Talk:Terminology
== TOTh workshop 2021
Notes, remarks and ideas arising from the TOTh workshop 2021.
What is a term?
Observation: Sometimes, terminologists and non-terminologists have different intuitive interpretations of "term". This should be made explicit in order to ensure consistent conversion.
Suggestion: introduce an explicit termlex:Term class for better orientation for terminologists.
Assumption (tbc.): a term has one single language and one or more possible (grammatically related) forms. That is, termlex:Term rdfs:subClassOf LexicalEntry
"lexicographic article"
Observation: In terminological resources, terms can be grouped into together into groups of meaning-equivalent terms. The lexicographic article will typically provide a definition, but can also provide information that does not refer to an underlying unit of meaning, e.g., historical notes on the use of the term, cf. https://www.termcat.cat/en/cercaterm/power?type=basic, second result ("Note" on earlier use of "power" [cat. poder] by Max Weber -- does that refer to the coining of the term or to the meaning?)
Suggestion: add lexicog:Entry to the diagram as a device for grouping terms. In terminology, the lexicog:Entry is usually (always?) defined by a (possibly implicit, but usually defined) concept. Note that no termlex-specific vocabulary is required.
Question: what would be typical attributes of a lexicographic article that unambiguously cannot apply to a concept
termlex:Concept
Observation: LexicalConcept is used for a broad band-width of phenomena, sometimes these are translation equivalent (WordNet Synset?), sometimes not (VerbNet predicates/frames???). We must provide a more concise, domain-specific definition. Also, the term "LexicalConcept" (coined to distinguish from externally defined "ontological concepts") was felt to be misleading/counterintuitive to terminologists (Rute Costa, p.c.).
Note: should always carry a definition
Suggestion: Add termlex:Concept as a designated subclass of LexicalConcept
Do we need termlex:lexicalizedConcept ?
Observation: ontolex:LexicalConcept is a subclass of skos:Concept, so, stating that a given skos:Concept
- is* an ontolex:LexicalConcept (resp., termlex:Concept) is equivalent with termlex:lexicalizedConcept and
does not require an independently defined entity.
If such a distinct entity is required, this can also be expressed by `owl:sameAs` or `skos:exactMatch` (etc.). These do not require novel vocabulary.
Suggestion: remove termlex:lexicalizedConcept
Giorgio Maria Di Nunzio and Federica Vezzani model
Seems to be directly interpretable against OntoLex concepts.
- Characteristic [what is that?]
- @name
- @variety
- --group--> Characteristic
- Term ~ LexicalEntry [designation is confirmed to be unique]
- @identifier
"e.g. URI" - @designation
"characters"
?single designation?: confirmed
- @identifier
- Concept ~ LexicalConcept
- @identifier
- -1-denoted-n-> Term ~ termlex:isEvokedBy ?
- -1-intension-n-> Characteristic
- --hierarchical--> Concept # hypo
Some representative resources
(Thanks to Pascal Vaillant, Rute Costa & TOTh participants)
- TermCat
- FranceTerm
- https://www.nlm.nih.gov/research/umls/knowledge_sources/metathesaurus/index.html: https://www.ncbi.nlm.nih.gov/books/NBK9685/table/ch03.T.concept_names_and_sources_file_mr/; documentation https://www.ncbi.nlm.nih.gov/books/NBK9676/
- https://www.ortolang.fr/market/terminologies, e.g., linguistics
Cimiano (2015) model
The following was developed in BPMLOD: https://www.w3.org/community/bpmlod/wiki/Converting_TBX_to_RDF#Mapping_the_TBX_Data_Model_to_the_ontolex-lemon_model
Probably also the basis of
- https://github.com/cimiano/tbx2rdf
- http://tbx2rdf.lider-project.eu/converter/tbx2rdf.html
- http://copyrighttermbank.linkeddata.es/
Maybe also the much more recent Pret-a-LLOD implementation
Patricia's (2020) model
Not sure this is properly documented anywhere, but it's still in the history: