Document the community's preferred technical specification for linking trait data to occurrence data - Githubissues

hiscom / business

HISCOM Business

2 stars 1 forks source link

Document the community's preferred technical specification for linking trait data to occurrence data #85

Closed ben3000 closed 3 years ago

ben3000 commented 6 years ago

Document the community's preferred technical specification for linking trait data to institutional and/or AVH occurrence records. The resulting document should be in HISCOM's Business wiki.

ben3000 commented 3 years ago

We need to find or define a methodology that adopts the vocabulary approach so that we can share this data using Darwin Core styles.

ben3000 commented 3 years ago

We just want to be able to store this data at the specimen level if a research scientist makes measurements.

AaronWilton commented 3 years ago

There are some separate parts to this 1) ensuring that trait data sets are given good metadata (that describe the provenance, link to pages etc at data set level) and stored somewhere "secure/permanent" 2) add annotation or some type linkage between the specimens and the data set 3) gold standard: document each field in the data set with methods, units etc etc

ben3000 commented 3 years ago

In regard to my comments about specimen-level traits, I've found this paper by Henning et al. (2018) to be very interesting reading.

ben3000 commented 3 years ago

@AaronWilton what metadata standards are you referring to? EML?

ben3000 commented 3 years ago

There is currently no single standard for representing traits as a vocabulary of terms. Attempts have been made over the years to define a core set of traits for taxon-level descriptions, but the issue continues to be that each trait can be defined in a different manner by different systematists. A trait service is needed that can store and provide access to the following details of a trait definition:

The existence of a defined trait
The methodology used to define the trait, e.g. leaf length and how it is calculated
The units of the trait; important for combining measurements across projects
Any further suitable metadata

With this, projects would then begin using the available traits as URIs in their data systems in the same way we currently use Darwin Core terms.

afuchs1 commented 3 years ago

HISCOM AGM discussion Ely advised that the ALA is not planning to address Traits in the Atlas. This will be better considered in discussions about the ‘extended specimen’. AusTraits is dealing with some trait data. ANBG has delivered morphological data on seeds to AusTraits. Close issue – will address in the future when required.