Webinar Recap: Video is the New Document

Think video is the defacto form of communication on the Web? You're not alone.  The explosion of online video has inundated the Web in much the same way documents once resulted in information overload. Video is, in a sense, the new document.

Video That is Easy to Find has Value

The popularity of video tutorials alone shows how powerful online videos have become, but businesses can do much more with video than just training, Nate Treloar, VP and General Manager of Enterprise at RAMP said during our recent webinar.

Treloar showed off the technology behind RAMP, a SaaS-based video platform, and talked about how making video discoverable and exposing it to search could pay off for businesses during the Video is the New Document webinar.

The infrastructure for supporting video (like cheaper storage) has improved so much, and the barriers to creating video have lowered greatly. As a result video content is running rampant across the Web.

Companies can use it for training and HR, and for meetings and communications of course, but video can also be productive on the customer side, and not just for media companies, Treloar said.

Make video easy to find, and it will have value. Companies can make video more engaging by allowing for navigation within a video he said. Video also needs to be versatile, meaning it should work on mobile devices.

One of the challenges of video is inherent in the format: it usually lacks textual documentation, which is necessary to open video up to search and potential engagement. RAMP makes video actionable by converting the audio into text, which then becomes metadata. This opens the video via the text to be exposed to search, making it more valuable.

Video as a Document

Video has become cumbersome to some companies in a way that documents once were (and still are), but that isn't the only way video is the new document. Once the sound from a video has been converted to text, that video now literally has a document associated with it, and it can be stored and processed in the same way, Teloar said. This allows easy retrieval of video assets through a search of the associated metadata.

The metadata is time-coded, meaning the searches can be performed within the video itself, pinpointing the exact time when key phrases or words were used. Once a video is transcribed, the transcript can be hovered over, and a time code will pop up to show exactly where that phrase can be found. Additionally, when a thumbnail of a video is hovered over, a short preview spins up without even having to launch the video.

These metadata powered videos also allow for synchronizing polls and presentations with relevant videos through the MetaQ component. This component is rule-based, using the metadata to link contextual materials -- while Treloar emphasized the uses in a training and media setting, the potentials of this for digital marketing should be obvious. RAMP offers a human transcription service for those who want more specific metadata for these purposes.

Multilingual Transcriptions, Face Recognition and More

During the question and answer period, the subject of multilingual transcriptions came up, and Treloar noted that RAMP does use language models, and so it can learn virtually any language a customer might need. More or less out of the box, RAMP offers English, Spanish and a few other languages.

Face recognition also came up during the question and answer segment, and it turns out only a few in the security sector have shown much interest in that technology, Treloar said. Face recognition just isn't that advanced overall, however, he said. It could be done with RAMP, but the technology would be via a third party integration. Another third party integration would be for translating videos, something that is different from transcribing in various languages.

Other questions were on whether RAMP was on premises or cloud (cloud only for now), and on speaker turn recognition. That is, can it tell which speaker is talking and when? This was another instance where RAMP could learn to tell who is speaking when, but has to be taught that.

We've included the entire session below, so feel free to give it a watch and leave us a comment on how transformative you see video being or what your favorite kinds of videos are.

Image courtesy of Flickr user mikecogh via Creative Commons license.