Jennifer Zaino
SemanticWeb.com Contributor

EveryZing — whose search-engine-optimization expertise lies in its ability to understand the semantic aspects of publishers’ content through the consistent marking up of multimedia assets — said last week that Internet TV company Revision3 is using its solutions to provide enhanced search and discovery of the site’s broadcast-quality programming.

Founded by Digg.com founders Kevin Rose and Jay Adelson, Revision3 focuses on the technology space, with video programming aimed at consumer electronic products, video games, social bookmark hot topics, and hackers (in the good sense of the word), as well as comedy, music and movies.

The Internet TV channel, just a few years old, has about 11 shows in its programming schedule, and has accumulated about 2,000 episodes in its video library to date.

“As we sat down to look at the web site and functionality for users, it was clear there was no easy way to find an episode you saw that mentioned the Nokia phone that you’re now thinking about buying, for instance,” says Ron Richards, senior director of marketing and product management at Revision3. “We can tag our video posts with metadata, and keywords, and so on, but a 45-minute episode of a TV show is not well-represented by metadata.”

In search of more state-of-the-art search technology, Richards chose EveryZing’s ezSEO and ezSearch shared-hosting solution because of its ability to convert speech to text, making it possible for users to search not just on metadata but on exact words spoken to find the episode they’re looking for. Richards says the solution will make it easier both to search on Revision3′s site for episodes that feature specific content and via search engines like Google, as well, which can crawl the tagged and indexed text output derived from the video content to make information easier to find by the public at large.

So far Revision3 has completed the first phase of the project, which encompassed making all 2000 episodes to date indexed and searchable.

“Now we are further deepening this to integrate search on every part on the site, as well as make the transcripts of text searchable so you can jump right to it in the video once you find the word you’re looking for,” says Richards.

In addition to helping connect users with the content they want, better search matters to Revision3 on the revenue front, as well.

“This exposes more of our back catalog to users in an easy way,” he says, ideally driving them to watch more content than they might have been expecting to. “And the more they watch, the more ads we can deliver and monetize our shows.” The solution has only been live a couple of weeks, and Revision3 hasn’t seen a dramatic increase in page views, but they are slowly starting to creep up, Richards says. Revision3 says it plans to continue to find ways to leverage the data the solution affords in other ways — for example, understanding search results to identify the most popular searches or hot tpics to more efficiently do the work of serving up to users the exact content they are likely to want to view.

In the Web 3.0 world, says EveryZing CEO Tom Wilde, it’s imperative for publishers to ensure that each content object, whether an article, video or audio clip, is well marked up, so that content is integrated with the Semantic Web in the future. “When you think of the virtuous circle, if content is more easily found it is more easily consumed, and that kicks off the second leg of this, the social graph, which can’t happen if content can’t be discovered in the first place. They reinforce each other.”

On the roadmap for EveryZing this year, he adds, is continuing to leverage the technology’s ability to create first order metadata into second order metadata that drives metadata. As an example, he cites the ability to assign latitude and longitude to a video clip based on what is mentioned in the transcript, such as the word Indianapolis, as a step in weaving that content into the geographic aspect of the semantic web.

“Imagine using Indianapolis as an entity — the latitude and longitude for Indianapolis could be tagged to that clip and that clip discovered using a map interface.” He also expects that in 2009 EveryZing’s services, which have been focused on fairly large sites with lots of content, will also be tailored to fit the needs of mid-sized publishers with more self-service style applications.

Semantic Tech & Business Conference Returns to San Francisco

Semantic Tech & Business Conference returns to San Francisco in June! Join us from June 3-7 for complete coverage of Big Data, Linked Data, Extreme Information Management, and Semantic Web. From breakthrough approaches to solving business problems to the big data implications of fast–evolving technologies, SemTechBiz provides you with an unparalleled interactive experience and delivers tangible business value. We're offering a special early rate when you register by February 17. Sign up now!