Term request: email address + phone number, postal address

[zhengj2007]

From alanruttenberg@gmail.com on January 03, 2012 19:23:03

From: https://groups.google.com/group/information-ontology/browse_thread/thread/cb03fa62269dd995 Subject: [IAO] email address

From: Chris Stoeckert stoeckrt@pcbi.upenn.edu Date: Tue, Jan 3, 2012 at 6:15 PM To: IAO Discuss information-ontology@googlegroups.com

Hi (and Happy New Year) I have multiple use cases that require specification on an email address. 'postal address' (IAO_0000422) is actually quite close (defined as: A textual entity that is used as directive to deliver something to a person, or organization) however it is logically defined as 'is about' some 'geographical location' which does not fit. I propose: 1) adjusting the textual definition of postal address to reflect its dependency on geographical location. 2) addition of the term 'email address' with a definition such as " A textual entity that is used as directive to deliver something electronically to a person, or organization." For differentia from 'postal address', we need to indicate what 'email address' is about but I don't think we have something appropriate for an electronic account (not really a device).

apologies if this has been discussed before but hard to search for in the discussion list. Thanks Chris

---

From: Alan Ruttenberg alanruttenberg@gmail.com Date: Tue, Jan 3, 2012 at 7:20 PM To: Chris Stoeckert stoeckrt@pcbi.upenn.edu Cc: IAO Discuss information-ontology@googlegroups.com

Clearly a mixed message here. The reason that the textual definition doesn't mention geographical location is that there are postal addresses that are not geographically fixed, for example addresses on warships. I would say it is loosely about geographical location in that eventually it is a material entity that is delivered somewhere specifically located (to a site, perhaps). So the text should be adjusted, but with a mind to not confuse postal address with names for geographic location such as longitude and latitude coordinates. We probably should include something about email addresses used as intended, and for other uses. Legitimate uses are as identifiers for persons and organizations, such as when used as a login, and to record provenance, as when used in the "From:" field. Illegitimate uses are, for example, spoofing From: addresses by spammers. I think that the definition of email address should make reference to the protocols that define delivery, something along the lines of delivering information to a mailbox by certain protocols. The protocols are well documented on the wikipedia page: http://en.wikipedia.org/wiki/Email_address In addition, since there is a standard for email address URIs, these should be used for instances. I.e. the URI for my gmail box would be: mailto:alanruttenberg@gmail.com and we would then have assertions

< mailto:alanruttenberg@gmail.com > instanceOf(i.e. rdf:type) <iao:email address> (whatever the id is)

etc. Protege lets you explicitly set the URI of a term. I'm going to add an issue once I send this email, and copy the conversation (or leave pointers) there. If you agree with the strategy and want to have a go at a definition, please add it to the issue thread?

Original issue: http://code.google.com/p/information-artifact-ontology/issues/detail?id=130

[zhengj2007]

From alanruttenberg@gmail.com on January 03, 2012 16:46:55

Added provision ID http://purl.obolibrary.org/obo/IAO_0000429 and editor note pointing to this thread.

[zhengj2007]

From stoec...@upenn.edu on January 03, 2012 18:59:03

I had not realized that there is such a mechanism as creating addresses with '+' on the fly. Given that, I agree with you completely.

• Bjoern

On Tue, Jan 3, 2012 at 7:56 PM, Bjoern Peters bpeters@liai.org wrote:

I thought email addresses are 'staggered CRIDs', with DNS servers registering each domain (e.g. gmail.com), and in each domain there is a second CRID mechanism that deals with the part before the @.

There is a mechanism, but it is not necessarily CRID based - the gmail addresses that you can make up on the fly by adding + being an example. Or do you not agree that those are counterexample?

I would consider 'made up addresses' to be outside of the scope. Made up postal addresses should also be, otherwise what 'geographic location' would they be about?

For mail addresses I agree. For email addresses I'm not so sure. If they are defined in terms of how they are manipulated by protocols, then email addresses with valid domains but invalid parts before the "@" are an in-between case. The protocol for delivering them (SMTP) defines the behavior of mail servers even when the mail arrives and there is no mailbox. Various actions can be configured such as responding with an error message, forwarding, or specialized handling such as in the case of the "+" addresses.

I've taken the CRID case to mean there is a single database of some kind that has an entry for each valid identifier.

There's no question that email addresses have a somewhat organized way to be created, but my gut would be define another kind of identifier creation process that matches how email addresses (and similar identifiers) work.

-Alan

I'm having some doubts about the "about geographic location" axiom on postal addresses. Since about is in the eye of the creator and because creators would not know geographic location in some cases (the ships) it seems like it can't be universal.

• Bjoern

---

On Tue, Jan 3, 2012 at 7:23 PM, Bjoern Peters bpeters@liai.org wrote:

(and I would argue more specifically CRIDs, but that is not completely true for postal addresses)

Interesting question about CRIDs. Some points to consider:

• There isn't any single registry for email addresses. They are allocated by providers, people who run smtp servers, and users that are enabled to add email aliases.
• Consider google email addresses, e.g. alanruttenberg+@gmail.com. Anyone can make an email address with arbitrary "+" additions and they will all be delivered to the email address without the "+". There is no registration necessary for these addresses.
• We need to decide whether email addresses that are made up (e.g. by spammers), but syntactically correct, are IAO email addresses. This adds a further wrinkle.

-Alan

ps. I added provisional id http://purl.obolibrary.org/obo/IAO_0000429 > so you can use it in OBI pending getting the definition right.

-- information-ontology@googlegroups.com To change settings, visit http://groups.google.com/group/information-ontology

[zhengj2007]

From stoec...@upenn.edu on January 04, 2012 13:29:16

From: Bill Hogan hoganwr@gmail.com Date: Wed, Jan 4, 2012 at 12:49 PM Subject: Re: [IAO] email address To: Alan Ruttenberg alanruttenberg@gmail.com

Just to emphasize Alan's point, there is a huge difference between street address and mailing address for a significant number of individuals living in rural areas. See here: http://painintheenglish.com/case/3604 Bill

[zhengj2007]

From stoec...@upenn.edu on January 04, 2012 13:52:57

I agree with Alan about inclusion of protocol in the definition of email address as it obviates the issue of spoofing since SMTP servers don't use header addresses according to the wikipedia link Alan provided.
Proposed definition: A textual entity that is used by Simple Mail Transfer Protocol (SMTP) servers as directive to deliver something to an electronic mailbox. It is about some computer which is used to provide access to the email.

[zhengj2007]

From stoec...@upenn.edu on August 13, 2012 10:02:18

At ICBO, I discussed phone number with Alan Ruttenberg along with email address and postal address as these seem to all be of the same type in some way. They share a communication objective but have different modalities (phone~voice, email~digital, postal address~physical). They are also about some addressable device or devices associated with a person or organization (phone/ voice mail, email server, mailbox). Are these all "A textual entity that is used as directive to deliver something to a person, or organization" (postal address definition) but delivering different things (voice, email, letters and packages)?

Summary: Term request: email address + phone number, postal address
Cc: MBrochhausen@gmail.com