entity global identifier@UNDERSTANDING DATA
Alex Shkotin
Hi Kingsley,
One important topic of your presentation [1] is about entity global identifiers (EGI). You described different ways to get this kind of identifier. And there are additional topics about their usage.
Our experience reported here [2] may be of interest.
The key idea is material procedures to get EGI from the entity in the reality, and to find the entity in reality by EGI.
Alex
[1] https://www.openlinksw.com/data/pdf/Understanding%20Data%20--%20PDF%20Latest.pdf
[2] https://www.researchgate.net/publication/282944880_Finite_Systems_Handling_Language_YAFOLL_message_1
a newer version is here Finite Systems Handling Language.v.1.2 (-:PUBLIC:-) opened to comments.
This is the paragraph of interest
Populating sorts and basic attributes(NEW)
The rock samples processed and stored at a laboratory can be identified using various methods. The Ide type identifier assigned to the sample can be arbitrary within a FMAS, provided that three requirements are satisfied:
(GUI) global unique identification: the identifiers of different samples must be different for all samples in all laboratories in the world.
(I2SW) id to sample works: the sample can be found by the identifier and FMAS in the particular laboratory.
(S2IW) sample to id works: sample identifier in FMAS can be obtained processing the sample at the laboratory.
A unique number of the record for sample was taken from the Proba database and updated to be Ide to execute GUI. So, DB record number 32994 yields an Ide = _SAM32994.
References to reality works this way: basic attributes of the element of each sort enable finding the sample in reality by their values to support I2SW.
For example, let's introduce the function
Declaration authorial_number sample : S prime . --mapping sample representative to the string.
The authorial_number function supports the author's sample number, which is unique for the laboratory that stores the sample.
This function must be full, i.e. we have an axiom:
Axiom AN_full ( ∀ x : sample ( ∃ y : S ( authorial_number ( x ) = y ) ) ) .
A material algorithm is assumed to exist: how a sample can be found in the laboratory if its author's number and other basic attributes are known.
A material algorithm that is 'reverse' in a certain sense is assumed to support S2IW, i.e. to create sample representative (Ide) while being one-on-ones with the sample in the laboratory: values of the basic attributes of the sample unambiguously characterizing it in FMAS are located somewhere in the laboratory. This unambiguity is supported by the axiom of uniqueness of the totality of basic attributes.
Suppose for simplicity that the authorial_number is globally (for all laboratories) unique. Then the axiom of uniqueness will look as follows:
Axiom AN_uni (∀x:sample (∀y:sample (authorial_number(x)=authorial_number(y)) → (x=y))).
A set of basic attributes, material algorithms, and axioms of completeness and uniqueness of their basic attributes similarly exists for publication sort and place sort.
In addition, the following formalization method is used for geographical names (for example, Iceland, Atlantic Ocean): they are used to form an Ide (_Iceland, _Atlantic_Ocean) and to assign it to the place sort.
! _Iceland place !
! _Atlantic_Ocean place !