Child concept with no broader concept

[elcioabrahao]

The algoritm founded one ocorrence of a child concept without a broader concept on the simple classification sheet, line 2141 of version 11 of CLADIMED. The concept labeled G81ZZ did not have a broader concept G81Z breaking the tree hierarchy. I would like to know how to proceed in this case. Tks!

[elcioabrahao]

I process the hole file to found other ocorrencies:

Total of 2 on the SIMPLE CLASSIFICATION:

Simple Classif Begin conversion, find broader and narrowers...
No broader concept founded - PrimaryEntry [f=G, sf=81, g=Z, sg=Z, ns=, code=G81ZZ, rowNumber=2141, broaderConcept=G81Z, narrowerConcepts=[]]

No broader concept founded - PrimaryEntry [f=M, sf=99, g=Z, sg=, ns=, code=M99Z, rowNumber=2833, broaderConcept=M99, narrowerConcepts=[M99ZB, M99ZA, M99ZZ]]
Done...

and 5 on the COMPLETE CLASSIFICATION:

Complete Classif Begin conversion, find broader and narrowers...
No broader concept founded - PrimaryEntry [f=R, sf=60, g=Z, sg=Z, ns=, code=R60ZZ, rowNumber=4300, broaderConcept=R60Z, narrowerConcepts=[]]

No broader concept founded - PrimaryEntry [f=G, sf=81, g=Z, sg=Z, ns=, code=G81ZZ, rowNumber=2617, broaderConcept=G81Z, narrowerConcepts=[]]

No broader concept founded - PrimaryEntry [f=F, sf=58, g=A, sg=B, ns=, code=F58AB, rowNumber=2031, broaderConcept=F58A, narrowerConcepts=[]]

No broader concept founded - PrimaryEntry [f=F, sf=58, g=A, sg=A, ns=, code=F58AA, rowNumber=2029, broaderConcept=F58A, narrowerConcepts=[]]

No broader concept founded - PrimaryEntry [f=M, sf=99, g=Z, sg=, ns=, code=M99Z, rowNumber=3552, broaderConcept=M99, narrowerConcepts=[M99ZB, M99ZA, M99ZZ]]
Done...

[jonquet]

This is a good catch.
Indeed no broader concept for those.
It seems some are "new in 2017" concepts like G81ZZ, M99Z but some the others are not in 2017 (nor in 2016 for which I have a local version).

While waiting from the source file to be corrected, I would suggest to assign the broader concept of the broader concept as a fix. For instance, G81ZZ shall have G81 has broader concept.

I contact the Cladimed folks.

[elcioabrahao]

Ok ! Tks !

[elcioabrahao]

Founded other ocorrence: 2 level without broader:

Complete Classification line 2031 and folowers: there is no broader concept F58A and there is no concept F58, so I extend the broader to TOP CONCEPT: F.
Interesting to note that the LABEL said Abandon voir ... but there is no Observation on the H column about this set of concepts:

F 58 A A F58AA Abandon voir G62FA
F 58 A A 1 F58AA01 Abandon voir G62FA01
F 58 A B F58AB Abandon voir G62FB
F 58 A B 1 F58AB01 Abandon voir G62FB01

Please ask CLADIMED folks about this... TKS !

[jonquet]

For every concepts with "Abandon" in the label name, it means they have been classified as obsolete. Interestingly, they are not all mentioned in the "Abandon décembre 2016" tab, which probably means they have been declared obsolete in previous versions.

For those concepts, I would recommend to:

• tag them with the annotation property owl:deprecated = true
• tag them with the annotation property skos:changeNote = column G
• assign skos:label as the ID not column G
• tag them with the annotation property prov:invalidatedAtTime = 2016-12 if they are present in tab ""Abandon décembre 2016"
• assign rdfs:seeAlso of the URI corresponding to the ID in the sentence "Abandon voir ID"

[elcioabrahao]

Fonded 881 concepts with ABANDON on the LABEL column. I founded ABANBON word too (think it was a misspelling, but I consider too). Prodeded on the filter as your recomendation only for the itens witch have no information on the H column. If there is ABANDON on the LABEL and on the H column I considered the information on the H column instead of the label.

abandon_concepts.txt

[jonquet]

The extraction of the deprecated concepts shall be useful. Indeed "abanbon" is a misspelling.
I would just proceed as recommended for all these concepts when defining them in SKOS.