Skip to content

Meeting Notes 18.06.2021

Jump to bottom
Casper Welzel Andersen edited this page Nov 5, 2021 · 1 revision

Participants:

  • Casper Andersen
  • Antanas Vaitkus
  • Emanuele Ghedini
  • Matthias Büschelberger
  • James Hester
  • Jesper Friis
  • ?

New annotation-centric ontology

Emanuele shows the ontology that he and Jesper have developed based on the discussions at the latest meeting (21.05.2021). They developed ontology branches based on the CIF 2.0 formal syntax and grammar, to keep the data names present in this grammar.

There is an issue with Loop and different data names, however CIF enforces that a loop can only contain data names from a single category. Emanuele suggests to constrain a Loop logically so that if one finds any data name in the loop, the constraint will be that any other data names have to belong to the same group super class/concept. The key data name has to be the same (in CIF terms).

Loop sub-categories

Antanas mentioned Loop sub-categories, James explains: An example can be made with anisotropic displacement parameters (a set of data values that represent a matrix). These can be supplied in a separate loop other than the general loop for atoms and positions, even though they technically belong to the same loop table and key data name. Emanuele: One can probably relegate this issue via a pointer and disregard the syntax, instead navigating the data semantically. Essentially, one can create a direct relation between the “foreign key”-column, in this example the atom label.

Dictionary -> DictionaryDefinedItem

Steps:

  • Generate ontology from dictionaries
  • Populate the knowledge graph/triple store with actual examples - individuals for the ontologies (in OWL-DL or similar)

Example CIF: https://www.crystallography.net/cod/1000000.cif

Clone this wiki locally