🎯 Recommender Systems Collection

This repository contains two mini-projects built using Python and machine learning for content-based recommendation:

• 🎵 Song Recommender System
• 🎬 Movie Recommender System

Both projects demonstrate how similarity metrics and feature engineering can be used to recommend relevant items to users.

---

🔍 Project 1: Song Recommender System

📌 Description:

This project recommends songs based on audio features using content-based filtering. It analyzes attributes like danceability, energy, tempo, valence, etc., and uses similarity metrics to suggest songs that sound similar.

🧠 Methodology:

• Features Used:
  • Danceability
  • Energy
  • Valence
  • Tempo
  • Acousticness
  • Liveness
• Techniques:
  • Cosine Similarity or K-Nearest Neighbors
  • Feature scaling and preprocessing

✅ Output:

Given a song name, it returns a list of similar songs based on their audio characteristics.

---

🎥 Project 2: Movie Recommender System

📌 Description:

This system recommends movies based on genres, cast, director, and keywords using content-based filtering. The project builds a "tag" feature by combining various metadata.

🧠 Techniques Used:

• Text preprocessing and vectorization using CountVectorizer or TfidfVectorizer
• Cosine Similarity to calculate similarity between movies
• Metadata parsing (cast, crew, overview, genres)

🛠️ Input:

• User provides a movie name.

🎯 Output:

• Returns top 5–10 similar movies.

---

📁 How to Use

1. Clone the repository:

   git clone https://github.com/Prajjwal6969/recommender-systems.git
   cd recommender-systems

2. Install required libraries: pip install -r requirements.txt

3. Launch the Jupyter notebooks: jupyter notebook