“I express my network in a FOAF file and that is the start of a Revolution” said TBL here http://dig.csail.mit.edu/breadcrumbs/node/215 - which I want to happen.

So in an effort to make myself machine accessible I expressed my profile here http://www.realtea.net/Xpose_Me_Lightly - in an experiment to publish better data about myself in the hope that I can help marketers tune their messages to me - part of my projectVRM activity.

So when I create a FOAF file for myself, I want to be able to distinguish between my business contacts, personal friends, and online acquaintances. I consider the followers on Twitter online acquaintances - many of whom I have never met in person or even interacted with.

So reviewing the FOAF file generated by SemanticTweet, I felt I needed to better qualify the class.

Maybe it needs a subclass of that expresses what type of relationship I have - the comment from the FOAF site says:

“To provide additional levels of representation beyond mere ‘knows’, FOAF applications can do several things.

They can use more precise relationships than foaf:knows to relate people to people. The original FOAF design included two of these (‘knowsWell’,'friend’) which we removed because they were somewhat awkward to actually use, bringing an inappopriate air of precision to an intrinsically vague concept. Other extensions have been proposed, including Eric Vitiello’s Relationship module for FOAF.

In addition to using more specialised inter-personal relationship types (eg rel:acquaintanceOf etc) it is often just as good to use RDF descriptions of the states of affairs which imply particular kinds of relationship. So for example, two people who have the same value for their foaf:workplaceHomepage property are typically colleagues. We don’t (currently) clutter FOAF up with these extra relationships, but the facts can be written in FOAF nevertheless. Similarly, if there exists a foaf:Document that has two people listed as its foaf:makers, then they are probably collaborators of some kind. Or if two people appear in 100s of digital photos together, there’s a good chance they’re friends and/or colleagues.

So FOAF is quite pluralistic in its approach to representing relationships between people. FOAF is built on top of a general purpose machine language for representing relationships (ie. RDF), so is quite capable of representing any kinds of relationship we care to add. The problems are generally social rather than technical; deciding on appropriate ways of describing these interconnections is a subtle art.”

(http://xmlns.com/foaf/spec/#term_knows)

I followed the Eric Vitiello link - that provided this:

Abstract

relationship is a module for extending the usefullness of the foaf:knows element. This is accomplished by aliasing the foaf:knows element into elements that describe the relationship between people in more detail.

The RDF Schema properties

* rel:friendOf
* rel:acquaintanceOf
* rel:parentOf
* rel:siblingOf
* rel:childOf
* rel:grandchildOf
* rel:spouseOf
* rel:enemyOf
* rel:antagonistOf
* rel:ambivalentOf

(http://www.perceive.net/schemas/20021119/relationship/)

So the Twitter friends may be acquaintanceOf. What do you think?

Paul,

thanks for the feedback and suggestions.

To be honest, I don’t expect to be doing reminders of the kind that you propose. I’m afraid that this looks like too much work for what is simply a small project that I run on my own time for the benefit of anybody who might get use from it.

From my point of view, the real benefit of a service like this is that it can be linked to directly, and doesn’t have to be cut and pasted into some larger document.

Of course, this raises the issue that you have correctly identified - namely that simply because you follow somebody on Twitter doesn’t mean that you ‘know’ them.

One could imagine how control could be made more granular. One suggestion that I received was to make a method that showed the intersection of your followers and followees (if that’s a word), which might be a greater indication that you ‘know’ them. Or you could go deeper again, as you have suggested, and looks for signs of interaction through DMs, @messages, and the like.

Ultimately, these are all likely to be unsuccessful, my idea of a friend, as codified in SemanticTweet, or any FOAF generating web service, is likely to be different to yours or anybody elses. The FOAF spec itself is deliberatly vague on the use of the foaf:knows predicate. To quote:

“A foaf:knows relationship does not imply friendship, endorsement, or that a face-to-face meeting has taken place: phone, fax, email, and smoke signals are all perfectly acceptable ways of communicating with people you know.”

Anyway, hopefully you find that SemanticTweet to be a somewhat useful and interesting tool, and can exploit and extend it in interesting ways.