tdwg / ac

Audiovisual Core
http://www.tdwg.org/standards/638
Creative Commons Attribution 4.0 International
11 stars 6 forks source link

Implementation of subjectPart and subjectOrientation in Myriatrix #258

Open Archilegt opened 1 year ago

Archilegt commented 1 year ago

This comment will be edited. Placeholder issue for a test implementation of relevant Audiovisual Core Terms subjectPart and subjectOrientation in the Myriatrix Scratchpad. Issue open to comments: Yes. [This may change when/if extensive (re)writing is needed] subjectPart terms comments: Complete. subjectOrientation terms comments: Complete.

In this use case report, the contributing tester documents limitations to the implementation of the current features of the Audiovisual Core subjectPart and subjectOrientation vocabularies as Drupal vocabularies in the Myriatrix Scratchpad as per Baskauf et al. (2023). Software development is beyond the scope of the tester but CMS development is in scope. It should be noted that Scratchpads 2.0 comes with three default vocabularies for images: 1) Image keywords (free entry), and 2) Imaging technique and 3) Preparation technique, drawn from a restricted vocabulary defined within the Scratchpad that can be augmented by the user (see Smith et al., 2009). In the absence of controlled subjectPart and subjectOrientation vocabularies, Scratchpads can handle relevant image tagging through the image keyword vocabulary. It is not possible to define regions of interest in images in Scratchpads. One basic approach to partly address this limitation is to implement subjectPart tagging as multiple tags. The best way to achieve this in Myriatrix is to create a "Subject part" field, allowing multiple comma-separated values, and drawing from the Morphology glossary to autocomplete the terms while typing. The Biodiversity Heritage Library did use regions of interest for applying image tags (BHL, 2012). It was done through Flickr and "Box Tagging Multi-Species Images" but it could technically be applied to more granular image regions. As limitations, 1) the region is always a box (rectangle), 2) tags in the form namespace:predicate=value are not developed for the potential BHL-Flickr subjectPart and subjectOrientation AC equivalents, and 3) even if image searching and tagging remain available, the BHL Flickr account is on hiatus due to staffing changes and the account is neither monitored nor new images are uploaded.

Tasks:

General remarks

  1. The section "1.2 RFC 2119 key words" across subjectPart and subjectOrientation documents (and beyond) SHOULD be updated to "The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 and RFC 8174". Also, note that the RFC documents are citable. Example: [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/RFC2119, March 1997, http://www.rfc-editor.org/info/rfc2119.
  2. subjectPart terms can be reused from or implemented as morphological glossaries in databases. That is the case of the Morphology glossary in Myriatrix.
  3. The content of the subjectPart and subjectOrientation field "Definition derived from:" maps to the content of the Scratchpads field "Ontology Term URL". See Baker et al., (2019).
  4. The use of the subjectPart and subjectOrientation field "Has broader concept" (equivalent to an is_a relation) is acceptable for "male cone" and "female cone" (is_a cone), "forewing" and "hindwing" (is_a wing), "foreleg", "midleg", and "hindleg" (is_a leg), and "right side" and "left side" (is_a lateral side), but is counterintuitive for "flower" (part_of inflorescence), "cranium" and "mandible" (part_of skull), for which a field "Part of concept" may be better suited. In any case, the two meanings should not be mixed. Also note that, for consistency, "wing" and "leg" should have been created with a "Has broader concept" (part_of !!!) "thorax", and "antenna", "eye", and "skull" (when present) with a "Has broader concept" (part_of !!!) "head".
  5. It is recommended that in the Controlled value strings from SKOS Collections for ac:subjectPartLiteral a consistent listing order is chosen for entireOrganism and unspecifiedPart. See also comment on unspecifiedWholeOrPart and entireOrganism (body, wholeOrganism) below. Tentative listing order: 1) unspecifiedWholeOrPart, 2) entireOrganism, 3) unspecifiedPart.
  6. subjectOrientation term definitions excessively rely on the word "side", apparently imported from ontologies like the BSPO and neglect two key points: 1) Views are views from the point of the observer and not directly related to specimen structures and the word "view" instead of "side" is meant to be used in many cases. "Views" from the point of the observer belong in subjectOrientation while "sides" are parts and belong in subjectPart. 2) The word "side" is applicable to the halves of a bilaterally symmetrical body as defined by the median/longitudinal/sagittal plane of symmetry and should not be used instead of the word "end" to describe the anterior/posterior or rostral/caudal or proximal/distal or basal/apical halves as defined by a transverse plane. "Ends" are also parts and do not belong in subjectOrientation but in subjectPart. Please, use bio-ontologies critically. Many of them contain quite remarkable mistakes and inconsistencies.
  7. It may be useful to start looking at WordNet while compiling, maintaining, and translating TDWG vocabularies. It should be noted that the top noun in the WordNet hierarchy is "entity", which contrasts with subjectPart and subjectOrientation being concept schemes and their terms being concepts. Nevertheless, mapping of TDWG vocabulary terms to WordNet may prove useful, especially 1) while redacting term definitions and/or choosing among multiple definitions for one term, 2) for translating terms from English into other languages by taking advantage of the standardization of wordnets and uniformity of wordnets synsets across human languages, 3) for connecting TDWG vocabulary terms to ontologies through existing mappings between the WordNet synsets and the categories from ontologies, and 4) for looking into WordNet relations whenever advanced SKOS relations are needed for modelling TDWG vocabularies and keeping both sets of relations mapped. More specifically for Audivisual Core, its vocabularies, and future machine classification applications, and eye should be kept on ImageNet, an image database organized according to the WordNet hierarchy of nouns.

