StocksAndFisheriesKB

Thanks @marketak@ics.forth.gr
Please here follow two points.

1)if I am not mistaken, the JSON I see in the above example is not exactly reflecting the structure of the semantic identifier for a traceability unit.

The first area (asfis:BWA+fao:57;fao:81) is the assessment area (inherited from the stock ce69cabb-d2bf-30a6-bb79-08f4ce857de8) while the second area (fao:81) is the fishing area/management area (from this uuid 38fd8998-916a-3bac-b44c-b478e8223315).

If this is correct, then the JSON should contain one field for assessment area, and a distinct one for fishing/management area.

I made this comment also by checking the requirements summarized here https://support.d4science.org/projects/stocksandfisherieskb/wiki/GRSF_Model-Traceability_Unit

And I remember we agreed to split two types of areas in the semantic identifier and put in between the management authority, which in fact you already do
asfis:BWA+fao:57;fao:81+authority:NAT:AUS+fao:81+iso3:AUS+isscfg:09.31

In conclusion, if the above is fine, then the JSON should be modified accordingly.

2) "refers_to": it seems fine, but it does not say what is the stock and what the fishery, do you need to pass this info to CKan or will it be inferred somehow when building the traceability unit page?

Thanks Yannis, looks great.

For the short name,
e.g. "short_name" : "Bluenose warehou | Southeastern Australia | Ocean Trap and Line Fishery (NSW) | Australia | Set longlines - Bluenose warehou - Southeastern Australia",

I think that is rather a GRSF name which you build from the semantic identifier, while a short name might be added by the owner (e.g. FishSource) later via the management panel in case this is needed. What do you think?

It is not actually constructed from the semantic identifier, the traceability_unit_name is constructed using that.
The short name contains common names for species and areas. I think it is taken from the fishery's short name.
As you mentioned this can be changed afterward, so we have to provide something for the short name when publishing the record.

To avoid confusing Luca (because this is a contents-related issue) the field short_name should exist in the json.

ok fine with me and thanks.

@marketak@ics.forth.gr the example does not contain grsf_type field which should has as value the constant "Traceability Unit".

https://support.d4science.org/projects/stocksandfisherieskb/wiki/GRSF_Model-Traceability_Unit

@marketak@ics.forth.gr I have modified the wiki at https://support.d4science.org/projects/stocksandfisherieskb/wiki/GRSF_Model-Traceability_Unit to include the name of the json field provided by the KB.

Regarding the following fields

"referring_stock_record" : {
    "id" : "8141a8ea-a40f-4681-961d-4ffce9f7310a",
    "url" : "https://data.d4science.org/ctlg/GRSF_Admin/ce69cabb-d2bf-30a6-bb79-08f4ce857de8"
},
"referring_fishery_record" : {
   "id" : "63564232-dcbb-42f4-9580-6ba20fe572d3",
   "url" : "https://data.d4science.org/ctlg/GRSF_Admin/38fd8998-916a-3bac-b44c-b478e8223315"
},

I think they could be something like:

"referring_stock_record" : "https://data.d4science.org/ctlg/GRSF_Admin/ce69cabb-d2bf-30a6-bb79-08f4ce857de8",
"referring_fishery_record" : "https://data.d4science.org/ctlg/GRSF_Admin/38fd8998-916a-3bac-b44c-b478e8223315"

What do you think?

According to the wiki at https://support.d4science.org/projects/stocksandfisherieskb/wiki/GRSF_Model-Traceability_Unit

The Traceability Unit should have mandatory field:

• GRSF Semantic Name
• Management Reporting Area

I didn't find such fields in the example JSON.

Instead in JSON example provided I find:

"traceability_record_uri":"https://github.com/grsf/resource/traceability_unit/f2b97baf-4212-337e-9d88-ec5cc59376ab",
      "database_sources" : [ {
        "name" : "FishSource",
        "description" : "FishSource is an online information resource about the status of stocks and fisheries, that compiles and summarizes all the information that is needed from analysts to evaluate the sustainability.",
        "url" : "http://www.fishsource.com"
      }],
"fishing_area" : [ "Code: 81, System: fao, Name: Pacific, Southwest" ],

Can we include in the wiki to have information about the section, the expected label, if they are mandatory etc?

Luca Frosini wrote in #note-10:

"referring_stock_record" : "https://data.d4science.org/ctlg/GRSF_Admin/ce69cabb-d2bf-30a6-bb79-08f4ce857de8",
"referring_fishery_record" : "https://data.d4science.org/ctlg/GRSF_Admin/38fd8998-916a-3bac-b44c-b478e8223315"

I have included the catalog IDs, since I thought they were mandatory (i.e. for creating a link in the catalogue). If they are not needed, then I agree with you, let's remove them and keep only the actual URLs

Dear Yannis, thanks, I spoke with Luca today for a few clarifications, and please I have two questions.

