[Data-modeling] Chemist & other types that are professions -was Re: State of the commons (C-F)

[Jeff Prucher]

> -----Original Message-----
> From: data-modeling-bounces at freebase.com 
> [mailto:data-modeling-bounces at freebase.com] On Behalf Of Robert Cook
> Sent: Wednesday, July 01, 2009 5:37 PM
> To: Freebase data modeling mailing list
> Subject: Re: [Data-modeling] Chemist & other types that are 
> professions -was Re: State of the commons (C-F)
> 
> Typically, I think professional types should be created only 
> out of necessity.  If there is a lot of useful data around 
> that profession (say, which movies you acted in), then it's a 
> good thing to have.  If we add Chemists and Physicists, why 
> not Nurses, Physicians and Plumbers?
> 
> Speaking of professions and professional types, I recently 
> got another request from a startup using Freebase data to 
> synch up these denormalizations using automatic data 
> processes to copy data back and forth between the representations.
> 
> That is, if somebody is typed as politician, he/she should 
> have the profession set to politician.  Not all 
> profession/type pairs would have this mapping, and even fewer 
> would have them symmetrically (/film/ actor -> profession: 
> actor is definitely one way).

Yikes.  I'm not sure how many of these types could be said to reliably map
to profession. I don't know that I'd say that /film/actor -> profession:
actor is even one way. Frank Miller (the comics artist) is typed as an actor
because he's had several cameo appearances, but it's hardly his profession.
Similarly, Jello Biafra probably shouldn't have a profession of politician,
no matter how many times he's run for mayor of San Francisco. Maybe there's
a threshold that should be met before making such assertions?  (Although the
reverse mapping, e.g. Profession:writer -> /book/author, doesn't seem
problematic to me.)

Jeff