Specific remarks

subjectPart terms

Non-applicable subjectPart terms: 1) All plant terms: bark, bud, female cone, flower, fruit, inflorescence, leaf, male cone, seed, stem, strobilis (cone), twig. 2) Some animal terms: abdomen, cranium, fin, foreleg, forewing, hindleg, hindwing, midleg, shell, skull, thorax, wing.

Applicable subjectPart terms which will not/may not be implemented in Myriatrix: 1) unspecified part (unspecifiedPart): Hard to include it in an anatomical Drupal taxonomy in a way which makes sense. I have to give it a second thought. Also, note that this term implicitly assumes that there is a part, but implicit term hierarchy and usage also includes it being applicable to images which would contain an "entire organism". The implicit meaning and intended usage are better conveyed by a term unspecifiedWholeOrPart. Having the two terms would be perfect for machine-assisted applications. 2) genitalia (genitalia): 1) The term may work well for a few animal groups in which the structures may be continuous and unique, continuous and paired, or discontinuous but a single pair. It will not work so well for discontinuous paired structures in more than one pair, and/or a combination of unique and paired structures across multiple segments. For arthropods, having a term such as genital segment(s) may work better than genitalia in some cases, as segments and their appendages are continuous anatomical structures which are captured along with disconnected anatomical groups (paired gonopods, paired cerci, penis, ovipositor, etc.). 2) The addition of the word "outer" to the definition does not benefit its broader application, as in some groups the genitalia is internal and needs to be dissected and mounted in microscopic preparations (microslides) that will later be photographed or scanned. 2) The definition could be improved by changing the expression "sex organ" to "reproductive organ". "Sex" suggests copula, while genital structures such as the ovipositor are for egg-laying. For centipedes, the term genital segment(s) is preferred to genitalia and already implemented in Myriatrix but not yet in the ChilAO.

