I'm just leaving Sao Paulo after attending the ACM Symposium on Document Engineering. Great conference (people and technical content). I'd like to say more about a number of the other papers, but I'll focus here on the one that won this year's Best Paper Award: Enabling Adaptive Time-based Web Applications with SMIL State by Jack Jansen and Dick Bulterman.
The paper presents a very interesting mix of SMIL, XForms and other technologies to enable end-users to customize their experience of content where time dictates the major structure. One example given is of a video bike tour of Amsterdam created by the first author. Fragments of the video content can be identified and tagged with keywords. During run-time, an XForm is used to allow an end-user to select from the available keywords those which they find interesting, and the corresponding fragments of video are played.
This is the type of thinking that will clearly help make video content a first class citizen on the web. There is significant potential for the use of this technology for stream-lining educative experiences as well. Imagine, for example, a 2-hour audiovisual presentation that provides a thorough introduction to a topic, say CSS. But suppose the end-user really needs to know only about setting up borders on tables. By selecting the content, the user can find out what they need to know in 10 minutes. What's more, it is easy to imagine how the technology reported by Jansen and Bulterman could be refined to attenuate the video content download to the selected fragments.
A second example reported in  pertains to ad selection within video content. When you download free video for your favorite entertainment, you're going to get ads and you won't have the ability to skip them because otherwise how do people manage to make a living providing you with this free entertainment?? So, sprinkled through the video will be ad slots with default advertisements lined up. But, in addition there will be hotspots on the video where various product icons may appears from time to time, and if the user clicks the icon, then the default ad is replaced with an ad corresponding to the clicked icon. Thus, you could find out what's new at the electronics shop in lieu of an ad for laundry soap or some such.
It's a great example on the technical side of the need for interactive video, but this is because there's huge market potential for this idea. Not only does the user get a more pleasant ad experience with their free content, but the content provider gets to find out what their audience is most interested in, and the advertisers get to find out what content their customers are most interested in. Very Web 2.0.
I encourage you to download and read the paper.