A knowledge discovery technique for heterogeneous data sources

Knowledge discovery is the non-trivial extraction of implicit, previously unknown and potentially useful information from data. We present a model of how concepts are structured within data sources, after exploring current conceptual structures applied to represent concepts embedded within data sources. These techniques include formal concept analysis (FCA), conceptual graphs (CG), and structured concepts (SC). By developing a hybrid conceptual structure, we intend to capture the key features of FCA, CG, and SC. In the end of this paper, we also present a system architecture for conceptual knowledge discovery.