1- GRSF semantic name: de facto that is the GRSF Traceability Unit Name , being the labels of all codes of the semantic identifier. So, the question is: do we need such semantic name or is it redundant?

2- Referring stock and fishery records: should we add in the json also the names of such stock and fishery or would it be redundant?

Thanks in advance for your feedback.

Aureliano Gentile wrote in #note-14:

Dear Yannis, thanks, I spoke with Luca today for a few clarifications, and please I have two questions.

1- GRSF semantic name: de facto that is the GRSF Traceability Unit Name , being the labels of all codes of the semantic identifier. So, the question is: do we need such semantic name or is it redundant?

You are right. In fact, as you might have seen these fields ( traceability_unit_name and grsf_semantic_name ) contain exactly the same contents. Unless @luca.frosini@isti.cnr.it wants them both present, I would remove one of them.

Aureliano Gentile wrote in #note-14:

2- Referring stock and fishery records: should we add in the json also the names of such stock and fishery or would it be redundant?

The referring records will be included as resources in the record, which means that they will have a URL pointing to the actual record. Since their names can be quite long (especially for fisheries) I do not know if they can be represented nicely in the catalogue. But if you would like to add them, I'm perfectly fine with it.

Thanks, so for question 1 (semantic name, let's see Luca's answer, in case the semantic name should be the one to be removed for consistency with stocks and fisheries), and for question 2, ok we do not put.

(By the way, I noticed the competency query for TU does not contain the traceability unit name, but contains the stock and fishery names).

@aureliano.gentile@fao.org we have just discussed this with @luca.frosini@isti.cnr.it

We could add a name for the referring stock and fishery records. To avoid the long names we could use either the semantic ID of those records or the UUIDs.

@aureliano.gentile@fao.org what do you think ?

yes indeed, the semantic ID might suffice, (and to see all metadata then we will click on the uuid). So fine as per your suggestion, thanks a lot.

Aureliano Gentile wrote in #note-18:

yes indeed, the semantic ID might suffice, (and to see all metadata then we will click on the uuid). So fine as per your suggestion, thanks a lot.

@marketak@ics.forth.gr please remember to attach to this ticket the updated JSON.

Yannis Marketakis wrote in #note-20:

Thanks @aureliano.gentile@fao.org for your reply. Let's use the semantic IDs then.
@luca.frosini@isti.cnr.it attached you will find the latest version (v4) of the JSON representation which contains updates in the fields

• referring_stock_record
• referring_fishery_record
• database_sources

Let me know if is OK

@marketak@ics.forth.gr in the last example you provided (v4) I don't find the field grsf_type

For the moment, I added it to my downloaded version as following:

"grsf_type": "traceability unit"

@marketak@ics.forth.gr In the example provided, there is still the following field:

"grsf_semantic_name": "Hyperoglyphe antarctica - 57 - 81 - Australian Fisheries Management Authority - 81 - Australia - Set longlines",

Do we have to remove it?

Moreover, there is s field description which is not documented in the wiki.
Do I have to consider it as simple Metadata of the Record? Which section?

"description": "Bluenose warehou | Southeastern Australia | Ocean Trap and Line Fishery (NSW) | Australia | Set longlines - Bluenose warehou - Southeastern Australia",

Of course, as with all records in GRSF, there are too:

"grsf_uuid": "f2b97baf-4212-337e-9d88-ec5cc59376ab",
"citation": "© D4Science.org. Hyperoglyphe antarctica - 57 - 81 - Australian Fisheries Management Authority - 81 - Australia - Set longlines. In: Global Record of Stocks and Fisheries (GRSF). Consiglio Nazionale delle Ricerche, CNR. Version: 1.0. Updated: February 20, 2023. Accessed: [Date accessed and/or downloaded]. http://data.d4science.org/ctlg/GRSF_Admin/f2b97baf-4212-337e-9d88-ec5cc59376ab. Database sources: [FishSource, FAO SDG14.4.1 Questionnaire]",

@aureliano.gentile@fao.org Can you also include them in the documentation?

Hi @luca.frosini@isti.cnr.it

• grsf_type: we were using that for stocks and fisheries to define the detailed type of each record (assessment unit or marine resource for stocks, and fishing unit or other fishery for fisheries). Traceability units are not sub-typed therefore I do not think we need it.
• grsf_semantic_name: I agree we can remove it (the same piece of information exists in the field traceability_unit_name
• description: Currently it is always the same as with the short_name (this is what we do for stocks and fisheries). If we do not need it (because it is the same with the short_name we can remove it)
• grsf_uuid: We need it
• citation: We need it

Good day Luca, regarding the #note-22, as long these are single examples, yes they can be added in the documentation as examples.

Regarding the #note-23, I concur with Yannis, if description is the same as short name, then no need to maintain it.