[LispSea] database of personal interests

[Brandon J. Van Every]

As I've talked, I've realized I have 10..15 different personal interests 
of relevance to LispSea.  And that's not even counting all the ones of a 
purely social value, such as sociocultural anthropology, world history, 
martial arts, film, etc.  A group such as LispSea is not just a 
technical organization, it is also a social one, so I don't see the 
listing of such interests as out of bounds.  SeaFunc meets over beer for 
a reason, after all.

It occurs to me, however, that my list of interests gets complicated.  
Let's say all the other active LispSea participants are equally 
complicated.  How can all these interests be made accessible?

One way is socialization.  People blab at each other, people find out 
about their common interests.  That's largely the format that SeaFunc 
has followed.  Although I would note, significantly, that I have not 
personally engaged in any social activities with anyone from SeaFunc, 
outside of SeaFunc.  Or MLSIG-Seattle before that.  The accident of 
personal chemistry?  Or geography?  Or available time?  Anyways, it says 
to me that the meme of socialization is not fully utilized.  On the 
other hand, I have a pretty good mental inventory of most of the 
regulars' technical interests.  So it's not under-utilized either.  I 
couldn't tell you much about people who just showed up a few times though.

So, socialization is not entirely sticky.  It requires the ongoing 
ritual of showing up to a meeting, and furthermore actively talking to 
people.  Not difficult at SeaFunc since that's all we do, but it will be 
more of a challenge for LispSea, since we're aiming at more lecture formats.

A bio page is the quick way to get information about 1 person.  It's 
not, however, the quick way to get information about 300 people.  My 
main objection to looking at the Lisp Questionnaire is, "You expect me 
to look manually through 230 people in hopes I find something?"  No way, 
man.

How do I search other people's interests, in a painless way?  Tooltips 
spring to mind.  I crank up the website, there's a randomly featured 
member in a sidebar.  A few of his major interests are given.  I suppose 
the definition of interests is the whim of the database designer, so 
that people can have checkboxes to click on, and they aren't required to 
play "guess the term."  I suppose we'd need ongoing maintenance by 
humans, designing new and useful categories according to user feedback.  
But I also wonder if Lisp can be used to construct some categories out 
of thin air.  That would be damn useful, because the data could 
potentially be extracted from any old crapola a Lisper was willing to 
write, and database designers might need to invent less and do less 
maintenance.

Emphasis on "might."  I'd happily start with augmented manual labor 
rather than AI or heavy duty text swizzling.

How do you even get a Lisper to write some sample crapola, so that 
there's even something to sample?  No participation, no data!  Eliza 
springs to mind.  If the extraction of user data is implemented as a 
silly game, perhaps users would bite the lure.

So now that I've laid out 1 year of R&D for someone, and the basis for 
an internet startup ala Meetup.com... anyone see the quick and dirty 
version of any of this?


Cheers,
Brandon Van Every