[Developers] [Data-modeling] Data load issues

[Brian Karlak]

On May 29, 2009, at 6:38 AM, Iain Sproat wrote:

> On a similar note, how do we deal with the resolution of duplicated  
> CVTs?  These are harder to fix through the client. (particularly as  
> they don't have a topic name with which to use the merge flag  
> against.)
>
> I assume the duplication of CVTs is a common scenario when uploading  
> properties from various data sources. I've just uploaded a bunch of  
> Irish Barons twice by accident.  And now there are two topics  
> linking the Peerage of Ireland to the relevant Noble title.
>
> Is there a bot which detects CVT's with identical properties and  
> merges, and CVTs with zero or one property and deletes?
>

Hello Ian --

For all of the reasons you describe, duplicate CVT cleanup needs to be  
an automated task.

Our gardening framework is being extended with logic to find and merge  
duplicate CVTs.  It should be completed soon; in the meantime I would  
not worry about trying to manually clean up duplicate CVTS.

Brian 
  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.freebase.com/pipermail/developers/attachments/20090529/5e093aed/attachment-0001.htm