Behind the scenes of the DMCA takedown notice


By Elizabeth Bagot

If you are in digital media, chances are you have heard about the creative industry’s clamoring for Google to apply stricter anti-piracy measures to its search results. Google recently announced the receipt of its 235 millionth DMCA takedown notice, up from a mere 50 million in 2012.

Why is this happening now? Why are copyright holders detecting – seemingly suddenly – such an enormous quantity of illegal material floating around online?

The answer can be summed up in one word: technology.

Just a few years ago, the process of submitting a takedown notice involved a person seated at a computer, doggedly entering various keyword combinations, manually gathering a list of illegal URLs, and notifying the infringer, one letter at a time.

These days, specialized technology has rendered the human element obsolete. Sophisticated servers have the power to detect hundreds of millions of URLs at a time and send thousands of takedown notices a day.

This technology, known as automatic content recognition, uses a “fingerprint” to identify licensed digital media. A fingerprint is a unique identifier that amalgamates parameters such as duration, color and lighting, scene changes, and other specifications into an ID tag for content. This fingerprint is compared with URL content found online to determine the existence of infringing material. If the comparison yields a positive match, action can be taken to have the material removed.

International tech company WebKontrol, based in Russia, takes a so-called “mapping approach” to automatic content recognition. This involves graphically displaying scene changes in a video clip, then superimposing that visual representation over a licensed fingerprint. If the two images match, the URL is positively identified as a copy of the original content.

The mapping approach’s value lies in its ability to handle video distortions such as cam rips, subtitles, splash screen, changes in brightness, etc., with exceptional accuracy. This virtually rules out false positives, ensuring that all takedown notices are legitimate. While powerful, automatic content recognition alone is not sufficient to tackle Internet piracy. Content has to be found before it can be compared and matched with fingerprints. Technology is needed to automatically scan the vast expanse of the web and interpret massive quantities of information.

Sophisticated content search engines perform this function. Here is a look at the numbers. WebKontrol’s Content DiV technology “crawls” 300 million webpages per day for content, simultaneously extracting detailed information for each relevant URL, including upload date, view count, title, duration, and quality. After passing through the initial filtration phase, 35 million URLs are submitted for the automatic content recognition phase.

The system performs 700,000 verifications per day. Action – in many cases, the submission of a DMCA takedown notice – is taken exclusively for comparisons that yield positive matches. That is a staggering burden for Google, but a boon for copyright holders that are serious about cracking down on piracy.

YouTube has already taken action. Its pre-upload filter technology, Content ID, compares all incoming user content against a database of licensed fingerprints, flags infringing material, and bars that content from being uploaded.

Content ID has been both a blessing and a curse for copyright holders. By virtually eradicating piracy on YouTube, it has shifted the piracy burden elsewhere – particularly to the Google search engine. Users can often find their desired content with a simple keyword search in Google. This situation has led to today’s controversy and the veritable avalanche of DMCA takedowns onto Google’s shoulders.

One fact is undeniable: as WebKontrol’s Content DiV and YouTube’s Content ID show, the technology does exist to address piracy on a massive scale. Whether or not Google is ready and willing to – and whether, from the point of view of net neutrality, it even should – cater to corporate interests is a question for another article.

About author

Elizabeth Bagot consults for key accounts at WebKontrol.

About company: WebKontrol is a leading tech company that offers full-spectrum digital media content protection solutions on the Internet and provides customized business analytics to inform the investment decisions of copyright owners. Its DiV Action service utilizes breakthrough Content DiV and Digital Fingerprinting technology to detect, identify, and verify pirated content, the removal of which opens up monetization opportunities for copyright owners. WebKontrol’s clients include nearly all of the Hollywood studios and global software producers.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s