Wordnik claims to be the world's largest online dictionary, with billions of words, billions of example sentences, and hundreds of thousands of pronunciations, related words, and comments.
Today the company announced that it is using this powerful dataset to do something new -- create a highly intelligent news reader called Reverb, which makes personalized content recommendations based on its understanding of words and concepts.
"We essentially built a word graph, and if you know the coordinates for a lot of words and how they fit together, you can do a good job of figuring out what people want to learn next," founder Erin McKean told VentureBeat. "Because it is based on the words that people use to convey their ideas, it is less game-able than the other ways people discover content."
Reverb's technology understands the underlying concepts in articles and organizes content from across the Web into separate streams based on top news, social streams, personal interest, and location-based information.
The top news section does not get personalized, because the focus is on the most important news of the day. Reverb will also see what your friends on Facebook and Twitter are reading and prioritize it for you.
Readers click on topics and stories that appeal to them on the app's "word wall." Reverb collects this information so it can gain a greater understanding of your interests over time and lead you from one article to another.
The Internet is home to a seemingly endless array of content and plenty of startups that want to help you navigate it all. Digg, Prismatic, Reddit, Flipboard, and others take various approaches to content discovery, but McKean said Reverb's strong technology foundation enables it to offer more intelligent recommendations and leads to "limitless" exploration.
"There is so much fantastic stuff out there you could read, and at the end of the day you have half a dozen tabs open and still didn't catch up on topics that interest you," she said. "We want to make it easier to discover things you didn't even know you wanted to know."
The goal is to help you stay on top of current events, what your friends are talking about, what's happening near you, and your own unique set of interests, with minimal effort on your part.
McKean formerly worked as the editor-in-chief for the American Dictionaries at Oxford University Press. She gave a TED Talk in 2007 in which she presented her idea for a better online dictionary. The audience expressed excitement about the prospect, and over the past six years, sheand her team created the massive dictionary, as well as a community of 100,000 word "enthusiasts."
She said Wordnik's API served over a billion calls last week.
Wordnik saw an opportunity to open up this technology to a wider audience by applying it in a new way, to content discovery.
Reverb's app is now available on the App Store, and it offers tools for publishers and developers. It is based in San Mateo, Calif., and backed by $6.3 million from Baseline Ventures, Floodgate, Manatt Venture Fund, Lucas Venture Group, and others.