How do we differentiate between data capture steps and observation assertions steps/translations?
[ ] template descriptions (copy paste a "blank" taxon description)
[ ] matrix based methods (fill a matrix)
[ ] "instance" assertion -> just make observations in some tool and record the data
[ ] what is a template, how does a matrix approach differ from a template based
How do these relate to
[ ] open/closed world assumptions
[ ] serialization to instance/ontology based assertions
What does this mean for
[ ] getting out of the paradigm of character/character state constaints/limits
Classic example: blue and red should never be in a bin together, they are not end product results of processes that are remotely related.
A system for decsribing phenotypes should have principals (what's their approach, and why):
Eliminating pre-classification is key
removing the gaps between ontologies and descriptions
we want to exend definitions...fractal all the way down the rabbit hole, click-> zoom, click -> zoom
ontologies as macros for expanding definitions vs ontologies for discovering unseen relationships
trying to increase the understanding that the end product is a persistence layer, a bucket of completely logical statements, seperating observations from ontologies (buckets/files) is really just a practical operations step
ontologies as a "focus", a means of coming together to better expand our knowledge about a certain domain. I.e. AISM goal is to focus and improve our understanding of insect anatomy, the fact that it is persisted as a logical construct is completely bonus, what's important is the effort to describe and update the anatomy
Notes from conversation Sept. 14/2020:
How do we differentiate between data capture steps and observation assertions steps/translations?
How do these relate to
What does this mean for
Classic example: blue and red should never be in a bin together, they are not end product results of processes that are remotely related.
A system for decsribing phenotypes should have principals (what's their approach, and why):
Eliminating pre-classification is key
removing the gaps between ontologies and descriptions
we want to exend definitions...fractal all the way down the rabbit hole, click-> zoom, click -> zoom
ontologies as macros for expanding definitions vs ontologies for discovering unseen relationships
trying to increase the understanding that the end product is a persistence layer, a bucket of completely logical statements, seperating observations from ontologies (buckets/files) is really just a practical operations step
ontologies as a "focus", a means of coming together to better expand our knowledge about a certain domain. I.e. AISM goal is to focus and improve our understanding of insect anatomy, the fact that it is persisted as a logical construct is completely bonus, what's important is the effort to describe and update the anatomy