• 

  Movie-Match

• About the project

Movie Match

Inspiration

Our inspiration for creating Movie Match starting from a common confusion while deciding what movie to watch. We wanted to save time and make it easy to choose the movie by building a movie recommendation system that provides personalized suggestions based on individual preferences.

What it does

Movie Match allows users to input a movie of their choice and receive recommendations for similar movies. Using text vectorization and cosine similarity techniques, the system analyzes movie data to generate relevant suggestions according to each user's movie taste.

How we built it

We developed Movie Match by collecting data from kaggle and preprocessing movie data, merging and cleaning it for analysis. Used text vectorization techniques like bag-of-words and used cosine similarity to find similarity between movie vectors to identify the top 5 nearest movies for recommendations. The tech stack is Python, Streamlit, NLTK (natural language tool kit), Pandas, Numpy , and Scikit-learn.

Challenges we ran into

One of the main challenge was to use Streamlit as it was new for us. Also processing and analyzing large datasets created difficulties but through research and experiments, we were able to overcome these challenges and implement the necessary system.

Accomplishments that we're proud of

We are proud to successfully built a functional movie recommendation system on my own that provides accurate and personalized suggestions to users. Also we were happy that we handled complex challenges and learn new concepts and technologies during the development process.

What we learned

Throughout the project of Movie Match, we learned how to use Streamlit for building interface of web applications, methods of text processing, vectorization, and similarity calculations. We also gained experience in handling the whole project for the first time by just the two of us.

What's next for Movie Match

In the future, I plan to modify Movie Match by adding features like user profiles, filter options, and more accurate recommendations by improving the logic. I also aim to improve the performance of the system and explore with other movie-related APIs for more movie recommendations.

Built With

• nltk
• numpy
• pandas
• pickle
• python
• scikit-learn
• streamlit