Applicable subjectPart terms with incompatible definitions: 1) mandible (mandible): The definition "A dentary bone that is the only bone in one of the lateral halves of the lower jaw" may work for humans/mammals/vertebrates but it does not work for animals which do not have bones/endoskeleton. The current definition for "mandible" in Chilopoda is "one appendage of the first pair of the mouth-parts", which should be probably improved to "one appendage of the first pair of gnathal appendages". Term already implemented in Myriatrix and in the ChilAO. 2) egg (egg): 1) This term does not belong in the subjectPart vocabulary but in the AC Taxonomic Coverage Vocabulary, where the term dwc:lifeStage and the example "egg" are already included. 2) Also in Scratchpads, in the Specimen/Observation content type, the field "Lifestage" (sic!) displays the suggestion "The age class or life stage of the biological individual represented by the catalogued object or observation. Examples: "adult", "mature", "juvenile", "eft", "nymph", "seedling", "seed", "egg". A controlled Scratchpads vocabulary does not exist, but if developed it would be applicable to both specimens and images. 3) The definition "The hard-shelled reproductive body produced by an animal" leaves out animal groups which produce eggs without a shell, e. g., arthropods, molluscs, fishes, amphibians, mammals. 4) The word "body" mentioned in the definition of "egg" should be changed to "structure" (or similar word) to avoid ambiguity with the word "body" mentioned in other three definitions (see point 4 in entire organism). In the definition, it could be made explicit that the egg is considered a life stage. The term "egg" will not be implemented in the Morphology glossary of Myriatrix and implementing a controlled lifeStage vocabulary is not planned at the moment.

Future subjectPart terms with incompatible definitions: 1) trunk (trunk): Not yet a term, but if added as it "should" (see Baskauf et al., 2023, table 3), then as a plant term its definition will conflict with the definition(s) of trunk as an animal term. This term is required for higher-rank groups such as Myriapoda (a subphylum) and Onychophora (a phylum) and the term itself is high in the anatomical hierarchy, e. g., it is a direct children of "body". Term "trunk" already implemented in Myriatrix. "Trunk" is a term explicitly mentioned by Lucking et al. (2021), who were focused on multiple annotation of words: "With regard to the word-level, mapping a single term to more than one annotation category could prove helpful (...) to resolve ambiguities (trunk as an anatomical feature of an animal or a plant). Thus, by assigning both annotation categories MORPHOLOGY and PLANT, the term trunk can be clearly identified as a plant characteristic." The same applies to annotation of images. Disambiguating "trunk" in image annotation should not be difficult as long as multiple annotation is used.

Observations on relevant subjectPart terms: 1) entire organism (entireOrganism): There is more than one issue here. 1) Note that wholeOrganism sounds more natural than entireOrganism, and it would also be more consistent with an unspecifiedWholeOrPart term (see above). The notion of a "whole organism" is also present in the definitions of the subjectOrientation terms "right side" and "left side" (it should also be in "lateral side" for consistency). 2) The term is intuitive, but "organism" is not really a term for anatomy and anatomical parts, at least in Zoology. For animals, the term "body" reflects the whole (e.g., anatomy of the human body, not the human organism). For an unicellular organism, the term "cell" may be enough. 3) An "entire organism" with the definition "An entire multicellular organism" neglects that an image of a unicellular organism also depicts an "entire organism". 4) The definition seems too much derived so as to cite CARO_0000012 multicellular organism as the source and also raises the issue that CARO does include uni- and multicellular organisms under cellular organisms (e. g., entire organisms) while subjectPart does not. 5) Word "body" is mentioned in the definition of four terms: "head", "thorax", "abdomen", and "fin". In Myriatrix, the term "body" is the parent term of the Morphology glossary. For the time being and to test new fields, I linked it to the term "body" in the Hymenoptera Anatomy Ontology (HAO). The term "body" in HAO has a definition which excludes antennae, legs and wings. The HAO definition is not aligned with any WordNet "body" definition, a partial match being definition 5 (the body excluding the head and neck and limbs). The HAO definition is also not aligned with the all-inclusive definition that the term will have in the Chilopoda Anatomy Ontology (ChilAO), which does match WordNet "body" definition 1 (the entire physical structure of an organism (an animal, plant, or human being)). "Body" is a term explicitly mentioned by Lucking et al. (2021) as one of the "WordNet’s 26 top-level entity categories (i.e., the unique beginner synset for nouns) for ontological classification". I checked the WordNet download page and downloaded the compressed file for WordNet 3.1. 1) In the WordNet database that I downloaded, there are 26 "noun." files, closely matching the list of 26 WordNet top-level entity categories mentioned by Lucking et al. (2021), e.g., 25 matches. There is one mismatch, it being the "noun.Tops" file in the database (that of course is not a category to be listed in the paper) and no noun.society file, while "society" is listed as one of the WordNet top categories in the paper. There is also no "society" entry in the "noun.Tops" file. 2) In the "noun.Tops" file that I downloaded there are not 26 but 51 "top-level" nouns, and like "society", the term "body, corpus" is not one of them. Looking into the "noun.body" file that I mentioned above, the term "body" is related to noun.Tops:natural_object and noun.Tops:animal, but that is all. This mismatch is of especial relevance, as biologists working on morphology ontologies, like myself, could rewrite from the paper something like "The term "body" is one of the 26 WordNet top-level entity categories (Lucking et al., 2021), and therefore we chose it as the parent term for our taxonomy of morphological terms.". Update: Dr. Andy Lucking provided a clarification on 23.06.2023, after consulting with colleagues: "There was a strange kind of misunderstanding. I knew that the categories are those below noun.Tops. However, "below" did not refer to the noun.Tops index (which I assumed) but to the table in this documentation: https://wordnet.princeton.edu/documentation/lexnames5wn ! So its file number 04 to 28. For some reason "society" had been added, however...". I replied that "I checked the lexnames5wn list and I think that it may also explain why "body" was listed. Files 04 to 28 are noun files, and then... file 29 is verb.body! So it may be that the word made it into the main noun list of the article by means of consecutive association."

