Existing ontology (metadata) to describe Business Transactions triples.

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

Existing ontology (metadata) to describe Business Transactions triples.

cracoras
Hi there,

I am working on a project that involves business transactions. My intention is to have an ontology with metadata in a triple format (Subject->Predicate->Object) will describe relationships between MetaData (MD) and Transactions (TR) objects. Like for instance:

Invoice X (TR) [is_shipped_from] Warehouse Y [MD].

My original metadata in XML (based on OAGIS - https://schemas.liquid-technologies.com/OAGIS/8/) describes all possible relationships and I was able to compile .yaml mapping file with all of them and have most of them defaulted to a "has_a".

Invoice:
  /*/DataArea/Invoice/InvoiceHeader/BillFromParty/Contact/ID:
    description: ''
    noun_object: ContactMaster
    predicate: has_a
  /*/DataArea/Invoice/InvoiceHeader/BillFromParty/Location/ID:
    description: Check with Jack
    noun_object: Location
    predicate: has_a
  /*/DataArea/Invoice/InvoiceHeader/BillToParty/BuyerContact/ID:
    description: Is this needed at invoice level?
    noun_object: ContactMaster
    predicate: has_a
  /*/DataArea/Invoice/InvoiceHeader/BillToParty/Contact/ID:
    description: Bill to Party contact on the invoice
    noun_object: ContactMaster
    predicate: has_a
  /*/DataArea/Invoice/InvoiceHeader/BillToParty/Location/LogisticsLocation/ID:
    description: Check with Jack
    noun_object: Location
    predicate: has_a
  /*/DataArea/Invoice/InvoiceHeader/BillToParty/PartyIDs/ID:
    description: Bill to Party on the Invoice
    noun_object: BillToPartyMaster
    predicate: has_a


The description is there, it could help me infer some of them manually, but is not very consistent, some are missing and Some of them even suggest to check with "Jack" the person who created the original metadata.  
To make it more challenging some of these predicate/relationship meanings may vary depending on the domain. For example, a transaction between two entities in a manufacturing company will have slight different terms than the same transaction in a healthcare business, etc.

I am trying to avoid to create these manually as there are at least over a 1,000 of those. I understand from my initial readings about ontologies that I could link my ontology to another ones that describes this domain.  

My questions are:

1) Am I right on assuming that I could leverage existing ontologies for the triple definitions ?
2) Is anyone aware of an open data set where I could obtain this ?

Thank you in advance for any information.

Regards,

--MD.