Generating explicit orderings for non-monotonic logics

摘要: For non-monotonic reasoning, explicit orderings over formulae offer an important solution to problems such as 'multiple extensions'. However, a criticism of is that it not clear, in general, from where the should be obtained. Here we show how can derived statistical information about domain which cover. this provide overview prioritized logics--a general class logics incorporate formulae. This has been shown elsewhere capture wide variety proof-theoretic approaches and particular, highlight role preferences-both implicit explicit--in proof theory. We take one particular logic, called SF describe experimental approach for comparing logic with example does use preference-namely Horn clause negation-as-failure. Finally, present results companson, showmg more skeptical accurate than