Co.Create

Trapit Promises Better Browsing And Targeted Content On IPad

Uncorked Studios and Siri siblingTrapit collaborate on an iPad app that offers a new, machine-learning-driven content discovery and reading experience.

Until tablets hit the market, browsing on a mobile device, while possible, was hardly ideal. Small screens and a miserable user experience meant phone-based browsing failed to fulfill the promise of mobile news reading. Then the iPad changed all that.

Trapit For iPad is the latest in a wave of news-reading apps designed to make finding and reading online content on a tablet easier, more intuitive and elegant. Released by Trapit, which previously existed as a desktop application, and designed by Uncorked Studios, a boutique software design and development house in Portland founded by former Wieden + Kennedy Executive Interactive Producer Marcelino Alvarez, Trapit For iPad uses some very well-known technology to differentiate it from its competitors.

As a sister company to SIRI, the makers of Apple’s voice-command assistant, Trapit For iPad draws on a very powerful artificial intelligence engine in order to continually refine the content it serves up. So, when the owners of Trapit walked into Uncorked’s studio--a mere flight of stairs below them in the same building--and said, ‘hey, wanna develop an app for us?’, naturally Alvarez said yes. “It was our first walk-in for an iPad app,” says Alvarez. Here’s the skinny on how it works and why it matters to you.


What is it? Trapit for iPad is described by the company as a “personal assistant for the web” but can be more clearly defined as a content browsing app that uses machine learning to adapt to your reading preferences.

How does it work? When users download the app, they’re first encouraged to create a new “trap,” or topic they’re interested in following. Traps can be as broad as media or technology or as niche as you like. Not interested in defining your interests right off the bat? No problem. Featured Traps allow you to browse curated topics, which at the moment range from women’s soccer, to hurricanes, chocolate, Frank Ocean and Israeli-Palestinian Negotiations. Once within a trap, users are able to browse countless stories served from 100,000 sources and are encouraged to rate the content as a way for Trapit for iPad to learn how accurate its results are.

Why is it innovative? Because Trapit for iPad finds articles on very specific topics that you’re interested in rather than just serving up content from your own social feeds, like competitor Flipboard. As the folks at Uncorked Studios say, “born from DARPA-funded research at SRI and a sibling to Apple’s SIRI assistant tech, Trapit’s machine learning bypasses the web’s SEO-driven priority by harnessing peoples’ behavior to hone discovery of articles that match the user’s passions.”

“Trapit combs through its hand-picked sources using relevant keywords,” explains Alvarez. “At first it provides you with a huge list of content it thinks might be relevant to you. The case of ‘apple’ is a good example. There are multiple uses of apple out there and more often than not it’s going to be the tech focused Apple but it might also give you results for the food. So when it delivers the first batch of content you can specifically train it find just find articles about Apple the tech company. And then even more so, if you’re interested in apple iPad apps, you can get very granular with what you’re searching for.”

What to watch out for: At the moment, Trapit for iPad is not perfect. Version one has struggled with stability, particularly on first-gen iPads, but Alvarez says version 1.2, which is currently under review with Apple, “makes significant improvements to stability.” First-time users should also be aware that instructional overlays, directing you to use its thumbs-up rating system, get in the way of immediate access to the content. Just keep tapping and this informative intrusion will be gone.

Why you should care: Though Trapit for iPad’s user experience could be improved by shortening the number of clicks from a trap to the content (something Alvarez says Uncorked is working on and originated out of the desire to showcase content in its native format), there’s huge potential in its machine-learning engine to deliver articles on super-niche topics from unexpected corners of the web.

Add New Comment

0 Comments