Property talk:P5587

From Wikidata
Jump to navigation Jump to search

Documentation

Libris-URI
identifier for an item in the catalogue of the Swedish National Library
RepresentsLIBRIS (Q1798125)
Associated item
Data typeExternal identifier
Domainhuman (Q5), version, edition, or translation (Q3331189), work (Q386724) and fictional character (Q95074)
ExampleAugust Strindberg (Q7724)tr574vdc33gk2cc (RDF)
Formatter URLhttps://libris.kb.se/katalogisering/$1
Related to countryFlag of Sweden.svg Sweden (Q34) (See 57 others)
See alsoSELIBR ID (P906), LIBRIS editions (P1182)
Lists
Proposal discussionProposal discussion
Current uses60,770
Search for values
[create] Create a translatable help page (preferably in English) for this property to be included here
Single value: this property generally contains a single value. (Help)
Exceptions are possible as rare values may exist.
List of this constraint violations: Database reports/Constraint violations/P5587#Single value, SPARQL, SPARQL (new)
Distinct values: this property likely contains a value that is different from all other items. (Help)
Exceptions are possible as rare values may exist.
List of this constraint violations: Database reports/Constraint violations/P5587#Unique value, SPARQL (every item), SPARQL (by value), SPARQL (new)

Not only human constraint[edit]

The point of the new URI is that it's supposed to be used for both authorities and books (replacing both LIBRIS editions (P1182) and SELIBR ID (P906)). So the constraint/domain should be expanded – both with instance of (P31)version, edition, or translation (Q3331189) (edition-level items, such as zg86cd0921xzp3q), and also instance of (P31)book (Q571) (work-level items, even though they haven't been fully implemented in Libris yet but they exist 53hlp2kp2xqdq5g). Ping @Salgo60, André Costa (WMSE): --Alicia Fagerving (WMSE) (talk) 06:52, 7 September 2018 (UTC)

On second thought, obviously instance of (P31)book (Q571) for work-level items is too limiting because all sorts of materials can be included in a library catalog! --Alicia Fagerving (WMSE) (talk) 08:29, 7 September 2018 (UTC)
@Alicia Fagerving (WMSE): yes agree its confusing but I hope LIBRIS will be the hub of a lot of Linked data in Sweden. What about a subject like we have a book about a church parish or a city? I miss some vision what a Linked Library system can do see todays tweet
  • I guess we will have LIBRIS-URI for subjects also? But this is just a guess I feel status with LIBRIS XL is rather confusing
On page about we can read in swedish
Varje begrepp, som t.ex. ett ämnesord, identifieras med hjälp av en HTTP-URI. URI:n består av en adress och en textsträng som tillsammans utgör begreppets unika identifikator. Exempel: https://libris.kb.se/86lnrdxs0h826ss.
Du som använder begreppen i egna system bör lagra de kompletta strängarna (https://...) i din databas.
Vissa begrepp får även en mer "läsbar" identifikator baserat på etiketten för begreppet. Exemplet ovan, som identifierar ämnesordet "Skidåkning", har fått den alternativa identifikatorn: https://id.kb.se/term/barn/Skid%C3%A5kning. Men du bör använda formen https://libris.kb.se/86lnrdxs0h826ss när du lagrar den lokalt i ditt system.
Looks like the links dont work ;-(
- Salgo60 (talk) 08:41, 7 September 2018 (UTC)
The links work if you use katalogisering as prefix, e.g. https://libris.kb.se/katalogisering/86lnrdxs0h826ss for your example. So yeah, concepts are also included. Oh, and so are events and organizations. More and more confusing :| --Alicia Fagerving (WMSE) (talk) 09:41, 7 September 2018 (UTC)
I error reported it see fel-url-i-om-idkbse - Salgo60 (talk) 16:40, 7 September 2018 (UTC)
I feel a fully working Linked Library system will have all concepts we have in Wikidata and in other systems. The major problem I see coming is to match different "concepts" in a useful way and to do that I feel we need to communicate A LOT... and as I have learned in project wmse-riksarkivet-tora is that most institutions as LIBRIS, RAÄ , Riksarkivet, Lantmäteriverket, SCB, SOFI has no working changemanagement like we have in Wikidata with Phabricator ==> it will be a mess to track a change in one system that is dependent on a change in another system see my <flame on>tweet<frame off> ;-)
My feeling is that people inside Wikidata are highly skilled, we have Phabricator ==> we have a controlled change process so maybe we can do some skill transfer...
In an earlier post at LIBRIS I tried to explain how good Phabricator works for Wikidata see post and when it works for Facebook who has continous delivery and +500 releases per day link rapid release at massive scale I guess it will be enough for Libris...or RAÅ or Lantmäteriverket... - Salgo60 (talk) 11:48, 7 September 2018 (UTC)
I missed this topic but went ahead and extended the constraint with a few more types. I agree that organisations, events and probably other types need to be added but it's probably easiest to investigate which as we come across them. /André Costa (WMSE) (talk) 20:37, 22 October 2018 (UTC)