2) head (head): The definition is "Anterior-most division of the body". The use of "division" restricts the definition and the word should be changed to "region" to keep "head" as broadly applicable as possible. Some organisms do not have a head visibly "divided" from the rest of the body. The addition of "...frequently bearing structures such as eyes, antennae, and mouthparts" to the definition may be useful, and there are related terms (eye, antenna, mandible) in subjectPart. Term already implemented in Myriatrix. Additional consistency remarks: These remarks are based on head, thorax, and abdomen terms. 1) The previously addressed word "division" (head definition) is also present in the final, more consistent (than raw version) definitions of "thorax" and "abdomen", which previously used the word "section" (thorax definition) and the word "tagma" (abdomen definition, incorrect singular when plural was intended). As the two definitions are focused on insects, "tagma/tagmata" is the most correct word to use there, the word "division" would be fine but the word "region" would be best because it is also applicable to "head". 2) Regarding "Anterior-most division" (head definition), "The middle division" (thorax definition) and "The posterior-most division" (abdomen definition), note that A) adding "The " to the "head" definition would improve consistency, and B) whenever only two or three divisions/regions/tagmata are involved, using the suffix "-most" for "anterior" and "posterior" is unnecessary.

3) leg (leg): No remarks at the moment. Term already implemented in Myriatrix.

4) antenna (antenna): No remarks at the moment. Term already implemented in Myriatrix.

5) eye (eye): No remarks at the moment. Term already implemented in Myriatrix but not yet in the ChilAO. Relevant terms for Chilopoda (centipedes) are compound eye and ocellus (plural ocelli).

Human-readable collection list of controlled value strings to be made available for centipedes in subjectPart

centipedes https://ac.tdwg.org/part_collections#centipedes An effort was made to align the subjectPart vocabulary for centipedes with the higher-hierarchy structures used by Bonato et al. (2010): cephalic capsule, antenna, clypeus and labrum (as clypeolabrum), mandible, first maxillae (as the singular first maxilla), second maxilla (as the singular second maxilla), forcipular segment, leg-bearing segment, leg, ultimate leg-bearing segment, and terminal part of the body (as the singular genital segment). Terms which are not higher-hierarchy structures in but are listed by Bonato et al. (2010) include the appendage terms forcipule, ultimate leg, coxopleuron, and gonopod. Other terms not in Bonato et al. (2010) but important for the overall structure of the Morphology glossary in Myriatrix and for the centipedes subjectPart vocabulary include: head, eye, trunk, forcipular apparatus, and genital segment. All centipede appendages are included in this subjectPart vocabulary.

subjectOrientation terms

Non-applicable subjectOrientation terms: 1) Plant terms: lower side, upper side. 2) Cnidarian and echinoderm terms: aboral side, oral side.

