Home // HUSO 2020, The Sixth International Conference on Human and Social Analytics // View article

YouTube Video Categorization Using Moviebarcode

Authors:
Recep Erol
Rick Rejeleene
Richard Young
Thomas Marcoux
Muhammad Nihal Hussain
Nitin Agarwal

Keywords: Moviebarcode, Video Categorization, YouTube, Social Computing Tool

Abstract:
Every minute more than five-hundred hours of video content is uploaded to YouTube, and we can only expect this number to increase. Although YouTube is the most popular video sharing website, studies conducted on this platform are sparse. The lack of effective video analysis techniques presents a tedious challenge for researchers and has hindered overall research on this platform. Due to this, research conducted on YouTube primarily focuses on analyzing text-based content or video metadata. With recent advancements in the development of moviebarcode, a technique that shrinks a movie or video into a barcode, we have developed a tool designed to extend the capabilities of moviebarcode as a forensic technique for systematically categorizing YouTube videos. We use moviebarcode to summarize an entire YouTube video into a single image to help users understand a video without even watching it and later use cluster them based on similarity. We analyzed six video collections and using moviebarcode only and without looking at the video content, we were able to achieve an accuracy of 75%. Using our method, an analyst can quickly group videos into bin computationally reducing the overhead of manually doing it

Pages: 15 to 19

Copyright: Copyright (c) IARIA, 2020

Publication date: October 18, 2020

Published in: conference

ISSN: 2519-8351

ISBN: 978-1-61208-800-6

Location: Porto, Portugal

Dates: from October 18, 2020 to October 22, 2020