[foaf-dev] mapping of foaf:knows to contact vs acquaintance (was Re: Announcing the new 'testparse' method)

[Dan Brickley]

(changing subject line as this is a vocab mapping issue, albeit  =

triggered by the new tool)
+cc: foaf-dev

On 27 Mar 2008, at 15:13, Brad Fitzpatrick wrote:

> The Social Graph API now includes a way to parse a page on-the-fly  =

> for testing:
>
> http://code.google.com/apis/socialgraph/docs/testparse.html
>
> The parser that's running there is the same as the one that's been  =

> parsing the web regularly, but it's slightly newer.  The FOAF parser  =

> was recently rewritten, and the XFN parser got a bunch of love as  =

> well.  The index will be reparsed with the new parsers soon.

 From a quick look, the FOAF parser looks much improved, nice work :)

Having this tool really helps with transparency too, which is great  =

when people are occasionally puzzled by counter-intuitive results. I  =

have lately been investigating ways of having some FOAF idioms that  =

are closer to the document-url model used by SGAPI and  =

Microformats.org; but that's for a separate post.

Some notes here just on the vocabulary mapping aspect. Basically I  =

reckon it should be emitting 'acquaintance' rather than 'contact' if  =

you're using XFN definitions.

Reading http://gmpg.org/xfn/11
[[
Friendship (pick at most one)

contact
     Someone you know how to get in touch with. Often symmetric.
acquaintance
     Someone who you have exchanged greetings and not much (if any)  =

more =97 maybe a short conversation or two. Often symmetric.
friend
     Someone you are a friend to. A compatriot, buddy, home(boy|girl)  =

that you know. Often symmetric.
]]

It seems you map foaf:knows to "contact". If this notion of "contact"  =

is supposed to be xfn:contact, I suggest xfn:acquaintance would be a  =

more faithful mapping of foaf:knows as we define it, ie.
in http://xmlns.com/foaf/spec/#term_knows
[[
knows - A person known by this person (indicating some level of  =

reciprocated interaction between the parties).
]]


Reading http://microformats.org/wiki/xfn-clarifications#friends_acquaintanc=
es_and_contacts =

  I see [[
is contact a better lowest common denominator

Q: Is rel=3D"contact" a better lowest common denominator than  =

rel=3D"acquaintance" for services that only offer just one level of  =

friending (AKA "are you my friend yes or no")?

A: The rel=3D"contact" XFN relation is the lowest level of the  =

"friendship" axis in XFN, but the semantic both as expressed by their  =

user interfaces, and as implied by users and their usage patterns of  =

social network services is closer to 'acquaintance' than 'contact'.
]]

This (in addition to the definition) suggests that xfn:acquaintance is  =

by far the closest fit to foaf:knows, both as specified and as  =

deployed. I wish I found 'acquaintance' easier to spell :)
The English word 'contact' is a reasonable approximation to foaf:knows  =

as well, but the meaning of 'contact' in XFN is pretty specific and  =

substantially weaker than knows/acquaintance.

Probably not a huge deal but I figure it's worth nailing now, before  =

too many downstream apps appear...

cheers

Dan

ps. XFN says, "Friendship (pick at most one)"; it is worth pointing  =

out that aggregating multiple descriptions of a friendship, you might  =

get a mix of 'contact', 'acquaintance' and 'friend' descriptions. This  =

seems to me perfectly fine, and we should read the 'at most one' as a  =

rule about what happens within a single link, or single document,  =

rather than an observation about the wider world.    .... ie. last.fm  =

might only know we're contacts; flickr might know we're friends; and  =

this is just fine.


--
http://danbri.org/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.usefulinc.com/pipermail/foaf-dev/attachments/20080327/2eb=
6f769/attachment.htm