Applicable subjectOrientation terms which will not/may not be implemented in Myriatrix: 1) apical side (apical): A part and not a view, an end and not a side. It does not belong in subjectOrientation but in subjectPart. In the case that conversion into a view is needed, options are "apical view" and "apical side view". 2) basal side (basal): A part and not a view, an end and not a side. It does not belong in subjectOrientation but in subjectPart. In the case that conversion into a view is needed, options are "basal view" and "basal side view".

Observations on relevant subjectOrientation terms: 1) unspecified orientation (unspecifiedOrientation): No remarks at the moment. 2) anterior side (anterior): From the point of view of a transverse plane, it is not a side but an end. From the point of view of the observer, it is not a side but a view. In Myriatrix, the implemented term will be anterior view (anterior). 3) posterior side (posterior): From the point of view of a transverse plane, it is not a side but an end. From the point of view of the observer, it is not a side but a view. In Myriatrix, the implemented term will be posterior view (posterior). 4) lateral side (lateral): From the point of view of the observer, it is not a side but a view. In Myriatrix, the implemented term will be lateral view (lateral). 5) right side (right): From the point of view of the observer, it is not a side but a view. In Myriatrix, the implemented term will be right side view (right). 6) left side (left): From the point of view of the observer, it is not a side but a view. In Myriatrix, the implemented term will be left side view (left). 7) dorsal side (dorsal): From the point of view of the observer, it is not a side but a view. In Myriatrix, the implemented term will be dorsal view (dorsal). 8) ventral side (ventral): From the point of view of the observer, it is not a side but a view. In Myriatrix, the implemented term will be ventral view (ventral).

New subjectOrientation terms 1) ectal view (ectal): See "ultimate leg" term for an explanation. 2) mesal view (mesal): See "ultimate leg" term for an explanation.

Human-readable collection list of controlled value strings to be made available for centipede parts in subjectOrientation

unspecified whole or part https://ac.tdwg.org/orient_collections#unspecified-whole-or-part New subjectPart term.

entire organism Equivalent term in Myriatrix is "body". List trimmed (oral, aboral, apical, and basal are out). Adding the list here for Myriatrix's reference.

unspecified part https://ac.tdwg.org/orient_collections#unspecified-part Apparently represented by unspecified morphological feature.

head Without changes. Adding the list here for Myriatrix's reference.

cephalic capsule https://ac.tdwg.org/orient_collections#cephalic-capsule New subjectPart term.

antenna Originally only the "lateral" view was listed.

eye Originally only the "lateral" view was listed.

clypeolabrum https://ac.tdwg.org/orient_collections#clypeolabrum New subjectPart term.

mandible List trimmed (lateral is out). Adding the list here for Myriatrix's reference.

first maxilla https://ac.tdwg.org/orient_collections#first-maxilla New subjectPart term.

second maxilla https://ac.tdwg.org/orient_collections#second-maxilla New subjectPart term. New subjectOrientation terms (ectal, mesal), see "ultimate leg" term for explanation.

trunk https://ac.tdwg.org/orient_collections#trunk New subjectPart term. Animal term. Posterior view is possible but unlikely to be used for "trunk".

forcipular segment https://ac.tdwg.org/orient_collections#forcipular-segment New subjectPart term.

forcipular apparatus https://ac.tdwg.org/orient_collections#forcipular-apparatus New subjectPart term.

forcipule https://ac.tdwg.org/orient_collections#forcipule New subjectPart term. New subjectOrientation terms (ectal, mesal), see "ultimate leg" term for explanation.

