Data Model: Attribute

Attribute

An attribute is is the quality of a substance (from Barry White)

Objective Attributes

Objective attributes are things that are not subject to individual subjective interpretation, they could be automated without human intervention. These are measurements like dimension (height, width, length) or mass/weight. You can take a ruler or scales and everyone following the same protocol should get the same result within parameters.

Environment Attributes

Environment attributes should be fairly easy to define, as at this time these are easy things like temperature, percent humidity and light (LUX or PAR). Since we have few sensors we will have few attributes. I suggest we work through this category before moving to phenotype attributes.

Phenotype Attributes

Phenotype starts off simple, but quickly gets quite complicated. I recently reviewed the OBO Ontology, and realized that the plant ontology (PATO) no longer exists! So much for consistent standards. As much as we can I suggest we follow some standard for naming plant parts, growth stages and the like. What we need to avoid is allowing everyone to come up with their own subject and attribute names and value units.

Subjective Attributes

Subjective attributes are determined in a person's head: "Do I like the flavor?", "Does it have good color?" Subjectivity should be avoided whenever possible, but it cannot be avoided and can be helpful at times. A good example of subjective attributes is wine tasting, there are even ISO standards for the drinking glass. Possible attributes could be:

• Taste
• Sweetness
• Bitterness

Subjective Measurement Scales

I would recommend that several subjective value scales be agreed upon using a 1 to 10 scale. These scales could include:

• Like - Dislike
• Agree - Disagree