Mayor Nenshi ballparks Closed Captioning data as 85% accurate, and requiring a “big disclaimer”. I’ve added such a disclaimer in the start of YYCCC 2010-12-13’s captions.
The following represents a “Web 2.0 friendly” approach to archiving Council sessions. And as of this archive, finally includes a reasonably accurate transcript.
This embed isn’t just pertinent to my blog post, it illustrates how YouTube embedded on a web page can start at a precise moment in the video. Simply add the following parameter to the HTML: &start=18320 (That’s the # of seconds to start at.)
(The start parameter currently has no impact on iPhones, hopefully that is on Steve Jobs’ to-do list.)
DJ Kelly had been a strong proponent of Open Data at city hall. Some datasets have obvious value (realtime location of busses, budget details). I’d like to add City Council audio, video & transcripts to that list. Of particular interest to me is Closed Captioning data.
The workflow I’m now using to upload video archives to YouTube does not parse Closed Captioning in realtime… the video is captured as a single large (1GB/hour) MPEG-2 file, which can be processed once council session ends.
Currently that is due to a file-locking issue with CapDVHS. But CCExtractor was created to parse files in realtime… in theory it would be possible to transmit Closed Captioning data to an RSS feed or Twitter as it is received.
Open Data requires both a friendly license (so that people can re-purpose the data without worry about being sued), and data in an easy to parse format.
Pulling Closed Captioning off SHAW Digital 89 is anything but easy (in realtime anyway). But it is possible. If the City of Calgary had to justify disallowing re-use of each dataset, instead of Calgarians having to argue why a given dataset should be made public and re-useable, I believe we’d see a lot of interesting applications appear.
When City of Calgary rolls out ELMS video, it will offer appropriate video playback for the minute/item being reviewed by the user. Here I illustrate that YouTube could also be integrated in a similar fashion to the City of Calgary’s website.
While I would be happy to see the addition of any video archives to Calgary.ca, there is a reason the majority of embedded internet video is hosted on YouTube. They’re backed by Google, therefore YouTube technology receives more R&D than any other vendor can afford. YouTube has a public API. YouTube facilitates access for the hearing impaired. YouTube has a bit of the Web 2.0 going on.
In short, if it says YouTube, YouTube, YouTube on the label, label, label, you will like it, like it, like it on your table, table, table.
Offering up video archives (with a Microsoft video server) represents meeting what needs to be done city employees, but does not facilitate discovery of what could be done by Calgarians.
My sample Calgary.ca integration is a hack, and is only intended to show the end-result of integration (it does not represent a sane workflow). However, up to the point of ELMS integration, this process is quite efficient. If ELMS (as Gregory of City Clerk’s office indicates) records a time-stamp for every motion put forward, then it should be easy to calculate (in .NET or JavaScript) the needed start parameter to play the appropriate portion of a YouTube video.