It’s as if the world’s celebrities, politicians, and companies were your Facebook friends. News aggregation website Wavii, launched today, distills current affairs into a feed of the kind of pithy, easily digested updates seen on the social network.
Users of Wavii (pronounced “wavy”) who logged in yesterday would have been informed of the biggest technology news of the day with a short update similar to those that Facebook produces when two people enter a relationship: “Acquisition: Facebook acquires Instagram for $1 billion.” Clicking on the update yields more information and a link to online articles about the event; clicking on a company name calls up a profile page showing all recent updates about it.
Wavii creates its newsfeed by digesting information from online news sources and turning them into short summaries. It adds photos, charts, and maps as appropriate. Users choose a set of interests, companies, or people that they want to see updates about in their newsfeed.
Wavii’s founder and CEO, Adrian Aoun, says Facebook’s approach to summarizing social news was a direct inspiration. “It’s so efficient that in three minutes I can stay up-to-date with a thousand friends. I get everything in context, with maps and photos, and I can click on people’s names,” he says. “But I only get these updates on my friends, not the world.”
Giving the news a Facebook-style makeover is not straightforward, though, because unlike Facebook, Wavii cannot rely on its users to neatly classify world events into defined categories. Nor are news articles easily digested by software. “We have to do it by crawling the real-time Web and teaching the computer how to read,” says Aoun.
Aoun and his colleagues used machine-learning techniques to teach their software how to categorize the events described in news articles: for example, that two people got married, or a company released a new product. The news can then be succinctly summarized in the most appropriate way, including the addition of relevant photos or quotes.
Wavii’s system was trained using data sets that provided basic knowledge needed to identify public figures and companies, and how events are described. Then it was set free on live data from the Web and tasked with identifying new events worth reporting. The things it came up with were reviewed by humans. That process, over time, taught Wavii’s software to accurately recognize over a thousand different types of event, says Aoun, so it can constantly feed on online news sources and summarize the most important news of the moment.
The software can even deal with ambiguities such as articles referring to people in different ways (Jennifer Lopez and J-Lo), detect sarcasm (to avoid reporting false information), and tag celebrity updates as “rumor” when the sources aren’t clear.
Even if Wavii can’t fit an event reported by several sources into one of its templates, it will try to summarize the news in a single sentence. For example, a common follow-up story to Facebook’s acquisition of Instagram wasn’t classified into a category, but was still properly summarized: “Kevin Systrom will make $400 million as a result of Facebook’s acquisition of Instagram.”
Being able to understand the content of news events allows Wavii to be better customized for specific users, says Aoun. He says Wavii can make money by matching ads to what users are interested in.
“I think the idea of applying an activity stream view to the news is an interesting test,” says Nova Spivak, founder and CEO of Bottlenose, a Web service that can intelligently filter streams of social or other data to present users with updates on particular topics. Spivak has met with Aoun and tried early versions of Wavii. “They seem to be creating a new news-consumption experience, and we’ll see if it catches on,” he says.
Web users certainly need tools able to help them navigate the growing volume of online information, says Spivak. “There’s an overwhelming onslaught of data.”
Wavii is not the only startup trying to make online news reading more efficient. Flipboard and Zite are just two of the other startups in this area. But Wavii’s attempts to digest and understand the content it aggregates make it different. Spivak believes, however, that taming all of that data will require a platform capable of intelligently digesting and filtering all types of data, for example social updates as well as news, something he is trying to make possible with Bottlenose. Aoun says that although Wavii is focused on news for now, it will ultimately be able to tell you about events from other sources, such as local listings.
What to know about this autumn’s covid vaccines
New variants will pose a challenge, but early signs suggest the shots will still boost antibody responses.
DeepMind’s cofounder: Generative AI is just a phase. What’s next is interactive AI.
“This is a profound moment in the history of technology,” says Mustafa Suleyman.
Human-plus-AI solutions mitigate security threats
With the right human oversight, emerging technologies like artificial intelligence can help keep business and customer data secure
Next slide, please: A brief history of the corporate presentation
From million-dollar slide shows to Steve Jobs’s introduction of the iPhone, a bit of show business never hurt plain old business.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.