This project is a content based recommender system which calculates the similarity matrix of 4805 movies in the TMDB database using cosine similarity. It then gives top 5 recommendations based on highest similarity scores (between -1 and 1)

Project link - Streamlit Cloud project

If the above link doesn't work, then please follow the instructions given below from here to setup the project in your system.

1. The 2 .csv files in the dataset were merged
2. All important columns were retained
3. Transformations using functions were performed to convert all column values to lists
4. The columns were merged to create a "tags" column
5. Transformations like lowercase, stemming, removing stopwords was done on the "tags" column
6. Cosine similarity matrix was calculated and a "recommend_movie" function was created to recommend the movies

1. A PyCharm project was created with a virtual environment
2. movies.pkl and similarity.pkl files were generated from the .ipynb file itself and copied to the PyCharm project folder
3. A profile was created on TMDB to generate an API key to fetch posters of recommended movies
4. app.py was designed to fetch posters from TMDB and generate movie names
5. Movie names and posters were displayed
6. App was deployed to Streamlit Community Cloud

1. Clone the repository:

   git clone https://github.com/rohanrvpatil/movie-recommendation-system.git

2. Install the required packages:

   pip install -r requirements.txt

3. Get a TMDB API key from here and store it in .streamlit/secrets.toml:

   # .streamlit/secrets.toml
   [general]
   tmdb_api_key = "your_tmdb_api_key"

4. Access the API key in your app.py:

   import streamlit as st
   
   tmdb_api_key = st.secrets["general"]["tmdb_api_key"]
   # Use the API key to fetch data from TMDB

5. Run the Streamlit app:

   streamlit run app.py

1. TMDB 5000 Movie Dataset
2. API key creation
3. YouTube link of project