Contribution to the Classification of Web of Data based on Formal Concept Analysis

Justine Reynaud (1) , Yannick Toussaint (1) , Amedeo Napoli (1)

1 ORPAILLEUR - Knowledge representation, reasonning

During the last decade, the web has taken a huge importance in everyday life, and has become what is commonly called a web of data. The available resources can be used by human agents but also by software agents, as it is the case for very large ontologies such as YAGO or resources such as DBpedia. These particular datasets can be linked together for constituting the Linked Open Data (LOD) cloud, where basic data are expressed as (subject, predicate, object) triples. One issue of main interest is knowledge discovery within LOD, which can help information retrieval and knowledge engineering. Formal concept analysis (FCA), which is a mathematical theory allowing classification and data analysis, was already used to classify LOD elements. In this research work, we are interested in analyzing the different approaches (extensions) based on FCA for knowledge discovery in the web of data. One objective is to study the efficiency and the applicability of the existing approaches and to propose some improvements.

Mots clés: Formal Concept Analysis; Pattern Structures; Triadic Con- cept Analysis; Linked Open Data

Domaines: Intelligence artificielle [cs.AI]

Justine Reynaud, Yannick Toussaint, Amedeo Napoli. Contribution to the Classification of Web of Data based on Formal Concept Analysis. What can FCA do for Artificial Intelligence (FCA4AI) (ECAI 2016), Aug 2016, La Haye, Netherlands. ⟨hal-01420466⟩