Measures of inset information of the open domain. I: Inset entropies and information functions of all degrees (Q1075289)

This is one in a series of papers [for parts II and III see the reviews below (Zbl 0591.94013 and Zbl 0591.94014)] studying measures of information in the so-called ''mixed'' theory of information (i.e. considering the events as well as their probabilities) on the ''open'' domain (i.e. without empty sets and zero probabilities). In this paper we find all \(\beta\)-recursive, 3-semisymmetric inset entropies on the open domain. We do this by solving the fundamental equation of inset information of degree \(\beta\) (\(\beta\in {\mathbb{R}})\) on the open domain. One result presented is that all \(\beta\)-recursive, 3-symmetric inset entropies on the open domain are of the form \[ \begin{alignedat}{2} &\sum^{n}_{i=1} g(x_ i) p_ i^{\beta} - g\left(\bigcup^{n}_{i=1} x_ i\right) \quad&&\text{if \(\beta\neq 0,1\);}\\ &\sum^{n}_{i=1} [g(x_ i) + L(p_ i)] p_ i - g\left(\bigcup^{n}_{i=1} x_ i\right) \quad&&\text{if \(\beta =1\);}\\ &\sum^{n}_{i=1} [g(x_ i) + \ell(x_ i,p_ i)] - g\left(\bigcup^{n}_{i=1} x_ i\right) \quad&&\text{if \(\beta =0\);} \end{alignedat} \] where \(g\) is an arbitrary real-valued function on the event space \(D\), \(L\) is logarithmic, and \(\ell:D\times]0,1[\to\mathbb{R}\) is an additive set function in the first variable and logarithmic in the second variable. This is the first place such \(\ell\) terms appeared.