Movie Recommendation Service

An end-to-end movie recommendation system built with Svelte, SvelteKit, PostgreSQL, and pgvector.

About Database & Dataset

This project uses a processed version of the latest MovieLens dataset. The original dataset includes a rich collection of movie ratings and tags from thousands of users. For this project, the data was processed and inserted into a PostgreSQL database, optimized for efficient querying and recommendation generation. To get more details about processing and inserting the data, please refer to github repo.

Overview

This project is a hybrid movie recommendation service that leverages both content-based and collaborative filtering techniques. It utilizes a vector database for efficient similarity searches, providing users with personalized movie suggestions.

Key Features

• Hybrid recommendation system combining content-based and collaborative filtering
• Vector similarity search powered by pgvector
• Interactive user interface built with Svelte and SvelteKit
• Real-time movie search functionality

Technology Stack

• Framework: SvelteKit (full-stack)
• UI Library: Svelte
• Server-side Logic: Node.js (via SvelteKit API routes)
• Database: PostgreSQL with pgvector extension
• API Integration: TMDB (The Movie Database) for movie details and posters

Getting Started

1. Clone the repository:

   git clone https://github.com/hasibuldog/movie_recommender_svelte.git
   cd movie-recommendation-service

2. Install dependencies:

   npm install

3. Set up environment variables: Create a .env file in the root directory and add the following:

   DB_URL=your_postgres_connection_string
   TMDB_API_KEY=your_tmdb_api_key
   

4. Start the development server:

   npm run dev

5. Open your browser and navigate to http://localhost:5173 to see the app in action.