Project Medley 🎵

A basic music recommendation engine built with Python, Pandas, and Scikit-learn by using content based filtering techniques.

This script analyzes a dataset of songs, finding similar tracks based on their audio features (like danceability, energy) and their genre. It then combines this similarity score with a song's popularity to provide recommendations that are both relevant and popular.

---

How It Works

The recommendation logic follows these steps:

1. Feature Selection: It selects key audio features (danceability, energy, loudness, tempo, etc.) and the track_genre.
2. Scaling: All numeric audio features are scaled to a 0-1 range using MinMaxScaler so that no single feature (like loudness) can dominate the others.
3. Genre Encoding: The track_genre column is converted into a one-hot encoded vector using pd.get_dummies. This turns categorical genre names into a numeric format that can be used in similarity calculations.
4. Master DataFrame: The scaled audio features and the encoded genre vectors are combined into a single master_features_df.
5. Similarity Calculation: When a user provides a track_id, the script calculates the Cosine Similarity between that song's feature vector and all other songs in the master dataframe.
6. Hybrid Scoring: To improve results, the final recommendation score is a weighted blend of two metrics:
   • Similarity Score (Alpha = 87%): How similar the songs are.
   • Popularity Score (1 - Alpha = 13%): How popular the songs are. This helps to surface popular, relevant tracks and push down very obscure ones.
7. Cleaning: The final list is cleaned to remove the original song and any duplicate tracks.

---

How to Use

1. Prerequisites

• Python 3
• A dataset.csv file (see below)
• The Python libraries in requirements.txt

2. Setup

1. Clone the repository:

   git clone [https://github.com/HilariousSoupXD/Project_Medley.git](https://github.com/HilariousSoupXD/Project_Medley.git)
   cd Project_Medley

2. Create and activate a virtual environment:

   python3 -m venv venv
   source venv/bin/activate

3. Install dependencies:

   pip install -r requirements.txt

3. Data

This script requires a dataset.csv file in the same directory.

The file can be obtained from Kaggle: https://www.kaggle.com/datasets/maharshipandya/-spotify-tracks-dataset?resource=download

The CSV must contain the following columns:

• track_id
• track_name
• album_name
• artists
• popularity
• track_genre
• danceability
• energy
• loudness
• speechiness
• acousticness
• instrumentalness
• liveness
• valence
• tempo

4. Run the Script

To run the default example, simply execute the app.py file:

python app.py