2 ms·
I would say that OWA is an interesting feature, but still far from being perfect. In the real world cases, we need sometimes to perform some sort of validation,
by haddr 11y ago
I would say that OWA is an interesting feature, but still far from being perfect. In the real world cases, we need sometimes to perform some sort of validation, where we need true/false answer rather than Unknown. This is where OWA falls short. For instance in the case of cardinality problem. (For those who doesn't know: In OWL you can say that a person can have only 1 father, but then you might add 2 fathers for one person, and the reasoner would be perfectly fine with that. It assumes that both persons might be actually the same instance)
- woodman 11y ago> ...where we need true/false answer rather than Unknown. If you've set things up with that sort of constraint, then you've set yourself up for failure - Gödel's incompleteness theorems [0]. I think OWA is the only way around that, you just need to gracefully handle unknowns. Of course, you can convert an unknown into a known pretty easily with: ¬u :) > For instance in the case of cardinality problem. That is exactly when I'd want my reasoner to ignore a graph edge, or else you get a non sequitur that ends with Nike belonging in the human set. Measure twice, cut once. [0] http://en.wikipedia.org/wiki/G%C3%B6del%27s_incompleteness_theorems http://en.wikipedia.org/wiki/G%C3%B6del%27s_incompleteness_t...