[Data-modeling] Company histories -- more fun than a barrelofdinosaurs mating

[Jeff Prucher]

I've gotten lots of approving comments for this schema, and nary a
complaint, so I'm going to start the process of implementing the changes.
Most of the changes are additive, so they won't affect anything, and will be
coming to a graph near you very soon.  The date-mediation of the
company/subsidiary relationship requires some refactoring, so somebody
(probably Bryan) will send an annoucment before we do that.

Jeff 

> -----Original Message-----
> From: data-modeling-bounces at freebase.com 
> [mailto:data-modeling-bounces at freebase.com] On Behalf Of Jeff Prucher
> Sent: Friday, August 01, 2008 1:36 PM
> To: 'Freebase data modeling mailing list'
> Subject: Re: [Data-modeling] Company histories -- more fun 
> than a barrelofdinosaurs mating
> 
>  
> 
> > -----Original Message-----
> > From: data-modeling-bounces at freebase.com
> > [mailto:data-modeling-bounces at freebase.com] On Behalf Of John 
> > Giannandrea
> 
> > Id suggest that 'companies merging' be named 'merged with' 
> > since the CVT doesnt display the current topic and that 
> property name 
> > makes sense also from the 'product of merger'
> > side.
> 
> Neither choice seems very satisfactory to me. I agree that 
> "companies merging" is weird from the "involved in merger" 
> side. But I think that "merged with" is just as weird from 
> the "product of merger" side -- the product didn't merge with 
> anything. 
> 
> > also 'involved_in_merger' could conceivably be on defunct_company 
> > rather than company if you always mark a merged company as defunct.
> 
> It is possible for a company in a merger to continue as a 
> subsidiary of the new company; this is what happened with 
> Sperry Gyroscope when it merged to form Sperry Corp.  I think 
> I've also seen evidence of two companies technically merging 
> but continuing on under the corporate identity of one of the 
> two companies, so I think these should stay on Company.
> 
> Jeff
> 
> > 
> > On Jul 30, 2008, at 4:44 PM, Jeff Prucher wrote:
> > 
> > > After much banging of my head into the wall, I have a few 
> proposed 
> > > additions to the company schema that (I hope) will be able
> > to address
> > > some of the issues people have had about representing the
> > history of
> > > companies through M&A, divisions, name changes, and the
> > like.  It's a
> > > bit complex, so please bear with me.
> > >
> > > The biggest change is the addition of a CVT between companies and 
> > > their subsidiaries; this allows the entering of dates for the 
> > > subsidiary relationship, since subsidiaries have a tendency to be 
> > > acquired, sold, merged, and disbanded with some frequency.
> > I've also
> > > created a model for company divisions, which can also be 
> merged and 
> > > split, and which also have a dated relationship with their parent 
> > > company, since divisions can be transferred between
> > companies (as part
> > > of an M&A or just by themselves).
> > > I've also provided a link between an acquisition and division, to 
> > > allow entering companies that are acquired and turned into
> > divisions
> > > (as opposed to becoming subsidiaries or being completely
> > absorbed into
> > > the acquiring company).
> > >
> > > Most of the new types hang off of this one:
> > > http://sandbox.freebase.com/type/schema/business/company_division
> > > The other main one is the subsidiary relationship CVT:
> > > 
> > 
> http://sandbox.freebase.com/type/schema/business/subsidiary_relationsh
> > > ip
> > >
> > > Two use cases were put before me, which I think can be accurately 
> > > modeled in this revised schema.  One is from user
> > sprocketonline, who
> > > wanted to model the history of the engineering firm
> > Freeman, Fox and
> > > Partners. The text history is something like this:
> > > Freeman Fox merged with John Taylor & Sons in 1987 to form Acer 
> > > Consultants.
> > > Acer was purchased by Welsh Water in 1993 and merged with another 
> > > subsidiary (Wallace Evans & Partners) to form Hyder
> > Consulting (which
> > > went under a few names). Hyder was purchased by Western Power 
> > > Distribution in 2000, and bought itself out as an
> > independent company
> > > in 2001.  Here's how this looks in my model; these are 
> the links to 
> > > Acer Consultants and Hyder Consulting; everything else is one link
> > > away:
> > > 
> > 
> http://sandbox.freebase.com/view/guid/9202a8c04000641f8000000008d3d94e
> > > http://sandbox.freebase.com/view/en/hyder_consulting
> > >
> > > The other use case is our own Jamie Taylor's pet example of
> > an overly
> > > complex company history, the Sperry/Burroughs merger into
> > Unisys, and
> > > the attendant corporate histories.  Here's some of what I'm
> > trying to
> > > capture:.
> > > For Remington Rand: http://sandbox.freebase.com/view/en/
> > > remington_rand:
> > > The foundation from a merger of three companies; The 
> acquisition of 
> > > Eckert-Mauchly, which became the UNIVAC Division of Rand The 
> > > acquistion of Engineering Research Associates, which also 
> became a 
> > > division of Rand
> > >
> > > For Sperry 
> (http://sandbox.freebase.com/view/en/sperry_corporation):
> > > The foundation from the merger of two companies, one of
> > which became a
> > > subsidiary; Many name changes; The acquisition of Remington
> > Rand, and
> > > the consequent merger of the UNIVAC Division and
> > Engineering Reasearch
> > > Associates into a new division, Sperry UNIVAC The merger with 
> > > Burroughs to form Unisys
> > >
> > > For Unisys (http://sandbox.freebase.com/view/en/unisys):
> > > In addition to the aforementioned merger, the acquisition of 
> > > Convergent Technologies to form its Network Systems Division.
> > >
> > > I think I've captured everything, but I'm not sure how easy
> > it will be
> > > to traverse.  Comments and suggestions are most welcome.  
> Note that 
> > > since this is on sandbox, the examples will self-destruct
> > this Monday.
> > >
> > > Jeff Prucher
> > > Type Librarian & Ontologist
> > > Metaweb Technologies, Inc.
> > >
> > >
> > > _______________________________________________
> > > 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
> > 
> 
> 
> _______________________________________________
> Data-modeling mailing list
> Data-modeling at freebase.com
> http://lists.freebase.com/mailman/listinfo/data-modeling
>