First draft for PKN specification

Your comments are welcomed on the first draft of a proposed CG specification for the plausible knowledge notation, see: https://w3c.github.io/cogai/pkn.html 

This deals with an intuitive syntax for defeasible reasoning that creates a presumption in favour of the conclusions. However these may need to be withdrawn in the light of new information. Arguments in support of, or counter to, some supposition, build upon the facts in the knowledge graph or the conclusions of previous arguments.

The Semantic Web is founded on description logic and deductive proof - however that is an inadequate framework in respect to everyday knowledge subject to uncertainties, imprecision, incompleteness and inconsistencies.

There are plenty of open questions, e.g.

How to model higher order statements such as “Mary thinks John is lying when he says he loves Joan”?

How to model reasoning strategies and tactics declaratively using an intuitively understandable syntax rather than having to hard code them as part of the reasoner’s implementation?

Whether machine learning of large knowledge graphs together with reasoning strategies and tactics is better done with vector spaces  (i.e. neural networks) rather than symbolic representations?

What is the role of symbolic representations of knowledge in the era of large language models?

Best regards,

Dave Raggett <dsr@w3.org>