- From: Marijane White <whimar@ohsu.edu>
- Date: Thu, 18 May 2017 00:28:17 +0000
- To: "Maggie Pong" <mpong@microsoft.com>, "public-schemaorg@w3.org" <public-schemaorg@w3.org>
- Message-ID: <5036E256-EB20-4692-B6BD-F5DFBBF69071@ohsu.edu>
Hi Maggie, I have been using schema.org to model contact information for healthcare providers and clinics at Oregon Health and Science University, perhaps I can help. That said, I am not sure there is a best way to do this. Schema.org offers considerable flexibility in modeling contact information and the approach you choose may depend on your use case. For example, if the contacts you are modeling will be used in a queryable system, you might consider devising a list of queries that you or users would ask the system, as this can often reveal which approach is needed. For example, my own use case required identifying buildings that share the same street address and clinics that share the same phone number. This led me to reify pretty much every piece of contact information as its own ContactPoint or PostalAddress, related to Organization and CivicStructure entities with contactPoint and address as appropriate. I should probably also note that I am doing something that I understand is rare with schema.org – I am using a triplestore, and I am minting IDs for my entities, including the ContactPoints and PostalAddresses. This has made writing SPARQL queries against my knowledge base dead simple. But a different set of use cases could have just as easily led me to model them all as text strings. Could you perhaps explain what you mean by “trying to avoid capturing contact as a person or a business directly even though it may share many common properties”? I am not sure I understand. Do you mean your users have many contacts with the same contact information? Or do you mean you are trying to avoid placing properties like email and telephone directly on Person or Organization entities? I am also a bit confused by your remark about ContactPoint, because my own understanding is that contacts are particular entities and it’s not clear why you think it might not suit your use case. Best regards, Marijane White, MSLIS Ontologist Research Associate Ontology Development Group Oregon Health & Science University Library From: Maggie Pong <mpong@microsoft.com> Date: Tuesday, May 9, 2017 at 11:41 AM To: "public-schemaorg@w3.org" <public-schemaorg@w3.org> Subject: Modeling contact info in address book Resent-From: <public-schemaorg@w3.org> Resent-Date: Monday, May 15, 2017 at 12:39 AM I am looking into the best ways to model contacts that a user has in an address book. I see contact is a rarified relationship between two entities. Therefore, I am trying to avoid capturing contact as a person or a business directly even though it may share many common properties. I see the ContactPoint schema but that is more about contact information for a particular entity. Any recommendation is appreciated. Thanks. Maggie Sent from Mail<https://go.microsoft.com/fwlink/?LinkId=550986> for Windows 10
Received on Thursday, 18 May 2017 00:28:57 UTC