prantoamt/made-template/blob/main/project/report.pdf
- Slides: https://github.com/prantoamt/made-template/blob/main/project/slides.pdf
- Repository: https://github.com/prantoamt/made-template
Reference: Md Badiuzzaman Pranto. Sentiment-Driven Spotify Music Recommendation: Leveraging Social Media Posts and User Playlists for Personalized Music Experiences. MADE WS 2023/24. Friedrich-Alexander-Universität Erlangen-Nürnberg: 2024.