Tap on the Layers icon on the toolbar (in menu button in iPhone) to switch between map layers. Each layer shows how your favorite styles and beers vary across these attributes. The “My Preferences” layer is a personal heat map used to predict where your preferences lie based on the ratings you have given to beers you have tried.
Each of the 25,000 beers is represented as a unique pin on the map. A beer’s coordinates are determined based on how it tastes in relation to every other beer on the map. The closer the beer pins, the more similar the beers. Tap and hold on any one of the dots on the map to pull up a beer pin. Tap on the beer’s flag to find out more info on the beer, find similar beers, or add it to a custom beer list.
About BeerMapper - An Overview
Beer enthusiasts around the world have helped to make beer review websites a phenomenal success. Their reviews contain a wealth of information, but searching through reviews to discover new beers that you’ll like can be tedious and difficult for experts and novices alike. We knew there had to be a better way, and BeerMapper was born!
BeerMapper harnesses the crowdsourcing power of beer reviews to provide personal recommendations and intuitive visual maps that enable beer lovers to easily discover new beers, new styles, and better understand the nuances of their personal beer tastes. We have teamed up with RateBeer, home of one of the most complete beer databases on the internet, and applied state-of-the-art data mining techniques to identify patterns and relationships hiding in the data that even the most seasoned beer geek might not be aware of. These patterns were used to create the most detailed yet intuitive visual maps of the beer universe known to humankind. Beer exploration will never be the same, #DiscoverBetterBeer today!
Map Layers Visualize the Beer Space
How do you make a complex set of data (such as the 6 million reviews on 25,000 beers in RateBeer.com’s database) more intuitive? By making it visual! BeerMapper shows the entire beer universe on one screen using space and color to communicate the relationship between beers and beer styles.
We built an algorithm to look at every review of every beer in RateBeer's database and used cutting edge machine learning methods to determine the similarities and differences between all beers, based on the expert and nuanced judgments of beer lovers from around the world. Next, we applied advanced mathematical tools to build virtual beer maps that capture the subtle, and not so subtle, relationships between beers. Once the spacial relationship between all the beers is determined, the complex space of beers is charted into two-dimensional layers and key maps are generated based on attributes including color, maltiness, bitterness, and alcohol content.
Smart Lists = Power Self Exploration
Beer Lists enhance the exploration process by giving you simple tools to create complex lists. Navigating the beer space using “smart filters” allows you to slice and dice custom lists according to beer, brewery, style, geographical region, and more. You can even create personal lists of your favorites or lists based on any other theme you desire. We will even be rotating in some of our favorite lists to give you some ideas. However you build your lists, discovering new beers that you love will be a breeze
Filling in the Holes
BeerMapper is also helping us expand the scientific understanding of today’s complex world of beers. While RateBeer.com's database is very extensive, as with any real-world dataset, it isn't perfectly complete. For instance, some breweries do not report the IBUs (international bittering units) for their beers; which means we don't exactly know how hoppy their beers are.
Thankfully, many breweries do report the IBUs and BeerMapper uses this information to estimate the missing IBUs using the review-derived maps. The same technology can be applied to more perceptual properties, like the matliness and color, allowing us to fill in the holes of the database, with error-bars that quantify our confidence in the predictions.
A lot of apps claim to make recommendations, but we actually proved mathematically that no recommendation system, human or machine, can learn your personal tastes faster than BeerMapper! Our engine uses three powerful yet sensitive industry leading algorithms that optimize our recommendations based on how much information we have on your preferences. Simply by rating the beers you have tasted, we can predict how you will like other beers by picking up on cues and patterns in your ratings.
The more beers you rate, the more customized the recommendations become, and the learning evolves as your tastes do. This means that BeerMapper’s system can rapidly adapt its recommendations to match your preferences at the moment, be it a warm summer afternoon or a cold winter night. With BeerMapper, you will never have to order or buy the wrong beer again!