1 of 10

Tea Team

Cara Reedy & Katie Chastain

2 of 10

Overview of Ontology

  • Represents various teas and tisanes as SteepedDrinks with one or more Flavors and one or more BrewingTemperatures
  • BrewingTemperatures of unknown tisanes can be inferred from the part of a plant they are brewed from, allowing the ontology to deal with SteepedDrinks not yet represented
  • The main use case goal is to match combinations of SteepedDrink based on two criteria:
    • BrewingTemperature compatibility (straight matching)
    • Flavor compatibility (more complex)

3 of 10

Overview, continued

  • FlavorCompatibilities are a type of nanopublication, which we’re setting up to arrange in nanopub collections
  • Nanopub format includes assertion, provenance, and publication info
  • Nanopub collections bundle assertions together and are therefore useful for numerous assertions that have the same provenance and publication info.

​

  • Like for example … the recommendations in The Flavor Bible, or feedback from website users

4 of 10

Structure

5 of 10

SPARQL examples

“Does my Sencha green tea blend well with mint tea?”

​

ASK {

tea:SenchaGreenTea tea:hasFlavor ?flavorA ;

tea:hasBrewingTemperature ?temp .

tea:MintTea tea:hasFlavor ?flavorB ;

tea:hasBrewingTemperature ?temp .

?flavorA tea:isCompatibleWith ?flavorB .

}

​

Will it blend? That is the question….

6 of 10

SPARQL, continued

“I like the flavor of snow bud white tea and I want to blend it with rose hip tea, but their brewing temperatures are not compatible. What tastes like the snow buds, but will still make a compatible blend with the rose hips?”

SELECT DISTINCT ?tealabel where {

tea:snowBudWhiteTea tea:hasFlavor ?flavorA .

tea:roseHipTea tea:hasBrewingTemperature ?temp .

?othertea tea:hasFlavor ?flavorA ;

tea:hasBrewingTemperature ?temp .

?othertea rdfs:label ?tealabel .

}

​

7 of 10

SPARQL, continued continued

“I have a little bit of fancy Oolong tea left. What are all the possible things that would taste good in a blend with it?”

SELECT DISTINCT ?tealabel where {

tea:exampleOolong tea:brewsAt ?temp ;

tea:hasFlavor ?flavorA .

?othertea tea:hasBrewingTemperature ?temp ;

tea:hasFlavor ?flavorB .

?flavorA tea:isCompatibleWith ?flavorB .

?othertea rdfs: label ?tealabel .

}

​

8 of 10

Evaluation

9 of 10

Going forward

Modelling questions - blends

  • Originally, we were dealing with “atomic” steeped drinks: mint tea + sencha green tea
  • Now we are working out a way to deal with teas that are flavored but not necessarily blended, eg. Earl Grey tea = black tea + bergamot oil
  • Fortunately, we had already left ourselves with some wiggle room with the SteepedDrink metaconcept
  • It can also be considered a blend, however, which is how it's currently modeled: blend can be two+ SteepedDrinks or a SteepedDrink and a flavoring (such as spice or essential oil)

10 of 10

Other issues

  • For our linking to another ontology, we connected to the Plant Biodiversity Ontology, to add species information about the various plants we’re throwing in hot water
  • We examined another ontology, the Plant Ontology, which represents plant morphological structures and development (in a general sense to all species)
  • The PlantParts we use in the tea ontology are defined from a culinary perspective, though, rather than a botanical/scientific one
  • Botanically, a cucumber is a berry and a strawberry isn’t….