Hierarchical organization is a prevalent feature of many complex networks appearing in nature and society. Recent research works on hierarchy include technological networks, ecological systems, social interactions, and even neural networks. Several proposed models exist in network theory which are based on the observed growth patterns of real networks. A related interesting, yet less studied question is whether the evolution of hierarchies can be described similarly.
This thesis focuses on the analysis of how a specific hierarchy changes over time. I implemented my analysis on a given dataset, which is the yearly updated hierarchy of scientific subject terms, called Medical Subject Headings provided by National Center for Biotechnology Information (NCBI), which are organized into 16 different, yearly updated hierarchies such as Anatomy, Diseases, Chemicals and Drugs, etc. The natural representation of these hierarchies is given by directed acyclic graphs (DAGs), composed of links pointing from nodes higher in the hierarchy towards nodes in lower levels. Due to the appearance of new MeSH terms, the MeSH hierarchies are growing in time.
According to my results, not only growth but also restructuring of the already existing connections plays an important role as well in forming the shape of the DAGs. By defining variables for describing growth and restructuring and by creating several hierarchy attributes I revealed some significant evolution patterns (including preferential attachment) that are characteristic for MeSH hierarchies. Although the empirical studies in this work are restricted to the networks between MeSH terms, it is quite plausible that a part of these features are more universal and occur in the time evolution of hierarchical networks in general.