Group Querying in Tridimensional Social Networks

In this paper, we present a new approach based on Triadic Concept Analysis for discovering an approximation or full match of triples using an inverted index, eliminating the need for a pre-processing phase to transform triadic contexts into dyadic ones and avoiding the application of derivation operators. To classify the identified triadic concepts, we introduce a new similarity metric based on the user’s query. We conducted an empirical study mainly to illustrate the efficiency and scalability of our approach. Our proposed solution shows not only superior efficiency compared to existing techniques documented in the literature, but also better scalability, making it suitable for big data scenarios. Finally, we present a case study for analyzing a tridimensional social network to illustrate the meaning and the cost of the querying process of a triple among a collection of 3D-clusters, rather than identifying 3D communities describing individuals, their attributes, and the conditions under which these attributes occur.