leg-bearing segment https://ac.tdwg.org/orient_collections#leg-bearing-segment New subjectPart term. The posterior view is possible for the "ultimate leg-bearing segment", which is addressed separately.

leg Only a lateral view is included at the moment, which is remarkable considering that many insects have the locomotory legs aligned with transversal planes, and therefore the legs have an anterior and posterior side (and anterior and posterior structures such as spurs and claws) and leg images (especially those of detached legs) can be oriented in anterior and posterior views.

ultimate leg-bearing segment https://ac.tdwg.org/orient_collections#ultimate-leg-bearing-segment New subjectPart term.

coxopleuron https://ac.tdwg.org/orient_collections#coxopleuron New subjectPart term. New subjectOrientation term (ectal, applicable to a coxopleuron with developed coxopleural process), see "ultimate leg" term for explanation.

ultimate leg https://ac.tdwg.org/orient_collections#ultimate-leg New subjectPart term. New subjectOrientation terms (ectal, mesal). For paired appendages aligned with the sagittal body axis, such as the ultimate legs of centipedes, a general lateral view can be considered but it may be confusing. For clarity, there are four sides and views involved as per the leg orientation: ectal side and view (which can depict either the right or left ultimate leg) and mesal side and view (which can depict either the right or left ultimate leg). The ectal side of the ultimate leg corresponds to the anterior side of a locomotory leg and the mesal side of the ultimate leg corresponds to the posterior side of a locomotory leg. The "right side (view)" and the "left side (view)" of the "entire organism" or "body" SHOULD NOT be applied to ultimate legs of centipedes.

genital segment https://ac.tdwg.org/orient_collections#genital-segment New subjectPart term.

gonopod https://ac.tdwg.org/orient_collections#gonopod New subjectPart term. New subjectOrientation term (ectal), see "ultimate leg" term for explanation.

Implementation

Myriatrix's Morphology vocabulary

Name: Morphology Machine-readable name: morphology Description: Lexicon dealing with the form of Myriapoda and Onychophora, and with relationships between their structures. [Notes: Currently, this vocabulary is a modified implementation of the Chilopoda Anatomy Ontology (ChilAO) and the Audiovisual Core subjectPart vocabulary.] Biological classification: Not a biological classification Display options: [x] Display tab [x] Display facets Default taxonomy term: Body

Multilingual options - Translation mode: No multilingual options for terms. Only the vocabulary will be translatable.

Revision information: [x] Create new revision by default [x] Show revision attribution on taxonomy overview pages. Note: This may effect performance when displaying your taxonomy page. [x] Show revision attribution with revision log

Path: https://myriatrix.myspecies.info/admin/structure/taxonomy/morphology Term count: 22 [Note: The count refers only to the subjectPart subset, not to the entire set of terms in the Morphology vocabulary.] Term index: unspecified whole or part | body | unspecified part | head | cephalic capsule | antenna | eye | clypeolabrum | mandible | first maxilla | second maxilla | trunk | forcipular segment | forcipular apparatus | forcipule | leg-bearing segment | leg | ultimate leg-bearing segment | ultimate leg | coxopleuron | genital segment | gonopod

Morphology Myriatrix

1) unspecified whole or part (unspecifiedWholeOrPart) 2) body (entireOrganism): This term is equivalent to entireOrganism in AC subjectPart. 3) unspecified part (unspecifiedPart) 4) head 5) cephalic capsule 6) antenna 7) eye 8) clypeolabrum 9) mandible 10) first maxilla 11) second maxilla 12) trunk 13) forcipular segment 14) forcipular apparatus 15) forcipule 16) leg-bearing segment 17) leg 18) ultimate leg-bearing segment 19) ultimate leg 20) coxopleuron 21) genital segment 22) gonopod

Myriatrix's Subject orientation vocabulary

Name: Subject orientation Machine-readable name: subject_orientation [Note: The machine-readable name must contain only lowercase letters, numbers, and underscores.] Description: This vocabulary is a modified implementation of the Audiovisual Core subjectOrientation vocabulary. Biological classification: Not a biological classification Display options: [x] Display tab [x] Display facets

Multilingual options - Translation mode: No multilingual options for terms. Only the vocabulary will be translatable.

