raghavs16/movie_recommendation_system

Purpose and Runtime Topology

FilmoBot is a React single-page movie chatbot backed by a Flask API. It converts conversational movie requests into structured search intents, retrieves and enriches movie data, and supports authenticated persistence for chats, profiles, and watchlists. The browser runs from client/src/index.js; Flask is started through server/app.py.

Frontend Application Shell

client/src/App.jsx owns React Router routing and application-level state across the chat and account-oriented screens. The primary user journey begins in client/src/ChatPage.jsx; onboarding and authentication are handled by client/src/Splash.jsx, client/src/Welcome.jsx, client/src/Login.jsx, and client/src/Signup.jsx. API targeting is centralized in client/src/config.js.

Chat and Recommendation Presentation

client/src/ChatPage.jsx sends user prompts to the backend, renders conversational results, and participates in chat-history behavior. client/src/MovieCard.jsx is the reusable movie-result boundary: it presents returned metadata and exposes trailer and watchlist actions. client/src/Sidebar.jsx provides navigation and access to persisted chat sessions.

User State and Persistence UX

The client stores JWT authentication and configurable preferences locally; settings include language, result count, notification choices, and chat-history behavior. client/src/SettingsPage.jsx owns this preference surface, while client/src/WatchlistPage.jsx and client/src/ProfilePage.jsx consume backend-backed user data. The backend SQLite database is represented by server/filmobot.db; API and database routes/models reside in server/app.py.

Backend API and Account Boundary

server/app.py is the Flask composition point: it exposes the API, performs JWT-based authentication, handles account/profile/watchlist/history operations, manages password-reset OTP email flow, and accepts profile image uploads. Uploaded images cross into server-local storage at server/static/uploads/. Authentication uses Flask-JWT-Extended, passwords use Flask-Bcrypt, and reset email uses Flask-Mail with Gmail SMTP.

AI Intent-to-Search Pipeline

Movie-query handling enters server/app.py and delegates intent extraction to server/ai_processor.py. That module calls Groq’s LLaMA 3.1 8B API for structured intent and supplies a rule-based fallback when the LLM is unavailable. The resulting intent selects discover, direct search, similar-movie, person-filmography, or song-to-movie paths; fuzzy and phonetic recovery for misspelled titles/names is implemented in server/movie_search.py.

External Movie Data Integration

server/tmdb_client.py is the main TMDB boundary for discovery, search, recommendations, language/genre/mood mapping, trailers, and streaming-oriented metadata. server/tmdb_retry.py wraps TMDB access with retry behavior. Backend responses are enriched with OMDb data such as IMDb ratings and cast, then ranked before returning to the client. Required external services are Groq, TMDB, OMDb, Gmail SMTP, and client-side EmailJS for client/src/ContactPage.jsx.