[foaf-dev] Re: 'Interested In' field

Dan Brickley danbri at danbri.org
Tue Apr 8 07:29:06 BST 2008


+cc: foaf-dev (apologies to anyone trying to cross-post; signups are at 
http://lists.foaf-project.org/mailman/listinfo/foaf-dev and 
http://groups.google.com/group/opensocial-and-gadgets-spec and probably 
needed to x-post)

ernest wrote:
> As per a couple of conversations in the last hackathon at six apart,
> here is our proposal for the 'interested in' field to be included in
> the next version.
> Here is a little research of the main social networks and their use of
> the 'interested in/looking for' field in their user profiles.
> (http://spreadsheets.google.com/pub?key=pXn69uFFnb93d__8QKu77Pg)
> Of course, anyone can keep researching in this huge list of social
> networks. (http://en.wikipedia.org/wiki/
> List_of_social_networking_websites)
> A common structure for all of them seems to be the following:
>
> Static Class opensocial.Enum.InterestedIn
>  <static> object DATING
>  <static> object FRIENDS
>  <static> object RELATIONSHIP
>  <static> object NETWORKING
>  <static> object ACTIVITY_PARTNERS
>  <static> object RANDOM (?)
>
> An open question is whether or not to specify a target gender for each
> of them as some of the social apps do (see the spreadsheet)
>
> Now it's time for your feedback.
>   
It's really great to see this kind of analysis happening.

As for target gender, I think a separate (like everything, optional) 
field indicating somehow sexuality is the way to go. Of the 6 values you 
show, it is only really 'dating'  and 'relationship' where a target 
gender makes sense. Do people say that they're looking for 'male' or 
'female' people to network with, be friends with, etc?

BTW I've been thinking a little about how to add something to FOAF for 
this area. An idea is to have something close to the above structure, 
plus sexuality.  To capture a richer notion of the kind of people 
someone is interested in, also have an extra construct which relates a 
person to a 'Group' definition. The group would be associated with a 
SPARQL query, along the lines of the thinking in 
http://danbri.org/words/2008/01/22/260 ... this would be a akin to a 
stored search, or to using a stereotypical person-instance description 
as a template for matching (that's more or less how SPARQL itself 
works). I could store a query that defined the group of people whose 
bloodtype is A+ and who work for Microsoft, even. By having a mechanism 
like that with massive expressivity it might help take away the 
expectation that 'interested in' will cover everything. But that's for 
FOAF; the relationship to OpenSocial there is up for grabs.

Back to your list: some of the sites have willfully obscure labels, 
which lead users off in different directions. How comfortable are you in 
saying facebook's 'random play' == multiply's 'adult fun'? Are there any 
use studies around that cover how they're all being used in practice? 
and whether those interpretations are shifting?

The commentary in 
http://www.urbandictionary.com/define.php?term=whatever+i+can+get btw 
suggests 'whatever i can get' is more of a wildcard across the other 
values, ""It refers to an individual's desire for any type of 
relationship, specifically, "dating," "friendship," and/or "random play." ""

In the talk I gave in Cork recently on this stuff, 
http://danbri.org/words/2008/03/04/288 I mention a recent case of a 
(gay) police inspector who missed out on a promotion due to his Facebook 
profile. Newspaper coverage noted:

      "The inspector, from Bedford, said on the site that he was 
interested in men and looking for 'whatever I can get' "

see slides at http://www.slideshare.net/danbri/whatever-i-can-get/

Sometimes information is designed not to be taken out of context. Here 
the implied interpretation seems to be sexual, rather than the milder 
'I'm open to interesting events' wildcard reading. Within the context of 
a single site, such vague labels can have semi-stable meaning and 
interpretation. But these social sites are typically anchored in a few 
primary countries. Coming up with a cross-culturally applicable list is 
not going to be super-easy. It could look easy from a surface 
engineering level, but result in confusion. I think even between UK and 
US you may find different interpretations due to our British habit of 
finding sex-related double meaning wherever we can ;)  
(http://en.wikipedia.org/wiki/Double_entendre#British_comedy)

Despite that concern, nice work! I hope to see it evolve and perhaps 
folk can find some empirical / user study research to help stabilise the 
categories?

cheers,

Dan



More information about the foaf-dev mailing list