The rapid proliferation of digital media on the Internet has created unprecedented choice in audio and video content for consumers. But it has also brought the headache of making it difficult for consumers to find the sort of content that they like.
Helping consumers find the sort of content that they want is the goal of Scouta, an Australian-developed application currently in the final stages of development.
According to Scouta's co-founder, former systems engineer and Web developer Richard Giles, the idea came in 2005 after he was approached by the technical publisher O'Reilly to co-author a book that would review 100 podcasts. Giles says at the time he was able to find around 30 quality shows, but after that the pickings grew slim.
His initial idea was to create a Web site that rated and recommended different podcasts. But then after a few drinks at Fremantle's Little Creatures brewery with friend (and programmer) Graeme Sutherland, they hit on a more practical idea.
"It occurred to us that this was sort of a lame idea when you could use the wisdom of crowds and provide recommendations based on what everybody else is enjoying," Giles said.
Sutherland then set about building a proof-of-concept application that would be able to tie into a user's content database (in this case Apple's iTunes) to scan the content they had accessed and match that through a central database against content from other users. This matching process could then be used to send back recommendations that might match the users' tastes.
Around the same time, YouTube started to boom, making it clear that Scouta should also be recommending online video in addition to just podcasts. So, when the Scouta Web site was built at the end of 2006 it included the ability to embed MP3 and MP4 files. Around a dozen file types are now supported.
There are two elements to the Scouta system. The backend system has been created in Python and uses the TurboGears Web development framework. The database is currently housed in PostgreSQL, after a conflict between TurboGears and the original choice, MySQL, forced a change of plan.
The next stage of development will be a release of the client software, also being written in Python (although some of the code for the Windows version is being created using C#). Giles said he hopes to have the Mac client ready for release within the next month, with the Windows version to follow soon after.
The initial version of the client software will be capable of interrogating the user's iTunes database, meaning that it can be integrated with the iPod, AppleTV and the iPhone immediately. But Giles said Scouta is creating an open API in the hope that other developers will build linkages for interrogating other media players and devices.
"The grand plan is to have this work with any popular media player, whether it's hardware or software," Giles said.
Much of the functionality of Scouta is already demonstrated through the company Web site, such as the use of tags, ratings and groups to add a social networking element to the service. The company has been fortunate enough to benefit from the experience of Cluetrain Manifesto co-author David Weinberger, who has advised on many aspects of the service.
Giles said that all through the development phase both he and Sutherland had been conscious of the need for the server application and the database to scale quickly should Scouta prove to be a hit with consumers.
He said there is also scope for Scouta to become a recommendation engine to content providers.
"A TV station or even a telephone network might want to provide their customers with highly relevant recommendations, whether it is based on their own content or everyone's content," Giles said. "We expect it to be massive, and we are definitely writing the code and the architecture to ensure that scalability."