Skip to content

A novel approach towards video-ranking using intent and relevance feedback

Notifications You must be signed in to change notification settings

EleetCodeR/YTVideoRanking

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

36 Commits
 
 
 
 
 
 

Repository files navigation

A novel approach towards video-ranking using intent and relevance feedback

Abstract

The “Internet” today has rapidly morphed into a large platform for learning purposes. As a result, large amounts of instructional videos are being produced every day and are uploaded to video repository platforms like “YouTube”. The video-ranking methodology employed by such platforms largely focuses on video-description and user-ratings as a direct criterion. This often leads to less relevant videos being ranked higher than others and creates a “search-intention and relevance gap” between users’ search query and video results that are shown. As these platforms also allow users to express their opinion about videos, in this paper, we propose a video re-ranking methodology IRF (Intent and Relevance Feedback) to improve the rank of the relevant videos, where we emphasize determining the impact of using unexplored video aspects, like “user comments” and “video’s content” in the video ranking and index calculation