Looking into the future used to be a case of going to Delphi and talking to a gas-intoxicated woman sitting on a three-legged stool. And I’m not sure the science of it has really improved since then.
When I was growing up – the ‘80s and ‘90s – there was a lot of talk, from people who liked to talk about such things, about society, in particular its youth, becoming fundamentally illiterate. TV had rotted the mind of a generation, with some help from rock and pop music. As well as being pompous throwback windbags, these people were telling the future badly.
We have, to my mind, never lived in a more literate society. During that period the advent of the internet, personal computers, email, mobile phones and texting now means more people use the written word than ever before, frequently in inventive and unorthodox ways (more fodder for the pompous windbags – those guys never go away). Our main interaction with the web is via text, even if what we ultimately consume is in a variety of media formats.
But there was a problem with this renaissance of the written word; the channels of information – newspapers, terrestrial TV, radio – that had previously acted as gatekeepers to the flow and delivery of information and content, were suddenly swamped by a Babel of opinion and words. There was a load of interesting stuff available, and it was all over the place.
The barrier to entry for publishing content had dropped from the pole vault to how-low-can-you-go limbo. This was tremendously empowering – voices that had not previously been heard, niche interests that had not previously been catered to, the disempowered, the marginalised and yes, the eccentric, were all able to find a space.
It was also disorientating. Here Comes Everybody. I remember it well – ‘Where do I go to find things that I’m going to like?’ ‘Where are the people whose opinion I trust?’ And ‘Just where do they hide the interesting stuff these days?’ ‘Oh wow, look: cats.’
It’s a challenge that, for many, still requires a solution. In an on-demand environment, with an ever increasing number of competing sources of media and platforms, just how do you ensure that viewers find the content they want, when they want it? In order to solve that problem, content owners need to know more about their content than ever before. What’s in the archives? What is relevant to today’s news stories and how do we ensure the relevant clips are discoverable?
If this all seems a little remote to the process of captioning, it shouldn’t. As well as providing a core service to the deaf and hard-of-hearing, our captioning process, whether it’s for live or pre-recorded content, produces a complete textual transcript of all a broadcaster’s spoken content. That’s incredibly useful data. And our captioning system can store time, programme, channel and news story information alongside each word in its database. That’s incredibly useful metadata.
Data tells you what you’ve got. Metadata helps you find what you’re interested in.
Regulatory requirements, as well as the need to provide the best possible service to our viewers, mean the quality of this data is already mandated to the highest levels. By running language algorithms on that database content, we can generate additional metadata – keyword summaries, search engine identifiers, trends and cross-content associations – all of which allows broadcasters to open up their content in new ways to audiences, no matter where they’re viewing that content. As we all already know, text is the best way to discover all kinds of media content. It’s much easier to search textual databases than audio-visual ones.
It’s this that has got broadcasters excited. Recent responses in the US, a market that has traditionally been quite resistant to change, have been overwhelmingly positive. Captioning has gone from being a service they are required to provide, to being a resource that’s useful to them in all sorts of ways.
And you know, those Greeks weren’t so crazy after all. Because the thing about the Oracle at Delphi was that it worked. It must’ve done, right? It was the most revered place in the Mediterranean for centuries. That wouldn’t have happened if the woman with the wow had been dropping bum steers for a living. Best guess historians have at the moment? Delphi was powerful because everyone went there when they were deciding what to do. The priests who interpreted the prophecies were privy to pretty much every important decision made in the region for 400 years. There wasn’t a thing they didn’t know. What Delphi represented was a huge database of useful information that you’d go and query when you wanted to know what the best course of action was. If they’d had some additional metadata, who knows what else those Greeks could have achieved.
Tom Wootton, Senior Product Manager, Access Services