Skip to content

To calculate the number of a likes a YouTube video can have using prediction algorithms.

Notifications You must be signed in to change notification settings

harshita1017/youtube_like_prediction

Repository files navigation

YouTube_Video_Views_Predictor 📉 📈

Since YouTube's birth back in 2005, we all have witnessed the rapid boom in the popularity of this video hosting service company. As of 2021, there are approximately 1.86 billion YouTube users worldwide. Considering the wide reach that YouTube can offer to your content, many people have turned to content creators and are generating their income from it. Considering their huge numbers, it is very crucial for these creators to have insights regarding the performance of their videos in terms of popularity. This would encourage them to align their content delivery to have a good amount of views for the good.

TECH STACKS USED: 🌱

  • HTML, CSS, JAVASCRIPT for the frontend

  • PYTHON, AND FLASK for the backend

    Libraries used :

    • NUMPY
    • PANDAS
    • SEABORN
    • MATPLOTLIB
    • JSON
    • PICKLE

FEATURES OF THE MODEL 🌟

  • The user is required to enter the url of its video
  • Through Web Scraping, our model will extract important information like if the Comments & Ratings have been disabled or not , if the video has errors or if it has been removed
  • An interactive web platform to take the video's url
  • High accuracy achieved on the dataset of YouTube Trending Videos for the year 2019.
  • Machine Learning Model made using Random Forest
  • Output of the expected number of views of the video

CHALLENGES FACED 😟

  • Integrating the Machine Learning Model and Web Service
  • Choosing suitable attributes for a good prediction
  • Understanding Flask as we were completely new to it
  • Time management with the on going semester examinations

OVERCOMING THE CHALLENGES 🩹

  • Team work always overpowers the challenges faced
  • We followed th following resources to brace up in the technical domain

REFERENCES 📚

About

To calculate the number of a likes a YouTube video can have using prediction algorithms.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published