Triangles are among the most important features of networks. The clustering coefficient, which measures the fraction of closed triangles in a network, is one of the most significant and frequently analyzed descriptors of networks, second only to the degree distribution. It has numerous implications for stability, spreading, community structure, link prediction, etc. Despite their importance, little is known about the general distribution and dynamics of clustering coefficients. While some models, such as the Kumpula model (Kumpula et al., 2009), focus on triadic closure in social networks, a comprehensive theory of their dynamics is still lacking.

The aim of this thesis is to conduct an extensive analysis of clustering coefficients across various empirical datasets, develop a model that is as general as possible based on the observations, and understand the limitations and applicability of the proposed model.