We know the major search engines do a great job of indexing text to make it easier for you and me to search stuff, right? Putting "crawl blockers" aside for a moment, if it's a document with text in it, and it's on the web somewhere, then it has likely already been indexed. If so, it can be found through a boolean keyword search.
But what about words that are embedded within the content of the video or audio file itself?
To a degree, search engines have traditionally been able to index multimeda files primarily through the text-based keywords, tags, titles and descriptives we associate them with when they're uploaded. But, until recently, there hasn't been a widely available method to index the spoken word within those media files. Now, with Everyzing's ezSearch service, you can.
Recently, Everyzing, launched a service that converts all audio in a media file to text so they can be indexed made searchable. But that's not all. One slick feature is in the way the search results are presented.
When video, say, is found with the keywords you're looking for, the results page returns not just a simple link to the video, but also a pointer and timecode to the place in the video at which the keywords were found. The image below, for example, is a snapshot of search results on Everyzing's search site on keywords "New York" AND "bicycle". Notice the pointer and timecodes. (Click image below to zoom in.)
I think that's slick. It makes for some compelling applications for advertising. But more than just advertising, it also makes for some compelling applications in the eLearning and real estate professional circles I run around in.
For example, in much of the online learning content I help produce, it's not uncommon to have the audio/video content span several hours of total runtime. While that length is made much more "consumable" to our learners by breaking it up into manageable "learning chunks" of about 10-15 minute segments, it can still be a chore to try and recall in which segment a particular example was given or the segment in which a reference was made.
With ezSearch, all the learner needs to do is recall a spoken phrase and the search results will show the video segment, timecode and pointer to the spot in the video stream at which it was spoken.
Everyzing has implemented their technology while working with boston.com and weii.com. You can click the respective links above to see customer examples, or you can also search Everyzing's search site at search.everyzing.com to get a feel for how it works.
Definitely something I'll be recommending for my future rollouts of our online designation courses.
Check it out. Feel free to comment and let me know what you think.
