In my last post I talked enthusiastically about the Beer Maverick site and their detailed information about hops (> 300 types). That information got me thinking about creating a hop network where each hop would be positioned based on their relationship to each of the nine aroma types identified on the Beer Maverick site. Here’s a quick reminder for what those nine types are, as seen in their Strata profile:
The idea behind creating a network map is to view all 300+ hop varieties in one visualization, positioned based on their common characteristics. We can see above that Strata scores high on the Stone Fruit attribute (peaches, apricots, etc.) and much lower on Berry and Grassy. In a network graph, other hops with a similar profile will be positioned close to Strata, while those with radically different mappings will be in a different sector of the graph.
Here’s a quick example developed with Gephi network graph software, using just three hop varieties:
We have colored the hops in green and the aroma attributes in a shade of yellow to make it easy to distinguish between the two. As predicted by the earlier radar graph, Strata is positioned close to both the Stone Fruit and Tropical Fruit nodes, and is connected to all aromas to at least a small degree. Mosaic, on the other hand, has no connection to the nodes at the top of the graph, such as Herbal and Spice. Chinook connects to Herbal, but not to either the Stone Fruit or Tropical Fruit attributes. The network is designed to minimize crossings, so nodes are positioned closest to their relevant connections.
This is a really simple illustration; just wait until we get all 300+ nodes in the graph! Things will become more cluttered, but we do have ways of optimizing the graph through sensible design of the node and edge elements and by allowing users to hide unrelated elements when they are traversing the graph. I’m looking forward to that graph, and will offer some check-ins along the way as more data is gathered. As always, thanks for reading!