Broadcasting to machines

With the launch of channel4.com’s new Programmes area, our audience has a consistent way to access all the information we’ve got about a broadcast episode. Clips and pictures from the show, features written about the people in the programme, and transmission information are all browseable. Some programmes have more information available to the website’s visitors than others, but a standard set of data is emerging. And, because we’re taking steps to make our URLs consistent on the site, it also makes it easier for machines to find out the information about our programmes that external applications may want to use.

Screen-scraping and the like is one way for a ‘bot‘ to gather information from a website, but today, the preferred means is that of an API (application programming interface) that allows one computer programme to call on data that another computer programme provides.

Web frontrunners such as Last.fm and Amazon have offered public access to their data for years, but publishers and particularly broadcasters have been slow to respond. Rights issues, fear of losing audience share, and confusion about what benefits to the user an API might bring are all to blame for broadcasters’ hesitation at publishing open data APIs.

Several blue chip media companies have broken ground recently, offering software developers access to partial or full sets of relevant data. Take the New York Times’ Open network, which offers access to its whole history of news articles and (over 75 years’ worth of) movie review data, amongst other APIs. Developers keen to see what they can make with the Times’ data have already started experimenting with some applications. Meanwhile, music television pioneers MTV recently opened up access to its archive of videos. We’re waiting to see what new spin on the classic clip developers come up with in the coming months.

In the spirit of public service broadcasting and opening up data for all, Channel4 is aiming to release an area devoted to developers later this year where they can play around with some of the data at the core of the channel. It’s an ambition that the BBC have taken the lead on with the estimable Backstage project.

But before we embark on building it, we’d like to know what you’d like to see from a channel4 developer area. What data would you like to play with? What applications can you imagine building if you knew:

• all the music tracks that were played on The IT Crowd?
• how many people were leaving messages on a contestant’s page on the Big Brother site?
• which episodes of Hollyoaks such and such appeared in?
• which recipes have Gordon and Jamie cooked on TV?
• what films have the highest honours on Film4?

and so on?

Answers on a digital postcard below please. The suggestions that you make contribute towards a better API.

Permalink||(1) Comments