Assessing adherence to ethical principles and standards of normative behaviour?

I am studying how to integrate common sense reasoning with chunk based knowledge graphs, see:

	https://github.com/w3c/cogai/blob/master/demos/nlp/commonsense.md <https://github.com/w3c/cogai/blob/master/demos/nlp/commonsense.md> 

Common sense is needed to support natural language interaction and everyday reasoning. According to Jim Taylor, it can be defined as sound judgment derived from experience rather than study. In other words, it relies on general knowledge rather than specialised knowledge.

Humans are not a gold standard, as many people exhibit poor judgement, e.g. purchasing things they can't afford, smoking and eating junk food, holding irrational beliefs contrary to the evidence, as well as blatent prejudices against people from different backgrounds. Machine common sense needs to be assessed from a practical perspective, including adherence to ethical principles and standards of normative behaviour.

This raises the challenge for how to assess cognitive agents in respect to adherence to ethical principles and standards of normative behaviour, and in turn how can we codify such principles and standards?  Is it possible to include ethical principles and standards of normative behaviour as part of common sense and to attend to them as part of metacognition, akin to an inner voice?

Any thoughts you are willing to share on this challenge?

Many thanks for your contributions.

Dave Raggett <dsr@w3.org> http://www.w3.org/People/Raggett
W3C Data Activity Lead & W3C champion for the Web of things