On the Logical and Algebraic Aspects of Reasoning with Formal Contexts
A formal context consists of objects, properties, and the incidence relation between them. Various notions of concepts defined with respect to formal contexts and their associated algebraic structures have been studied extensively, including formal concepts in formal concept analysis (FCA), rough concepts arising from rough set theory (RST), and semiconcepts and protoconcepts for dealing with negation. While all these kinds of concepts are associated with lattices, semiconcepts and protoconcepts additionally yield an ordered algebraic structure, called double Boolean algebras. As the name suggests, a double Boolean algebra contains two underlying Boolean algebras. In this article, we investigate logical and algebraic aspects of the representation and reasoning about different concepts with respect to formal contexts. We first review our previous work on two-sorted modal logic systems KB and KF for the representation and reasoning of rough concepts and formal concepts, respectively. Then, in order to represent and reason about both formal and rough concepts in a single framework, these two logics are unified into a two-sorted Boolean modal logic BM , in which semiconcepts and protoconcepts are also expressible. Based on the logical representation of semiconcepts and protoconcepts, we prove the characterization of double Boolean algebras in terms of their underlying Boolean algebras. Finally, we also discuss the possibilities of extending our logical systems for the representation and reasoning of more fine-grained quantitative information in formal contexts.
Authors
- Churn‐Jung Liau (ORCID: https://orcid.org/0000-0001-6842-9637)
- Prosenjit Howlader (ORCID: https://orcid.org/0000-0002-8932-126X)
Institutions
- Institute of Information Science, Academia Sinica (TW)
Publication Details
- Journal
- ACM Transactions on Computational Logic
- Published
- 2025-05-05
- DOI
- https://doi.org/10.1145/3733832
- Citations
- 1
- Primary Topic
- Rough Sets and Fuzzy Logic
- Type
- article
- Field-Weighted Citation Impact
- 1.50