The Beer Mapper helps people choose the right suds.
When exploring the murky continent known as beer, it helps to have a map. Will one find pleasure in the bright territory of the lagers and pale ales? Or do treasures lie in the darker regions inhabited by stouts and porters?
Only the individual drinker can decide for certain, but he or she no longer has to search alone. Rob Nowak ’90, MS’91, PhD’95 and Kevin Jamieson PhDx’14 are here to help. Inspired by the algorithms the two developed in their research, Jamieson created the Beer Mapper, a new app available for mobile devices that is designed to enable users to find the beers that they’ll like best.
Nowak is a professor of computer and electrical engineering, and Jamieson is one of his graduate students. The Beer Mapper is the product of research that the two have been doing on active learning. “This is an interactive way of learning how humans judge a set of objects,” Nowak says, “of finding how you would efficiently gather information in order to learn how to predict something — in this case, beer preferences.”
The Beer Mapper draws on thousands of reviews of different beers from the website ratebeer.com to classify varieties using a range of characteristics, such as whether the beer in question is malty, bitter, fruity, hoppy, light, or dark. It then plots each beer on a map — the one pictured here — based on the aggregate of those descriptions.
Users can look for a particular beer variety they like, and then see what other beers are similar. Or, they can go through a series of questions to discover where on the map their tastes lie. “We ask them either/or questions,” Jamieson says. “Do you prefer A or B? Do you prefer [Leinenkugel’s] Summer Shandy or Tommy’s Porter?”
The Beer Mapper has incorporated the feature dimensions of each beer, and so after a few questions are answered, it can plot a user’s favorites on the beer continent.
Jamieson is working with an app developer named Kevin Clark to prepare the Beer Mapper for commercial use.