MtvInsight

[] []

Description

MtvInsight is a movie and TV shows prediction website that leverages a dataset comprising 10k movies and 11k TV shows manually collected from The Movie Database. The prediction process involves converting each movie or TV show into a vector and determining their similarity based on the direction of these vectors. A small angle between vectors indicates similarity in content.

Inspiration

I wanted to make a project that i can use and also use machinelearning init

Quick Start

• MtvInsight

Usage

• dataset option 1: installing manually using tmdbapi ==Warning== it takes alot of time to fetch all the data manually. code for fetching the movies, tvshow data from code for preparing and processing the similarities in jupiter notebook movie, shows

  option 2: you can access the the dataset directly from tmdb dataset a collection of 5k movies code for preparing and processing the similarities in jupiter notebook movie shows

• doployment

frontend(react) --> vercel backend(flask) --> google-cloud database(mongodb) --> atlas

• loading svgs

i used the cool library react-loader-spinner

process

first: download the dataset

1. manually (warning takes alot of time)

   • you can see code for fetching the datasets using the code at fetch_movies fetch_tvshows //note: change (tmdb.api_key) with your actual api key you can get it from tmdbapi

2. you can use already existing movies dataset tmdb dataset

   • note the data neet more processing in the next step

second: process and calculate similarity

• if manually installed the data use this for calculating similarities movie.ipynp show.ipynp

• if manually installed the data use movie_pre.ipynp

stack

backend

• flask

frontend

• react
• vanilla css

Contributing