Using Conceptual Scaling In Formal Concept Analysis For Knowledge And Data Discovery In Medical Text

Formal concept analysis (FCA) 10] is a technique for knowledge representation and unsupervised machine learning that has been applied to a wide variety of knowledge representation and exploration tasks in a disparate range of disciplines including civil engineeringg7], psychologyy8], and political sciencee9]. Several software tools have been developed to aid knowledge exploration using FCA. The most prevelant of these being TOSCANAA4] but also otherss6, 2]. In these tools the process of creating conceptual scales is performed before the analysis stage by experts in FCA after interviews with domain experts. This paper describes a tool to allow the creation of conceptual scales by a domain expert through the manipulation of a folding Hasse diagram. By providing the domain expert with a set of simple graph manipulations, an intuitive understanding of FCA can facilitate the creation of conceptual scales and navigation of the data.