Scalable Performance of FCbO Algorithm on Museum Data

Formal Concept Analysis – known as a technique for data analysis and visualisation – can also be applied as a means of creating interaction approaches that allow for knowledge discovery within collec- tions of content. These interaction approaches rely on performant algo- rithms that can generate conceptual neighbourhoods based on a single formal concept, or incrementally compute and update a set of formal concepts given changes to a formal context. Using case studies based on content from museum collections, this paper describes the scalabil- ity limitations of existing interaction approaches and presents an imple- mentation and evaluation of the FCbO update algorithm as a means of updating formal concepts from large and dynamically changing museum datasets.