Monday, April 4, 2011

Man vs. Machine: Curating with YouTube APIs

If you are reading this, chances are you are building an app which includes video. Given that over 35 hours of video is uploaded to YouTube every minute, finding the best of it can be challenging. In this post we’ll discuss some interesting approaches and illustrate them with applications recently submitted to our YouTube API Project Gallery.

Use the feed

To get started quickly, API developers will find that using standard feeds is a great way to find content. In case you have not noticed, we have added two new experimental feeds: “most shared” and “trending videos”. They provide more of a real-time pulse of YouTube and expose some of the functionality behind YouTube Trends so that it can be accessed programmatically.

Storytelling apps

While standard feeds provide a convenient mechanism, content curation around common narratives calls for more complex apps. Let’s take a look at a few examples. Storify [1] and Memolane are two interesting applications addressing different aspects of creating a social media narrative around a story (the former) or chronologically-organized series of life experiences (the latter). History of Jazz is a good example of a mobile app using YouTube built around the concept of curation. The application uses iOS video playback techniques described in our earlier blog post on the topic and is a great learning tool for those of us who like to discover new music. Shortform is a curation platform for video DJs (VJs), where users can organize interesting content for their viewers to enjoy across a wide range of topics. And yes, you too can be a VJ!

Searching for a deeper meaning with semantic analysis

The apps presented in the previous section focus on empowering users to organize content. Another approach worth highlighting is a hybrid of curator’s selections and related content recommended by an algorithm. To achieve this, ViewChange.org by LinkTV, a non-profit committed to improving the world with storytelling, utilizes semantic analysis. As shown in the screen shot below, top-level categories are decided upon (or “bootstrapped”) by curators, however, while navigating the site users are presented with automated recommendations as well.

Wow, how did they do that?

To learn more about these applications, check out the project gallery. Additionally, some of these developers made more information available about their API experience so we’d like to take this opportunity to highlight it as well. Memolane put together a blog post about their usage of YouTube API, which you can read here . Shortform also blogged about their use case; follow this link to learn more. If you would like to understand how ViewChange integrated semantic processing in their video site, you can find their blog post about the topic here.
As always, if you have an interesting YouTube API project you would like to share with the developer community, please submit it to the gallery. We would love to hear from you.

Cheers,
—Jarek Wilkiewicz, YouTube API Team

[1] At the time of this writing Storify was still in Beta, but we have arranged for a special invite code for our readers. Use youtubeapirocks and you’ll be able to create your own stories.