Revision information: [x] Create new revision by default [x] Show revision attribution on taxonomy overview pages. Note: This may effect performance when displaying your taxonomy page. [x] Show revision attribution with revision log

Path: https://myriatrix.myspecies.info/admin/structure/taxonomy/subject_orientation Term count: 10 Term index: unspecified orientation | anterior view | posterior view | dorsal view | ventral view | lateral view | right side view | left side view | ectal view | mesal view

Subject orientation Myriatrix

1) unspecified orientation (unspecifiedOrientation) 2) anterior view (anterior) 3) posterior view (posterior) 4) dorsal view (dorsal) 5) ventral view (ventral) 6) lateral view (lateral) 7) left side view (left) 8) right side view (right) 9) ectal view (ectal) 10) mesal view (mesal)

Acknowledgements

The Natural History Museum, London continuously supports maintenance of Myriatrix. Dr. Andy Lücking (Text Technology Lab, Goethe-Universität, Frankfurt am Main) provided clarifications regarding the use of the term "body" in his BIOfid multiple annotation paper.

References

baskaufs commented 1 year ago

I haven't looked at this comprehensively, but wanted to note that the labels and definitions for existing subjectPart and subjectOrientation terms have already been translated into several languages, including Spanish. See https://tdwg.github.io/rs.tdwg.org/ for links to the JSON files containing the translations.

Archilegt commented 1 year ago

Hi, @baskaufs. I think that now the text in the first comment is informative enough so as to allow a first round of comments. Re the translations that I mentioned in my first comment: I meant translations of new terms and new or changed definitions. But that is sort of a final task. The first task that I would like to get done is to "Create a subjectOrientation Drupal vocabulary in Myriatrix, with all relevant orientation terms for centipedes", but just the taxonomy of terms at the moment. Data fields for Drupal vocabulary terms can be created later. I would like to create as many Drupal vocabulary data fields as needed to align and record the information from the AC subjectPart and subjectOrientation terms' data fields (e. g., "AC Part Term IRI", "AC Part Label", "AC Part Definition", etc.). I am also available to meet online if you prefer. Is there any upcoming AC meeting?

baskaufs commented 1 year ago

Hi @Archilegt . Thanks for going through the terms carefully and I'm excited that you are experimenting with implementing it in Scratchpads. A few responses:

General Comment 1. Thanks for this information. We will work on updating the RFC 2119 statement on our standards documents as we are able, probably as we publish new versions of the documents.

General Comment 4. By modeling TDWG controlled vocabularies as SKOS ConceptSchemes, we are being less precise in the semantic relations between terms than might be the case for a formal ontology. In the case of skos:broader, the discussion of broader/narrower relationships in the SKOS Primer notes that they are used "To assert that one concept is broader in meaning (i.e. more general) than another...". In particular, the commentary notes "Not covered in basic SKOS is the distinction between types of hierarchical relation: for example, instance-class and part-whole relationships." So asserting that there is a broader term just means that the broader term is more general without being specific about the way it is more general.

The detailed comments that you have made are very useful and I am grateful for you taking the time to give such a careful look at the terms and their definitions. One thing that I will note as a general comment is that the task group made a conscious effort to make the term labels and definitions as uncomplicated as possible and to avoid being more granular than necessary. This was a practical decision intended to make the controlled vocabularies usable by people who were not technical experts, both on the side of assigning values to images (for example by a student worker) and on the side of someone who was searching based on values in the vocabularies. So for example, although there may be technical distinctions in the names applied to analogous parts in different organism groups, we chose to use simple terms like "leg" even though there might be more specific names used in some groups (because that's the kind of term a general user might search for). We also opted not to be too granular in the parts that we specified for groups. During development, we looked at a system that described neuropteran parts in great detail, but concluded that most users would not be using the vocabularies at that level of granularity. If through use of the vocabularies it was determined that they were not granular enough, additional more specific terms could be added if they were needed. I won't say more about this here since our discussions on the topic are documented in the task group meeting notes.

