A single-point characterization of representable uninorms (Q695242)

The paper gives a characterization of representable uninorms by an additive generator. It starts with the definitions of uninorms and representable uninorms. The next section presents some results about the representability of proper uninorms (with neutral element strictly between 0 and 1) that are necessary for the next parts of the paper. For a uninorm \(U\) with neutral element \(e \in ]0, 1[\) such that the underlying t-norm \(T_U\) and the underlying t-conorm \(S_U\) are strict and \(E=[0, 1]^2 - ([0, e]^2 \cup [e, 1]^2)\), \(\check{E}=E \cap ]0, 1[^2\), the attainment by \(U\) of one of the bounds \(\min\) and \(\max\) on \(\check{E}\) is determined by the value at an arbitrary point in this region (Section 3) and the values \(U\) taken on \(\check{E}\) cannot by bounded by the neutral element \(e\) (Section 4). The main result concerning the characterization of representable uninorms \(U\) is given in Theorem 4 in Section 5 and affirms that a uninorm \(U\) with neutral element \(e \in ]0,1[\) such that both \(T_U\) and \(S_U\) are strict is representable iff there exists a point where \(U\) takes a value strictly between the two arguments.