Automatic subtitling comes to YouTube – the implications for search
First and foremost, the news that YouTube is rolling out automatic subtitling is good news for the hard of hearing:
The machine-generated captions will initially be generated in English. At first they will only be found on 13 channels.
These include National Geographic, Columbia, as well as most Google and YouTube channels…
Currently YouTube offers a manual captioning service but video makers tend not to use it. [BBC]
There are two other benefits it may bring. More widespread captioning of YouTube films means more of them will be practical to watch in situations where you can’t have the sound turned on (such as when using a work computer without headphones).
But also, by generating text from words, as the BBC is doing for coverage of various Parliaments and Assemblies, YouTube will be generating the raw material to make its films much more findable via search.
YouTube itself is already the second most popular search engine (after Google) but whether searching it directly or via another search engine, you only search the text that is available. Imagine if searching for blog posts meant you only got to search the text of the headline, excerpt and tags. Useful – but a whole lot less useful than being able to search the full text.
Auto-captioning opens up the possibility of providing search engines with a much richer set of information, making videos come up in search results more often and therefore be viewed more often.
Still have fond memories of the machine-generated captions used for the marriage of the blessed Diana and Charles Windsor way back in 1991. The had gems such as the newly-minted princess ‘foaming out of the class coach’ and a wonderful character identified as the ‘Duck of Ed In Borrow’.
The auto-system used for the news now in 2009 still needs a human over-ride safeguard to prevent such wonderful nonsense. I suspect we are in for some entertainment on ‘You Too’.