YouTube Data Analysis and Prediction of Views and Categories

Abstract

Abstract: There's enormous growth and fashionability of YouTube. It has all implicit to move billion of lives encyclopedically as the number of observers is growing day by day. Nearly billions of vids are watched on YouTube every single day, generating a huge quantum of data daily. YouTube data is actually in unshaped form, so there's great demand to store the data, process the data and assaying the data. This analysis will help in discovering how people are performing on YouTube, one can fluently identify what contentworks best on YouTube. The primary purpose ofthis design is to find how real time data can be anatomized to get the rearmost analysis and trends in YouTube. The analysis is done using stoner features similar as views, commentary, markers, likes, and dislikes. Analysis can be performed using algorithms like direct retrogression, bracket and other machine literacy models and python libraries like pandas, matplotlib library to classify the YouTube data and gain useful information. Keywords: YouTube, titles, Prediction ofcategories, Data analysis.