Community features are only available to paying members.
+ Share this report
It’s a gorgeous day outside, but the Contextuall team is holed up in an East Village apartment, fine-tuning our prototype and continuing to add to our index of historical news headlines.
We’re still a few weeks away from an official launch, but for today we’d like take a break and jot down a few thoughts on our product vision. (Please feel free to reach out if you have any questions or feedback on this).
“Intelligent” Machines Reading the News
Imagine a world where machines read every single news headline generated by the media — news stories, Tweets, Facebook posts etc.
These machines turn each headline into a set of data points: What is being discussed? Who is saying it? When did they say it? Is the author optimistic or pessimistic about the issue?
Now, imagine if these same machines are constantly trying to “learn” from the interactions between media sentiment and real-world events, by using the latest artificial intelligence technologies.
For example, these machines could study historical news headlines on the economy, and “learn” how these sentiment changes affect the future performance of stock markets.
Alternatively, these machines could study historical sports pages to “learn” how the sentiment changes for the Los Angeles Lakers basketball team correlates to their on-court performance.
The possibilities here are endless…
This is Contextuall’s vision: extracting data from the news headlines that convey information about the world around us, and using machine learning technologies to find predictive relationships between media sentiment and real-world events…
The Media: A Rich Data Source
Our core assumption here is that the media is a reliable and useful data source. It remains to be seen if this is in fact true, and we intend to test this assumption thoroughly.
For now though, we’re assuming that journalists are reliable aggregators of information on the world around us. We’ll keep you posted on our progress.
Academic Studies We’re Loving
From the BBC: Researchers have developed software which could predict future events such as disease outbreak. The prototype software uses a combination of archive material from the New York Times and data from other websites, including Wikipedia. (Read More)
From WIRED Magazine: In Isaac Asimov’s classic science fiction saga Foundation, mathematics professor Hari Seldon predicts the future using what he calls psychohistory. Drawing on mathematical models that describe what happened in the past, he anticipates what will happen next, including the fall of the Galactic Empire. That may seem like fanciful stuff. But Peter Turchin is turning himself into a real-life Hari Seldon — and he’s not alone. (Read More)