Can We Predict What Will Trend on Twitter?
Twitter is one of the best ways of keeping abreast of what's happening in the world. News tends to break on the site faster than it does anywhere else, and as such is a goldmine for those looking to catch trends before they break.
So news has arrived that a researcher from MIT may have developed an algorithm that can predict what topics will trend on Twitter is very interesting indeed.
The application, developed by Associate Professor Devavrat Shah, can apparently predict trends hours before they appear on the official Twitter trending list.
The application uses a machine learning algorithm that compares real-time data with historical data. It then predicts outcomes based on past events that most closely matched current situations.
It's a service that Twitter is certainly interested in, as it appears that Shah's research assistant is none other than Twitter employee Stanislav Nikolov.
This interest is likely to revolve around selling ads at a premium to companies based on whether their topic is going to trend or not. Of course to do that the application would need to be tested on the full Twitter data set, but it is a certain possibility.
Advertisers might not be happy to pay premium rates for topics that fizzle out before ever becoming top trends however. Thus far, the algorithm has been trained using a set of 400 topics, half of which trended and half of which did not.
Shah has also suggested other uses for the application, including stock markets, movie ticket sales and public transportation as possibilities, but others might include combating cybercrime by identifying threats earlier or predicting the severity of disease outbreaks.