Current Articles | RSS Feed RSS Feed

Google, Yahoo!, Pixsy Video Search Indexes Explode with Gotuit

  | Share on Twitter Twitter | Share on Facebook Facebook | Submit to Digg digg it |  Share on LinkedIn LinkedIn 

The Internet has evolved from text, to images, to video and video publishers are hungry for solutions that will better drive search traffic to their video libraries.  However, accurate, relevant, and useful video search remains a difficult problem that the biggest technology companies in the world are still trying to solve.

The video search engines like Google, Microsoft, Yahoo!, Pixsy, Blinkx, Truveo, etc. all have differing approaches, but share the same fundamental unit of video in their indexes - the video asset.  At Gotuit, we thought about the problem differently, and use premium scene-based metadata to vastly increase the amount of information that describes a video within any search index.

First announced last month with Pixsy, Gotuit customers can use our Video Metadata Management System to author rich, structured metadata about each individual scene in their original videos.  We then syndicate an mRSS feed of the scene-level metadata to all the major search engines.  Importantly, there is no video editing required so it is much simpler for the publishers to operationally manage whole assets than thousands of small clips.  The metadata instead defines ‘virtual clips’ so that they can be indexed by the search engines using their current algorithms.

The video search impact is dramatic.  Imagine an example of a 30 minute asset with six searchable metadata elements that now has a metadata file describing the 15 scenes within the asset, and each scene being described with ten elements.  The result is a 25X increase in the amount of information that now exists in the video search engines.

Not only is there more data, but the data is more accurate and drives more relevant results due to the power of the human-authored metadata.  Details like specific characters, locations, events, emotions, scene types, and other keywords are now present.  Video is very subjective, and the person describing the scenes can add information that simply could not be derived by an automated solution, but would be very valuable to match against human-entered search terms.

Obviously, the ultimate goal of better video search is better monetization.  As Pixsy’s CEO Chase Norland said, “Allowing our search technology to work against each scene inside an asset, instead of just the asset itself, grows the search index for each video by an order of magnitude for the user to better search against and the website to monetize.”

Not only does Gotuit metadata explode the amount of information in a video search index, all of the scene boundaries are ad insertion avails.  Plus, the scene metadata can drive more targeted ads - both in-page banner and in-stream video ads.  Having seen the incredible economy that has arisen over targeted text-based ads, is there any doubt that there will be a similar evolution of the video-based ad market?

Premium, scene-level video metadata is what unlocks this opportunity for video search and content monetization, and look for more developments in these areas here at Gotuit in the near future.


Comments

There are no comments on this article.
Comments have been closed for this article.

Stay Connected

Subscribe to our blog via email

Your email: