One of the biggest issues that video-sharing sites such as YouTube face is dealing with issues surrounding copyright infringement.
Even though YouTube users are instructed to only upload content that belongs to them and that they have the right to use, copyrighted content is still uploaded thousands of times a day. Under the Digital Millennium Copyright Act (DMCA), YouTube isn't responsible for the copyright violations of its users, provided the company removes that content when notified by the rights holders.
Still, thanks to a barrage of lawsuits targeted at the service (including Viacom's $1 billion lawsuit against YouTube), the company has had to turn to technology to figure out ways to help deal with copyright issues, notifications and takedown requests.
To help counter these lawsuits and better work with rights holders, YouTube created its own content detection system known as Content ID.
Content ID Overview
Content ID is a system that allows rights holders to identify content that is comprised partially or entirely of their content.
YouTube started building Content ID in 2007 after it decided that video and audio fingerprinting technology could help the company in its various legal battles. The problem was that existing systems weren't good enough for YouTube's uses.
This video gives an overview of the content ID system.[embedded content]
To use Content ID, rights holders submit copies of content or ID files that are then run against user uploads. Copyright holders can assign various policies to their content in the event of a match. The policy options are:
Block — This means that if a content match is found, the video will not be viewable.
Track — The video can stay online, but the content owner will be able to track how many views it receives and from where.
Monetize — Rights holders can choose to serve ads on the content and they will receive revenue from those ads.
Policy options are available on a regional basis, which allows rights holders to block content in some regions, while keeping them accessible in others.
YouTube users can find out about any copyright notices or Content ID matches by visiting the "Copyright Notices" section of the video manager.
Does it Work?
So how accurate is the system? According to a 2009 study of YouTube's audio fingerprinting technology, it's quite robust. While Content ID doesn't work in all cases (though we should stress that is has been improved upon over the last three years), it is quite adept at identifying content from various reference files.
Likewise, with its video fingerprinting technology, YouTube tends to identify and assign policies to lots of content very quickly.
Moreover, YouTube and rights holders are making big money off of Content ID. According to The New Yorker, more than a third of YouTube's revenue comes from Content ID. This represents the best case scenario for rights holders, users and service providers because it allows content to stay online, while rights holders still generate revenue.
Content ID is a big part of YouTube's content identification strategy, but it isn't the only way the company works to protect rights holders. YouTube has created what it calls "Copyright School" as a way to inform users of the rules regarding copyrighted content and YouTube.
The program also works as a way for users who have had multiple copyright violations (or "strikes") against their account to remove those strikes and stay in good standing with the service. Users who are repeatedly found in violation of YouTube's content policies will be banned from the site, and their videos will be disabled/deleted.
What About Fair Use?
An automated system is a great way for a large service like YouTube to deal with copyright issues en masse. Still, some policy experts criticize the service for not having a way to account for what falls under the fair use doctrine. Rules regarding fair use (or ways to use copyrighted content legally, without having to obtain written permission) vary based on the type of content, the purpose of content and the location of the user. In the case Lenz v. Universal (where a mother sued Universal Music Group for forcing YouTube to take down a video of her baby dancing to a Prince song), courts ordered that rights holders had to consider fair use before issuing a takedown notice.
Still, the law regarding what is and what is not considered fair use is a very murky area, especially when it comes to user-generated content on sites such as YouTube. Some policy experts have criticized YouTube for not crafting Content ID in a way that looks out for fair usage claims.
Source : http://mashable.com/2012/02/17/youtube-content-id-faq/