Abstract: Formal Concept Analysis (FCA) is a prominent field of applied mathematics using formal concepts – maximal clusters of object-attribute relationships – to discover, process, and represent knowledge in so-called conceptual hierarchies. Its efficient algorithms and expressive power makes FCA suitable to unify methodologies and to provide an in-depth insight on knowledge structures [1], [2]. Electronic Health Record (EHR) systems are nowadays widespread and used in different scenarios. In this paper we consider the problem of improving EHR systems with new, FCA grounded features. For this, we start with some particular medical data sets and discuss the improvement of some features of EHR systems by using FCA. Two main methods have been taken into consideration so far. First, we consider the medical data sets as many-valued contexts. By using conceptual scaling, we build `knowledge landscapes’ [3] and show how these `landscapes’ might be used in the framework of EHR. A complementary approach is based on Triadic FCA (3FCA) approach. We exemplify these methods on several medical datasets and discuss how conceptual landscapes can be used to improve not only the integrated view of patient data (as an EHR system specific feature), but also communication and support future research.
Keywords: Formal Concept Analysis, Electronic Health Record, medical data, many-valued contexts, conceptual scaling, Triadic FCA
Acknowledgments: Diana Halita was supported by a doctoral research from POSDRU/187/1.5/S/155383. Special thanks to Dr. Daniela Grecea, medical investigator, and her team.
Presentation: SOFTCOM 2016 – Session 1: Education and Information Systems
