Fusing semantic aspects for formal concept analysis using knowledge graphs

Formal Concept Analysis (FCA) is a field of applied mathematics with its roots in order theory. Over the past 20 years, FCA has been widely studied. Knowledge Graphs (KGs) model factual information in the form of entities and relations between them to semantically represent the world’s truth. Since existing semantic FCA (or ontology-based FCA) only relies on ontological knowledge, ontology-based FCA is limited in scope and scalability. This paper theoretically and empirically investigates a new semantic FCA by exploiting KGs to solve the scalability of ontology-based FCA. Concretely, we propose a kind of novel semantic FCA (i.e., KGs-based Formal Concept Analysis, KGs-based FCA) by semantifing FCA based on KGs. We further expand KGs-based FCA and propose a KGs-based FFCA (KGs-based Fuzzy Formal Concept Analysis) by extending KGs-based FCA. We also investigate the properties of KGs-based FCA and KGs-based FFCA in theory. The experimental results show that our proposals (KGs-based FCA and KGs-based FFCA) significantly outperform other traditional FCA and semantic FCA in information retrieval, and is 15.96–16.30 higher in Retrieval Results Ratio (R3) than other traditional FCA and semantic FCA.