YinYang bipolar logic and bipolar fuzzy logic (Q704147)

One of the main objectives of a knowledge base is to be able to use the corresponding knowledge to answer queries. For a given query \(Q\), we may have no information about \(Q\) at all; in this case, the truth value of \(Q\) is unknown (u). It is possible that we can deduce \(Q\) from the knowledge base, in which case the truth value of \(Q\) is ``true'' (t). It is possible that we can deduce \(\neg Q\) from the knowledge base, in which case the truth value of \(Q\) is ``false'' (f). In practice, it is also possible that some part of the knowledge base supports \(Q\) while some other part supports \(\neg Q\); in this case, all we can conclude about \(Q\) is that our knowledge about \(Q\) is inconsistent (i). Belnap has proposed and analyzed a special 4-valued logic with these 4 values u, t, f, and i. The authors extend Belnap's logic to the fuzzy case. Namely, in their new logic, the truth value of a statement \(Q\) is characterized by two numbers \(t^-\) and \(t^+\) from the interval \([0,1]\): \(t^+\) is the degree to which our knowledge supports \(Q\), and \(t^-\) is the degree to which our knowledge supports \(\neg Q\). Here, e.g., Belnap's u corresponds to \(t^-=t^+=0\), Belnap's i corresponds to \(t^-=t^+=1\), etc. This setting is very similar to Atanassov's intuitionistic fuzzy logic (IFS), where a truth value of a statement is also represented by a pair \((t^-,t^+)\), but IFS is limited to \textit{consistent} pairs, i.e., pairs for which \(t^-+t^+\leq 1\). The authors extend IFS to arbitrary pairs.