w3c / rdf-star

RDF-star specification
https://w3c.github.io/rdf-star/
Other
119 stars 23 forks source link

Uncertainty representation #132

Closed djebR closed 3 years ago

djebR commented 3 years ago

As a Ph.D. student working on uncertainty, I want to use RDF* to represent uncertainty values for RDF statements, using mUnc ontology, This offers the possibility to evaluate the reliability of statements in a specific context.

following this: http://ns.inria.fr/munc/

hartig commented 3 years ago

Hi @djebR ! You may be interested in the following related work about capturing and querying uncertainty for individual RDF statements based on RDF*. While this work is in the context of RDF stream processing, the concepts carry over to the static case as well. The main paper about this work is:

djebR commented 3 years ago

Indeed, I attended that presentation and the author mentioned that use case and mUnc ontology in the perspectives. Unfortunately, we didn't have the time to discuss it.

I am planning on including RDF* as the main way to annotate statements with uncertainty values.

afs commented 3 years ago

Hi there - the use case has been added to the UCR document.

It would be good to hear your thoughts on RDF-star (the work of this group) where the triple statement mentioned is not also asserted and whether this can form a useful basis for use with mUnc.

djebR commented 3 years ago

Hi there - the use case has been added to the UCR document.

It would be good to hear your thoughts on RDF-star (the work of this group) where the triple statement mentioned is not also asserted and whether this can form a useful basis for use with mUnc.

One example I can think of is to assert the appartenance of an embedded triple to the context (the World) it was declared in. <<ex:Apple ex:hasColor ex:Red>> munc:inWorld ex:World_1 {| munc:hasMeta [a munc:Uncertainty; ... ] |}

TallTed commented 3 years ago

@djebR --

You may want to make that comment in a new issue, perhaps specifically targeted to the UCR document -- as comments made on closed issues are easily overlooked.