Example of the Linked data challenge[edit]

I am cleaning parish of the Church of Sweden (Q615980) see salgo60.github.io/Svenskaforsamlingar

For a church parish we add a picture of the church see example Q10711614#P18 were we add

When I started to search I found 2 objects

So what what we "normal" Wikipedia people call a church is when we connects with the specialists at RAÄ two concepts ==> in the WIkipedia article we describe the church and in Wikidata we will have 2 objects relating to 2 concepts at RAÄ and how will LIBRIS solve this? Paintings from the National museum gets in Wikidata that this painting depicts (P180) ==> if the National museums starts also to deliver Linked data with depicts they need to coordinate with Libris/Wikidata/RAÄ..... example from a blogpost SPARQL search in Wikidata finding people buried at cemetery Adolf Fredrik who are depicted on paintings/sculptures link


LIBRIS Utvängstorp[edit]

If we check i LIBRIS they have

  • book "Midnattsmässa i Utvängstorp 2017" that I guess is a recording in the church
  • book "Utvängstorps kyrka" is about the church
    • today in MARC format we have
      • 084 a Ic-cz Utvängstorp: Utvängstorps kyrka2 kssb/7
      • 245 1 0 a Utvängstorps kyrka /c [Rune Lindgren]
      • 610 2 4 a Utvängstorps kyrka
MARC format[edit]

A check in the MARC format book

  • 084 - Other Classification Number (R)]
  • 245 - Title Statement (NR)
  • 610 - Subject Added Entry-Corporate Name (R)

LIBRIS XL[edit]

How is it today mapped

==>

    • "hasTitle": "mainTitle": "Utvängstorps kyrka"
    • "classification":
      • "code": "Ic-cz Utvängstorp: Utvängstorps kyrka",
      • "inScheme": "sameAs": "@id": "https://id.kb.se/term/kssb%2F7/"
      • "subject": "@type": "Organization", "name": "Utvängstorps kyrka"

==> we have no links to an external concept

Lesson learned so far and what is needed[edit]

What I guess is needed[edit]

A) we need to document how a "concept" like church is defined in Wikipedia, Libris, National Museum.... Both in Swedish but also in English
B) we need in Wikidata try to describe/document what we thought when we imported/mapped it so the next person get an easy understanding. After speaking to Andre Costa the decision was to import "Anläggningspresentation" and some "Byggnadspresentation" 
C) we need to find a "standard" process inside Wikidata how to flag/track/report problems in a consistent way with external data providers see EPIC T202530  Feedback processes and tools for data-providers  
D) best would be if RAÄ, LIBRIS, Digisam?, en:National_Archives_of_Sweden starts to show what vision they have and set up some kind of shared workspaces for linked data in Sweden were we easy could
D-1) discuss concepts
D-2) document like Nobel data did see picture of db modell in Task T200668
D-3) maybe have a common platform to track changes or a possibility to forward helpdesk tasks between systems "upstreams" ==>
D-3-1) unique helpdesk numbers that everyone can use to communicate
D-3-1-1) one place to track the status of a task and who is in charge of the task and dependencies compare the usage of Phabricator for project Wikidata and wmse-riksarkivet-tora. 
D-3-2) a platform for sending in Change requests
D-3-3) a platform to discuss how we set up e.g. federated search between all linked data systems - compare Wikidata <-> Nobel Data task T200668

How to make Linked data possible[edit]

I suggest that we start to define the process maturity for handling en:linked data. I guess it will be the same as CMM Capability Maturity Model see video

As documented above getting no helpdesk tickets or having no process for asking Change requests ==> we are on the initial level were we hope finding "heroes" in the organisation that will help us. When we agree that we have a problem and a common vision then we can take step 1 together is my experience doing things like this before in organisations with a requirement to deliver a good result.....

My personal believe with such poor change management shown above people will not trust you and will not invest in time/money to do solutions with your data see tweet

Characteristics of Capability Maturity Model

- Salgo60 (talk) 06:27, 8 September 2018 (UTC)

LIBRIS XL roadmap[edit]

- Salgo60 (talk) 10:56, 21 September 2018 (UTC)

Formatting URL what is good or bad[edit]

@Alicia_Fagerving_(WMSE): what is the user case you address with your change 770027612 https://libris.kb.se/katalogisering/$1 ?

