FCA applied in social networks with use of BDD

In recent years, the increasing complexity of a hyper-connected world has directed the attentions of many types of research and because of that, we need new multidisciplinary complex network analysis. Our goal is to find and analyze users access patterns using a type of association rules named implications and enable large database handling through adaptation of Formal Concept Analysis (FCA) algorithms. Our results may help create security policies and network settings to better predict and control future users access. However, with the increasing amount of information from complex networks such as social networks, powerful computing resources are needed. We propose the use of binary decision diagram (BDD) to handle this situation. In this work, the BDD is adapted to FCA derivation operators allowing the handling of a quantity of information that without this technique would not be feasible. To evaluate BDD benefits, we used two types of algorithms, an incremental and a batch, both showed performance gains in formal context with high dimensionality and density.