Canonical Computational Models Based on Formal Concept Analysis for Social Network Analysis and Representation

Formal concept analysis is a mathematics research field introduced in the beginning of the 1980s by Rudolf Wille, that has been applied in several different knowledge areas, including Computer Science. FCA is a data analysis theory that identifies conceptual structures within data sets or formal contexts. In this work, we propose an FCA-based approach to build minimal implication rules-based computational models for social networks. As an application example, in this work we constructed canonical models using data extracted from user sessions in one of the most popular social networks in Brazil, Orkut. These models represent the patterns of access to Orkut, about a certain problem domain, and are composed by a minimal rule set.