As LIBRIS has no Landing page I prefer an URL like https://libris.kb.se/katalogisering/search/libris?q=$1

My arguments:

  1. With the above URL you get the "agent" as result number 1 ==> it makes sense for the user PLUS you get the related instances... see my comment lanka-librisuri/#c3083566
  2. We create solutions using the formatting URL in external tools ==> it should be useful for other tools
    1. Tool wmflabs HUB use the formatter URL ==> from an user perspective getting to an entry with no links adds no value
    2. Example how Swedish Church parishes use this tool from Family Search T200515#4683374
  3. LIBRIS are aware that have no "landingpages" maybe will fix it see kundo.se

Related properties to use

Related questions @Alicia_Fagerving_(WMSE)::

  1. is there any plans that LIBRIS XL will update its data with same as Wikidata/Wikipedia?
  2. status of LIBRIS Federated search see question
    1. Wikidata <-> Nobelprize.org is done and useful see SPARQL in action: Comparing 2 datasets using federated SPARQL

- Salgo60 (talk) 07:19, 22 October 2018 (UTC) phone: ++46705937579

Linking directly to the person's entry has the advantage of being able to see all the biographical information at a glance. If you link to the search page, you have to perform an extra click in order to get to the information that you expected to see in the first place. --Alicia Fagerving (WMSE) (talk) 07:31, 22 October 2018 (UTC)
Any status / activity with federated search? - Salgo60 (talk) 08:09, 22 October 2018 (UTC)
my question was what do you expect to see?
  1. to confirm it is the same as person?
  2. to learn something new about the same as person?
  3. to understand what LIBRISXL knows that Wikidata dont know?
As you are in Wikidata Osamu Shimomura (Q205345) you already know much more than LIBRISXL knows and presents on the page jgvz3wj23jlqq08
  1. LIBRIS XL tells me he won the Nobel prize 2008 but no easy way finding the books
  2. search jgvz3wj23jlqq08
    1. shows related items #LIBRISXL has about him i.e. Wikidata knows he is same as and LIBRIS tells us a little bit more
I feel #1 adds no value and its impossible for a "normal" user to find the related books with todays LIBRISXL design. Today they link auth/319922 which is even less user friendly than FAUT:319922
I asked LIBRIS XL if they have any user cases available so we understand why its designed as it is see question
For me presenting an author and the related books must be one of the main purpose of a Library system and it feels odd they design a system that dont give the users an easy way to get added value - Salgo60 (talk) 08:09, 22 October 2018 (UTC)
This is not the place to discuss the development of Libris. Please keep in mind that the Libris XL interface is still under development. The things you are pointing out will probably change a lot before Libris XL becomes the default interface. --Alicia Fagerving (WMSE) (talk) 08:14, 22 October 2018 (UTC)
yes I also hope that but I get afraid we will sit with yesterdays weather for the next year and therefor I have problem understand your change to something I think is less good and not at all user friendly... call me if you think you will discuss this and give an update of the LIBRIS activities ++46705937579 I think Wikipedia and LIBRIS XL can do a lot interesting see tweets #LIBRISXL - Salgo60 (talk) 08:45, 22 October 2018 (UTC)
I've always interpreted formatter URL (P1630) as being the page where the data provider expects (human) readers to end up if they look up the identifier. For a Wikidataitem that would be https://www.wikidata.org/$1 for Libris-URI. As such it should be https://libris.kb.se/katalogisering/$1. Linking to the search would be the equivalent to, for Wikidata, linking to the results of a sparqll query listing everything that has a relation to the item - sure it gives you more info but its not what the user requested or the provider intended. I can see a search link being used, but only when there are no landing pages at all, not the case for Libris-URI.
As you mention there are two other properties.
  • formatter URI for RDF resource (P1921) Should follow the KB spec, meaning it should be https://libris.kb.se/$1. The wmflabs HUB and other tools making used of linked data should primarilly make use of this property. If they don't the error lies with those tools, not with Wikidata. If the link doesn't give you rdf when you navigate there using a linked data content negotiator (i.e. not your browser) then that is a separate thing to raise with Libris.
  • third-party formatter URL (P3303) Here we are free to link to anything we feel gives a good view of the data (c.f.. Resonator for Wikidata). I have no opinion of how this one is formatted.
/André Costa (WMSE) (talk) 16:07, 22 October 2018 (UTC)
@André Costa (WMSE), Alicia Fagerving (WMSE):Now LIBRIS XL released 1.3 that at least have a klickable number see tweet - Salgo60 (talk) 07:30, 2 November 2018 (UTC)