[Data-modeling] Adoptive and other relationships

[Jeff Prucher]

Foster sounds like a good addition. We could add it a couple ways. One is with a single "Foster" topic, and acknowledge the fact that the meaning varies over time and location. The other is to create multiple Foster topics for the various kinds of fostering.

Jeff

----- Original Message -----
From: "Faye Harris" <faye at metaweb.com>
To: "Freebase data modeling mailing list" <data-modeling at freebase.com>
Sent: Friday, April 10, 2009 2:53:14 PM GMT -08:00 US/Canada Pacific
Subject: Re: [Data-modeling] Adoptive and other relationships

I like the model.

Adoption is a legal process today in most countries, but used to be 
less...formal. Should I just use "Adoptive" to describe foster 
relationships of yesterday where kids may never have been legally 
adopted, or is there a better word for it that should be added to the 
"type of relationship" enumeration? Even the word "foster" carries a 
legal meaning today. I'm thinking of -- who else -- Edgar Allan Poe, 
whose foster parents, the Allans, never became his legal adoptive parents.

-- Faye


Jeff Prucher wrote:
> Following up on a thread from while I was out (now also known as <https://bugs.freebase.com/browse/DA-694>), I've put up a model for mediating parent/child relationships, including a relationship type property to the sibling relationship, on sandbox.
>
> Schema: <https://www.sandbox-freebase.com/type/schema/people/person>
> The properties to note are "parents 2", "children 2", and "siblings".
>
> And what could be better for sample data than the British royal family?  Camilla's biological children are a particularly nice test case -- they each have 1 biological sibling, 2 step siblings (Charles' sons), and 1 half sibling* (from their father's second marriage).  They also have 2 biological and 1 step-parent.  See <https://www.sandbox-freebase.com/view/en/laura_lopes> for an example.
>
> The relationship types are enumerated lists; more values can be added if needed -- I just populated them with what I thought were likely values.
>
> And just to reiterate Robert's earlier comment, if we do do this, it will totally break any application that uses the parent/child relationship, of which there are probably many. Not necessarily a reason not to do this, but just to keep that in mind.
>
> Jeff
>
> *Note: technically they have three half siblings, but hey, it's only a model.
> _______________________________________________
> Data-modeling mailing list
> Data-modeling at freebase.com
> http://lists.freebase.com/mailman/listinfo/data-modeling
>
>   

_______________________________________________
Data-modeling mailing list
Data-modeling at freebase.com
http://lists.freebase.com/mailman/listinfo/data-modeling