Hacker News new | ask | show | jobs
by mistrial9 1825 days ago
A stealth tech company in Calif-US created a system to record as many news casts as possible, 24-7, and invoke the closed-caption system to annotate it, and make it searchable. The company was sold in the early 2000's for $1B+ ; I have never seen the name in any tech news.

source: colleague was a core engineer there

4 comments

SnapStream by the makers of BeyondTV started doing this around that time frame. As another commenter mentioned, it is WIDELY used in media for searching/archiving/clipping programs.
I've no idea if it matches your story, but I hear a lot of shows use SnapStream to find content? https://www.thewrap.com/the-secret-to-daily-show-colbert-rep...

And I remember back when Google Video was a thing that Google was indexing feeds of TV channels' subtitles. I remember there was a search engine at one point for subtitles? Presumably they still index live TV subtitles and content both to support YouTube TV but also to feed content into their YouTube Content ID program... I wonder if anyone's comparing TV subtitles to speech-to-text output to catch differences in either.

Here's an article about Google's early Video Search product for captions: http://googlesystem.blogspot.com/2006/08/brief-history-of-go... Apparently displaying and searching captions to the public for free lasted barely 6 months... Here's a screenshot showing a search result: https://informitv.com/2005/01/25/google-video-search-for-tel...

I would've said this was no doubt bought by a financial sector company looking for bottom up sentiment analysis, but the whole premise of the parent post is that there's hardly such a thing as bottom up sentiment in local news.
As far as I can tell about how the unlimited cloud DVR on YouTube TV works, google is already recording all local news broadcasts all the time and retaining them for at least 90 days.

I don’t think they have any plans to do anything other than what they do now (selectively offer playback to subscribers who opted in before the broadcast), but makes you think it’s a solved technical problem that governments and other players are also probably doing.