Holger,
Would you agree that a specific set of "Shapes" *are* what you call "the validation algorithm". Surely those concepts would have to be synonymous from either POV. If not, how might they be different?
Jeff
> On Jan 23, 2015, at 10:39 PM, Holger Knublauch <[log in to unmask]> wrote:
>
>> On 1/24/15, 12:57 PM, Karen Coyle wrote:
>> Holger, how I read what you are saying is that there are two options:
>>
>> - validation based on classes
>> - global validation
>>
>> If one is not uses classes, how is the "whole graph" defined? What makes a graph "the whole graph"? A single subject? Some shape? Or does whole graph mean absolutely every triple in your triple-store?
>
> The "graph" is every triple in the default RDF graph that was passed into the validation algorithm. A triple-store (dataset) may contain additional (named) graphs that are not part of the validation.
>
> Holger
|