Data-driven Flavor Profile
Explore automated tasting notes from reviews with the Data-Driven Flavor Profile.
The Tasting Intelligence Data-driven Falvor Profile
A flavor profile is a way to describe what you taste. You can use this to describe what you taste and compare the tastes of drinks. Use it to identify the taste and describe the main flavors.
Flavor profiles aren't an exact science. Tastes and notes may differ from one person to another. This depends on how developed a person's palette is and the experiences someone has. Your previous tastings are important with the association you make with a taste. In fact, when you taste the same drink at different moments, you might get different observations. You will develop your taste in time!
We have composed a flavor profile based on different available flavor profiles. Inspiration came from the book "Whisky, a tasting course" by Eddie Ludlow and Whisky Magazine's tasting wheel, and other sources. A flavor wheel is a common way to visualize the flavor profile. Please find the Tasting Intelligence Data-driven Flavor Profile in our interactive flavor wheel. You can hover over and click on it as it is an interactive visualization. Our flavor wheel is optimized for usage with our software. Find out how the optimizarion works in this article: word similarity
Flavors
Our data-driven flavor profile groups the flavors into these seven main groups:
- Woody
- Fruity
- Spicy
- Floral
- Cereal
- Feints
- Peaty
How does the software make an automated tasting wheel?
A tasting wheel is how you fill in the tasting of your drink inside the flavor wheel, it's personal. Making a tasting wheel from different observed notes gives interesting insights. Our tasting wheel represents the full experience. It's more than a factual description of the tasting notes. Everything written down in the review will be analyzed.
Our software follows these steps to make a tasting wheel based on reviews on the web:
- Search for reviews and scrape them from the web and store it (automatic copying)
- Combine all the reviews into one document
- Clean the text and make it ready to analyze
- Analyzing the text to discover the flavors from the data-driven flavor profile
- Make the tasting wheel
When analyzing we calculate the similarity of the document against our data-driven flavor profile. Our algorithm groups the results and maps them with the corresponding main flavor. Please find an example of an analysis in this article: 10 cane rum.
Links to other flavor profiles
As a reference we provide links to to other flavor profiles. You can use this to study flavors further and understand flavor profiles better. The links to the flavor profiles are:
