Concept Similarity in Formal Concept Analysis with Many-Valued Contexts

Abstract

Formal Concept Analysis (FCA) is a mathematical framework which can also support critical activities for the development of the Semantic Web. One of them is represented by Similarity Reasoning, i.e., the identification of different concepts that are semantically close, that allows users to retrieve information on the Web more efficiently. In order to model uncertainty information, in this paper FCA with many-valued contexts is addressed, where attribute values are intervals, which is referred to as FCA with Interordinal scaling (IFCA). In particular, a method for evaluating concept similarity in IFCA is proposed, which is a problem that has not been adequately investigated, although the increasing interest in the literature in this topic.