imdb-rating-classifier

An application that scrapes data from IMDB and adjusts rating based on some rulesets.


Keywords
imdb-api, rating-system, web-scraping
License
MIT
Install
pip install imdb-rating-classifier==0.1.3

Documentation

IMDB rating classifier

This is a simple IMDB rating classifier application that panalizes reviews in accordance with some pre-defined ruleset.

Overview

The application scrapes data from IMDB and adjusts the rating system according to some specific validation rules (review penalization).

The data is scraped from the IMDB charts API using the BeautifulSoup library.

The data structure of the parsed payload is as follows (example):

{
  "rank": "1",
  "title": "The Shawshank Redemption",
  "year": "1994",
  "rating": "9.2",
  "votes": "2,223,000",
  "url": "/title/tt0111161/",
  "oscars_won": 0,
  "penalized": false
}

We would then, extract the following fields, into a dataframe:

- rank (int)
- title (str)
- year (int)
- rating (float)
- votes (int)
- url (str)
- oscars_won (int)
- penalized (bool)

Using dataclasses, we can then, preprocess the data against some schema definition.

The rules are as follows:

schema = {
    "rank": {
        "type": "int",
        "min": 1,
        "max": 250,
        "required": True,
    },
    "title": {
        "type": "str",
        "required": True,
    },
    "year": {
        "type": "int",
        "min": 1900,
        "max": 2023,
        "required": True,
    },
    "rating": {
        "type": "float",
        "min": 0.0,
        "max": 10.0,
        "required": True,
    },
    "votes": {
        "type": "int",
        "min": 0,
        "required": True,
    },
    "url": {
        "type": "str",
        "required": True,
    },
    "oscars_won": {
        "type": "int",
        "min": 0,
        "required": True,
    },
    "penalized": {
        "type": "bool",
        "required": True,
    },
}

Requirements

  • Python>=3.8>=3.10
  • BeautifulSoup4
  • requests
  • pytest
  • tox
  • click
  • pre-commit
  • flake8
  • black
  • isort

and more...

Installation

For development purposes:

  • Clone the repository

    foo@bar:~$ git clone git@github.com/marouenes/imdb-rating-classifier.git
  • Create a virtual environment

    foo@bar:~/imdb-rating-classifier$ virtualenv .venv
  • Activate the virtual environment

    foo@bar:~/imdb-rating-classifier$ source .venv/bin/activate
  • Install the dev dependencies

    foo@bar:~/imdb-rating-classifier$ pip install -r requirements-dev.txt
  • Install the pre-commit hooks

    foo@bar:~/imdb-rating-classifier$ pre-commit install

For usage:

  • Install the dependencies and build the wheel

    foo@bar:~/imdb-rating-classifier$ pip install -e .

The application is publicly available and published on PyPI and can be installed using pip:

foo@bar:~$ pip install imdb-rating-classifier

Usage

  • Display the help message and the available commands
foo@bar:~$ imdb-rating-classifier generate --help
Usage: imdb-rating-classifier generate [OPTIONS]

  Generate the output dataset containing both the original and adjusted
  ratings.

  An extra JSON file will be generated alongside the csv file

Options:
  --output FILE               The path to the output file.
  --number-of-movies INTEGER  The number of movies to scrape.
  -h, --help                  Show this message and exit.
  • Run the application with the default number of movies (20) and the default output file (data.csv)
imdb-rating-classifier generate
  • Run the application with a specific number of movies
imdb-rating-classifier generate --number-of-movies 100
  • Run the application with a specific number of movies and a specific output file
imdb-rating-classifier generate --number-of-movies 100 --output some_name.csv

Testing

  • Run tests and pre-commit hooks
foo@bar:~/imdb-rating-classifier$ tox

CI/CD

The application is automatically packaged and distributed to PyPI, It is also automatically tested using tox as an environment orchestrator and GitHub Actions.

TODO

  • Add more tests
  • Add more validation rules
  • Add more documentation
  • Add more features
  • Publish the package on PyPI
  • Add oscar awards or nominations for the movies
  • Add a version switch for the cli

License

MIT License

Author

Marouane Skandaji