Searching A/V content
Stefanie Olsen over at CNet has been looking at the approaches to searching for A/V content being put forward by Yahoo, Google and Microsoft. You can see the business need for doing this. Audio-video content is increasingly used over broadband networks. And more and more people are going broadband. There were 4.5 million high speed broadband subscribers in the UK at
the end of June 2004 a figure which should have increased to 5.5
million with the connection rate of 50 000 week – this does not include
the 1.6 million cable broadband connections [source DTI]. As TV becomes more readily available on the web and bittorrent enables large file downloads the merging of TV and web as mediums is bound to happen through audio-video.
So what about the search models for A/V? The first thing to say is that they’re quite distinct. Google would seem to be deploying a specific tool that will search existing A/V content on the internet using closed caption text – which is pretty ambitious. Aside from the rights issues of using the text there are also rights issues around playing the content if Google wishes to preview or offer stills of the search outcome.
Yahoo, however, would seem to be looking to manage the source files and index existing A/V content to then search, as Olsen states:
The Web portal plans to collect XML feeds of video content from
third-party publishers. That way, it can index programming and make it
searchable to visitors. Yahoo’s database will rely on the title and
description of video content to deliver relevant results, as opposed to
actual language within the video.
The copyright questions are more clear in this proposition: Internet
content companies looking to drive traffic to their video can likely
sell more TV-like advertising if it works.
Microsoft‘s plan would seem to involve annotating future A/V to search. This is resource heavy but microsoft may want to do deals to pay for this by providing advertising that is "contextually relevant" to the programming. Hmm. It does however, have the benefit of being platfrom agnostic and working with broadcast TV, cable, and internet.
It would seem like Yahoo will be the first to the market as their model is theoretically doable now. It will be interesting to see whether Google’s offering [and even Microsoft’s] can offer the extra richness of depth of archive and also more useful search results to usurp Yahoo. Interesting stuff… Nice one Stefanie.