The Future of Social Media: Predicting trending Twitter topics with 95% Accuracy

taracb    November 18, 2012

Devavrat Shah, a MIT professor in the electrical engineering and computer sciences department, and Stanislav Nikolov, a graduate MIT student have created an algorithm for Twitter that is able to predict trending topics before Twitter is able to with 95% accuracy and up to 5 hours earlier.

The algorithm sifts through a large sample of tweets, popular trending ones and unpopular tweets.  The data is then compared to new information to see if there are any patterns.  The new tweets that look like older tweets, that have also trended means there is a chance a trend will be formed.

Lessons for Others:

The new algorithm created at MIT can be applied to any social media platform including, Pinterest, Facebook, LinkedIn and even with Google searches.  The ability to predict trending topics, before they have even become trending on social media is a significant resources.  Marketing professionals would be able to know what will be trending, which can help with future and current campaigns.

Web References:

Researchers Predict Twitter Trends with 95% Accuracy 

MIT Algorithm can predict which Twitter topics will trend

Submitted by : Tara Busse, SMBP Student University of Waterloo
To contact the author of this entry please email

If you have concerns as to the accuracy of anything posted on this site please send your concerns to Peter Carr, Program Director, Social Media for Business Performance.