On the Galois correspondence for Hopf Galois structures arising from finite radical algebras and Zappa-Szép products (Q2232002)

\textit{S. U. Chase} and \textit{M. E. Sweedler} [Hopf algebras and Galois theory. Berlin etc.: Springer-Verlag (1969; Zbl 0197.01403)] defined the notion of a \textit{Hopf-Galois structure} on a finite extension of commutative rings. When applied to a finite Galois extension \(L/K\) of fields, it may be regarded as a generalization of the group ring \(K[G]\), where \(G\) is the Galois group. Given a Hopf-Galois structure \(H\) on \(L/K\), they also described an injective Galois correspondence from the \(K\)-subHopf algebras of \(H\) to the intermediate fields of \(L/K\), by associating to each \(K\)-subHopf algebra \(J\) its fixed field \[ L^J = \{ x\in L \mid jx = \varepsilon(j)x \text{ for all }j\in J\}, \] where \(\varepsilon: H\rightarrow K\) is the counit map. Unlike classical Galois theory, this correspondence need not be surjective, and one can consider the so-called \textit{Galois correspondence ratio} defined as \[ \frac{\#\{\mbox{those in the image of the Galois correspondence for \(H\)}\}}{\#\mbox{\{all intermediate fields of \(L/K\)\}}}. \] By Greither-Pareigis theory [\textit{C. Greither} and \textit{B. Pareigis}, J. Algebra 106, 239--258 (1987; Zbl 0615.12026)], the Hopf-Galois structure \(H\) arises from a regular subgroup \(N\) of \(\mathrm{Perm}(G)\) which is normalized by the subgroup \(\lambda(G)\) of left translations. Further by [\textit{T. Crespo} et al., Publ. Mat., Barc. 60, 221--234 (2016; Zbl 1331.12009)], there is a bijection between \(K\)-subHopf algebras of \(H\) and the subgroups of \(N\) which are normalized by \(\lambda(G)\). The Galois correspondence ratio for \(H\) may then be rewritten as \[ \frac{\#\{\mbox{subgroups of \(N\) which are normalized by \(\lambda(G)\)}\}}{\#\{\mbox{subgroups of \(G\)}\}}, \] whose calculation is a completely group-theoretic question. It is known that there is a close connection between Hopf-Galois structures and \textit{skew braces}, which may be viewed as a generalization of radical rings. The Galois correspondence ratio in the radical ring and skew brace settings, respectively, was previously studied by the author in [New York J. Math 23, 1--10 (2017; Zbl 07405613); J. Algebra 511, 270--291 (2018; Zbl 1396.12003)]. In the paper under review, he in particular considers radical rings \(A\) with \(A^3 = 0\) and skew braces \((B,\star,\circ)\) constructed from semidirect products, in which case the skew brace is a \textit{bi-skew brace}, namely \((B,\star,\circ)\) with the roles of the operations \(\star\) and \(\circ\) reversed is also a skew brace. He computes the Galois correspondence ratio for various examples. It is interesting to note that there could be a drastic difference between the Galois correspondence ratios for \((B,\star,\circ)\) and \( (B,\circ,\star)\).