Abstract <p>In this paper, the formalization of heuristics used to generate explanatory typologies in the process of empirical typology has been considered. The basis for constructing this typology is interpretable models resulting from the constructive generation of causal relations in limited (but potentially expandable) data sets. For this purpose, an approach that implements plausible reasoning through logical means: a synthesis of induction, analogy, and abduction has been proposed. Some results of typology based on social monitoring useful for the development of theories and applied recommendations for social policy have been presented.</p>