Movie Recommendation App built in NextJS

Hosted here:

https://movie-vector-db.vercel.app/

A movie recommendation app leverages NextJS for a fast, server-side rendered frontend experience. At its core, the app utilises a vector database to store and efficiently query movie plot summaries as high-dimensional vectors. Similar to companies which use recommendation engines such as Amazon for products or Netflix for movies.

Key features:

User Interface: Clean, responsive design built with NextJS components for seamless navigation and movie selection.

Vector Embedding: Movie plot summaries are converted into numerical vectors using natural language processing techniques. Leveraged open AI's model: text-embedding-3-large.

Vector Database: Employs a specialised database (AstraDB) for storing and querying high-dimensional vectors. Integration with Open AI

Cosine Similarity: When a user selects a movie, the app calculates cosine similarity between the chosen movie's vector and all other movies in the database.

Recommendation Engine: Returns a list of movies with the highest cosine similarity scores, representing the most similar plot lines.

Performance Optimization: NextJS's server-side rendering and automatic code splitting ensure fast load times and a smooth user experience.

Comparison Page

This app provides users with an engaging way to discover new movies based on plot similarities to their favourites, offering a cool approach to content recommendation in the streaming era based on the latest techniques.