Handling incomplete information in formal concept analysis - a possibilistic approach

Abstract It is very common to find databases with missing information. Therefore, it is important to develop formal tools. This paper focuses on the contribution of Formal Concept Analysis to this fundamental goal. For this purpose, five forms of attribute implications are extracted from incomplete contexts, which are analyzed from two different approaches. The first approach follows the traditional recipe taking into account the well-known characterizations about the validity of attribute implications. The second approach goes further by considering possibility theory. Specifically, possibility and necessity measures are defined in order to establish the plausibility and certainty of relevant statements pertaining to an incomplete context.