Interactive Visualization of Formal Concept Lattices

Formal Concept Analysis (FCA) is suitable for use within organisations at different levels of maturity in information management. It takes as input a bigraph, into which both structured and unstructured data can be readily transformed, and produces a multiple-inheritance type hierarchy suitable for formal knowledge representation. Accordingly, FCA has been widely applied in areas such as information retrieval, knowledge discovery and knowledge representation. The multipleinheritance hierarchy produced by FCA is a complete lattice which can be represented as a labelled, directed, acyclic graph (DAG). We adopt a visual analytic approach to FCA by combining computational analysis with interactive visualisation. Scaling FCA to the interactive analysis of large data sets poses two fundamental challenges: the time required to enumerate the vertices, arcs and labels of the lattice DAG; and the difficulty of meaningful and responsive user interaction with a large digraph. This paper briefly describes three software prototypes which address aspects of these scalability challenges.