Core Concepts
The lack of a central location that provides and describes FCA data sets and links them to already known analysis results poses a problem for the sustainable development of the research field. This work analyzes the requirements for an FCA repository and proposes a simple, file-based solution to address this issue.
Abstract
The paper analyzes the current state of the dissemination of FCA data sets and presents the requirements for a central FCA repository. It highlights the challenges in establishing such a repository.
The main parts of the proposed FCA repository are:
Formal contexts: The central entity, including objects, attributes, and incidence relation.
Simple statistics, metadata, and usage information for each context.
Relationships between contexts, such as scaling and sub-contexts.
Collections of standard benchmark contexts.
Storage of formal concepts, concept lattices, and implication bases.
The repository is implemented as a file-based system, using a git repository for version control and collaboration. The metadata for each context is stored in a machine-readable and human-editable YAML file. The formal contexts are stored in the Burmeister format, which is well-supported by FCA tools and libraries.
The paper discusses implementation considerations, such as file naming, location, and content representation. It also proposes the establishment of a working group to develop a curation policy and drive the further development of the repository. Integrating the repository into the FCA ecosystem, by providing easy programmatic access from FCA tools, is another key aspect.
The scope of the repository is limited to formal contexts and their metadata, without aiming for a comprehensive modeling of all FCA data structures. The goal is to provide a resilient and community-driven solution that can serve as a foundation for further improvements.