Therapy Tunes is an innovative project exploring the intersection of music and mental health. Developed collaboratively by three colleagues, this platform offers personalized song recommendations aimed at enhancing users' emotional well-being.
Our primary aim is to recommend songs that align with users' musical preferences while addressing their mental states. Therapy Tunes harnesses music's therapeutic potential to aid in managing common issues such as anxiety, depression, and insomnia.
-
Data Collection: Gather user information to assess anxiety, depression, and insomnia levels, alongside musical preferences.
-
BPM Determination: Calculate an ideal BPM (Beats Per Minute) range based on collected data.
-
Data Analysis: Analyze an extensive song database using determined BPM and other parameters, employing clustering and PCA (Principal Component Analysis) for song categorization.
-
Song Recommendation: Identify the most suitable song segment based on the user's profile and recommend songs accordingly.
-
Astrological Integration: Incorporate current horoscope interpretations, extracted via web scraping, to enhance song recommendations.
- Python
- Machine Learning Models:
- XGBoost
- AdaBoost
- SVC
- RandomForestClassifier
- LightGBM
- Machine Learning Algorithms:
- Clustering
- PCA
- Web Scraping: Beautiful Soup
- Data Analysis and Visualization Tools
- Integrate user feedback for continuous improvement of the recommendation system
- Develop a mobile application
- Expand the music database
- Collaborate with mental health professionals to enhance platform efficacy
https://therapytunes.streamlit.app/
For inquiries or suggestions, please reach out to us via our LinkedIn profiles: