Theory of expertons and fuzzy logic (Q1111543)

Starting with a universe E of properties and a set of experts, each expert is asked to give his opnion about each element of E in the form of an interval of confidence, i.e. a closed subinterval of [0,1]. After some statistical calculations, that are not indicated in the paper, one obtains for each element of E and for each level\(\in [0,1]\) a closed subinterval of [0,1]. This double-indexed family of closed subintervals of [0,1] should be called (as it is not done explicitly) an experton. Then the classical algebraic operations of ``union'', ``intersection'' and ``complementation'' of expertons are defined (just as for fuzzy sets, flou sets, probabilistic fuzzy sets) and hence not surprisingly at all, they show some lattice structure. In fact the operations defined are nothing but Moore's interval calculus operations. The third and last paragraph concerns the extension of some multi-valued logic implications to expertons. Some comments: (i) The title of this paper is misleading: it pretends to give a theory of expertons and fuzzy logic on 9 pages! (ii) The author claims, however without proof, that the notion of experton, without defining it, generalizes probabilistic fuzzy sets and fuzzy systems. But the usefulness and the gain w.r.t. the generalized notions of this experton concept is nowhere indicated.