/FilmFlix

Microsoft Engage program 2022

Primary LanguageJupyter Notebook

title emoji colorFrom colorTo sdk sdk_version app_file pinned
FilmFlix
📈
gray
red
streamlit
1.9.0
app.py
false

FilmFlix

Link to the app: https://filmflix-0.herokuapp.com/

Overview

Home

It contains posters of Top 6 most liked and most popular movies.These movies are sorted based on number of like and popularity measures.

Recommendation based on similar movies:

This works based on content based recommender system using cosine similarity.Cosine similarity is a metric used to measure how similar two items are.

Recommendation based on user profile:

This section first performs exploratory data analysis on the dataset and some useful plots are displayed.

Then it builds three ml models:

  • KNearestNeigbour

  • GradientBoostingClassifier

  • LogisticRegression

Then it compares the accuracy of three models and suggests genre of the movie based on high accuracy.

Recommendation based on Genre

This section searches and filters the five genres of the movie.

  • Action
  • Romance
  • Thriller
  • Comedy
  • Drama

Recommendation based on your mood

This section recommends music based on language, mood and favourite singer of the user.

TechStack

1.Python

2.Streamlit

3.TMDb api

4.Html (inside streamlit.components.v1)

How to run the project?

1.Clone or download this repository to your local machine.

2.Then install the virtal environment in the file directory using the command: python -m venv my_venv.

3.Activate the virtual environment by the placing the relative path of Activate.ps1 in my_venv: my_venv\Scripts\Activate.ps1

4.pip install -r requirements.txt

5.streamlit run app.py

ScreenShots

Home