WE'VE BEEN TALKING, READING, AND are inundated with so many promises of a world of digital media content that is or will soon be at our fingertips. True enough, there is an amazing amount of audio and video content to be sought and sifted through. But, let's face it-- finding what you want in, say, an hour-long television interview is still really, really hard. First, you have to watch the video (yes, I know, what an inconvenience...). If you're fortunate, someone has tried to make it easier for you by indexing the interview with selected keywords that you can jump to. That works, as long as what that "someone" has indexed has some relation to your interests. But, if it's an interview where what you really want are only the sections where the interviewee mentions, say, "Watergate," and those sections have not been tagged for you, there is no other solution than to watch the content in a linear mode. And with this ever-growing sea of digital media at our fingertips, text continues to be the dominant method by which we search for content. In some cases, image recognition software may allow us to request comparisons-the "find me something that looks like this"-functionality that is certainly helpful. But, content publishers can easily be stymied by trying to figure out affordable methods to use which will result in their content being more accessible to search requests.
One of the more interesting product announcements in October was the introduction of EveryZing's MetaPlayer (www.everyzing.com). EveryZing, headquartered in Cambridge, MA, is focused on creating applications that result in an enhanced discovery of online audio and video content.
Based on core intellectual property in speech-to-text technology and natural language processing, EveryZing counts among its customers NewsCorp, Disney, CBS, Cox, and Thomson-Reuters.
One of the fundamental aspects of finding content that's relevant to you is the method by which you must search for that content. And that, of course, has traditionally been by using keywords. However, as already outlined, who determines what keywords are relevant? How many keywords should be indexed? And, with the sheer amount of digital media growing each day, simply adding people to manually enter keywords will clearly not scale. EveryZing, however, uses its core IP to generate metadata from audio and video content which, in turn, can be used for assisting the search process.
If we return to our earlier example of the television interview, if you had the entire transcript of the interview indexed and made available to you, searching for relevant words that are indexed to the exact video and audio moment becomes a far easier task.
As EveryZing's CEO, Tom Wilde outlined to me, MetaPlayer preserves for the content owner (publisher) the control of content, brand, and the monetization of that content. At the same time, the consumer has more control over content by virtue of having more meaningful search criteria that can be fulfilled.
EveryZing's application suite, consisting of ezSEO, ezSEARCH, and MetaPlayer are excellent examples of the functionality and packaging necessary to make search, retrieval, tracking, and monetization of content a reality. EveryZing's MetaPlayer and the technologies that power it can be integrated into third party applications and to extent "chromeless" players.
The bottom line: an impressive product set focused on addressing the vexing problem of searching through audio and video files which are taking an ever-increasing role in all of our online experiences.
| Tom Ohanian is Chief Strategy Officer, Signiant Corp. Tom was on the founding team of Avid Technology and is an Academy Award® and two-time Emmy® recipient |
No comments:
Post a Comment