On Concept Lattices for Numberings
Nikolay Bazhenov,Manat Mustafa,Anvar Nurakunov
DOI: https://doi.org/10.26599/tst.2023.9010102
2024-06-21
Tsinghua Science & Technology
Abstract:The theory of numberings studies uniform computations for families of mathematical objects. In this area, computability-theoretic properties of at most countable families of sets are typically classified via the corresponding Rogers upper semilattices. In most cases, a Rogers semilattice cannot be a lattice. Working within the framework of Formal Concept Analysis, we develop two new approaches to the classification of families . Similarly to the classical theory of numberings, each of the approaches assigns to a family its own concept lattice. The first approach captures the cardinality of a family : if contains more than 2 elements, then the corresponding concept lattice FC1( ) is a modular lattice of height 3, such that the number of its atoms to the cardinality of . Our second approach gives a much richer environment. We prove that for any countable poset , there exists a family such that the induced concept lattice FC2 ( ) is isomorphic to the Dedekind-MacNeille completion of . We also establish connections with the class of enumerative lattices introduced by Hoyrup and Rojas in their studies of algorithmic randomness. We show that every lattice FC2 ( ) is anti-isomorphic to an enumerative lattice. In addition, every enumerative lattice is anti-isomorphic to a sublattice of the lattice FC2 ( ) for some family .
computer science, information systems,engineering, electrical & electronic, software engineering