Youtube Video Ranking: A NLP based System

Abstract

YouTube is an acclaimed video information source on the web among various social media sites, where users are sharing, commenting and liking/dis-liking the video along with the continuous uploading of videos in real-time. Generally, the quality, popularity and relevance of results obtained from searching a query are obtained based on a rating system. Now and then few irrelevant and substandard videos are ranked higher because of higher views and likes. To address this issue, we put forth a sentiment analysis approach on the user comments based on Natural Language Processing. The suggested analysis will be helpful in providing a desirable result to the search query. The effectuality of the system has been proved in this paper using a data driven approach in terms of accuracy.