Nate Weiner

Intwition - Relaunched

• News Detection. Finds major topics as they are happening.
• See What People Are Discussing shows only the stories people are actively talking about on the web.
• Find New Stories Use the New Scoop Finder to see what’s being discussed at any given moment.
• Find Related Content Finds related blog posts, news articles, pictures, and video for each story.

Overview

Intwition analyzes internet chatter to detect breaking news stories and the topics that people are truly interested in.

Intwition’s Aim

After the ground stopped shaking in China, people started talking. They started calling their friends, they started texting their family, they started posting messages online. After 7 minutes, the first message appeared on Twitter, after 8 minutes local Beijing news agencies started reporting an earthquake. And after 10 minutes, the story was on Intwition’s demo homepage. Intwition is an algorithm that monitors internet chatter to detect waves of discussion and deliver the content related to it. News is being reported by deeply embedded reporters, all over the world, and at every given moment. These reporters are called “people”. And they are the reporters on the scene - anytime and everywhere. By listening to them, we can find stories as they happen and show what topics people truly care about. This is what the new version of Intwition aims to do.

On an Ordinary Day

Even on an ordinary day, when major events aren’t happening, Intwition is still useful. The stories that appear on Intwition’s homepage are the stories actively being discussed on the internet right at that moment which means these stories are indicative of what topics people really care about from the news that day. Rather than relying on user’s to submit stories and then manually vote up those stories, Intwition simply looks at what people are talking about. In an essence, to vote for a story, all you have to do is talk about it on the Internet. This method proves to be faster to deliver content than some previous social news models.

What Happened to the Old Intwition?

Let’s face it, web applications that track links from Twitter are simply a dime a dozen. There are plenty of sites out there that simply display links based on the number of times they appeared in Twitter and the first version of Intwition was one of them. I wanted to do more. I wanted to group related links together, I wanted to bring breaking stories up to the top faster, and I wanted to expand the source of data beyond just Twitter.

Intwition’s New Path

As it stands, Intwition is an experimental proof of concept. The first iteration of the program, the one that discovered the earthquake in China, had only been a few hours old and was being run for the first time overnight. Since then, it’s been changed dozens of times and continues to be tweaked each day to see how it handles the next wave of news. Without question, it has its flaws, but for a 10 day old program, it shows great potential. Initially, Twitter is being used as the primary source of chatter and as a result there is an obvious tech bias to some of the stories. I have plans to begin pulling from other streams of information as the system evolves. With this, I can display a greater number of stories that are filtered into categories. Moving forward into the future, I’m focusing on preventing duplicate and unrelated stories from appearing before increasing the speed of reporting. There is a delicate balance between speed, preventing false alarms, spam, and unimportant stories (for example, the fact that everyone is suddenly eating breakfast on a Monday morning). As I said before, the algorithm is experimental and I will be the first to admit it’s far from perfect but I wanted to release it into the wild and start pushing it harder to evolve. In the past weeks/months we’ve been witnessing how Twitter is slowly becoming a unique source of news and I wholeheartedly believe this trend will change the way we found out about news. Intwition is simply trying to lead the way.