The other thing that I would mention is that linking the terms to external ontology terms is a somewhat new thing for TDWG vocabularies and has only recently been done for several Audiovisual Core controlled vocabularies. So I'm not sure that there are very clear guidelines about how it should be done and about what the implications are of making those links. For discussion on this point, I would ping Jennifer Giron (can't find her GitHub username) who is interested in developing links to ontologies and @DavidFichtmueller who is organizing a group on mappings between standards.

I guess the main question that I have at this time is what action you are hoping for from the Audiovisual Core Maintenance Group? If the task group had received this kind of feedback during the development or public review phases of the vocabularies' development, it would have been relatively easier to incorporate some of your suggestions. However, at this time, the task group is disbanded and changes to the vocabulary would need to be made via a series of term change requests to the AV Maintenance Group using the process described here.

The AC Maintenance Group does not currently have regularly scheduled meetings, so to take actions for changing the vocabularies, the relevant change proposal issues would need to be made in the issue tracker and then a meeting would need to be called by @edwbaker . To make changes to the normative definitions of terms or to add new terms is not particularly simple since it requires following the process I linked above. Making changes to how the part groups are organized among organism groups does not require any particular process since those associations are managed outside of the official standard. The AC MG would just have to agree that they made sense and we could make the changes.

Archilegt commented 1 year ago

Pinging @JCGiron

JCGiron commented 1 year ago

@Archilegt Thank you for bringing me into this discussion and for taking the time to critically review the terms and definitions including in the Views Controlled Vocabularies. The first thing that strikes me is that you are using Scratchpads, which are soon to be retired. Do you have plans to migrate somewhere? I would like to learn more about ChilAO. Are you using the Ontology Development Kit? Also, you probably have seen it: we recently published a paper about the development of the Ontology for the Anatomy of the Insect SkeletoMuscular system (AISM): Girón JC, Tarasov S, González Montaña LA, Matentzoglu N, Smith AD, Koch M, Boudinot BE, Bouchard P, Burks R, Vogt L, Yoder M, Osumi-Sutherland D, Friedrich F, Beutel R, Mikó I (2023) Formalizing Invertebrate Morphological Data: A Descriptive Model for Cuticle-Based Skeleto-Muscular Systems, an Ontology for Insect Anatomy, and their Potential Applications in Biodiversity Research and Informatics. Systematic Biology: syad025. https://doi.org/10.1093/sysbio/syad025. I can send you the accepted version (in case you don't have access; it is not yet in the journal's format), but I'll need your email.

Archilegt commented 1 year ago

Hi, @JCGiron! Many thanks for writing!

  1. Scratchpads migration: Migration is not possible because there is nothing like Scratchpads. The people using Scratchpads (myself included) can only downscale whatever we are doing by moving to the next best virtual research environment. Which VRE is the next best one depends on the individual Scratchpad. In the case of Myriatrix, which is global in scope and aims at displaying as much data as possible online, the next best VRE seems to be the Living Atlases. I think that you may be asking because of TaxonWorks. TW was considered but it didn't come on top.
  2. The end-of-life for Drupal 7 has been pushed back to January 5, 2025. I have asked Vince Smith (NHM London) if Myriatrix could be kept active until then but he has not yet replied.
  3. ChilAO development: The ChilAO and the Morphology glossary in Myriatrix are developed with the philosophy of bringing broader communities closer to ontology building. Therefore, the basic file for the ChilAO is one Excel file which is later converted into the ontology proper. The ChilAO v1.3 is accessible here and here. My colleague José Aldana is the developer of the conversion script.
  4. Thanks for offering a copy of the AISM! I recently got it from Markus Koch, who developed the LepAO under the BIOfid project some time ago. I will read the AISM paper as soon as I have time.

Hi, @baskaufs! Many thanks to you too! I will reply in my next time window!