Technological advances in the natural and social sciences enable scientists, researchers, academics and other stakeholders to collect more data on complex networks than ever before. The speed and capacity of immense information gathering is continuously improving. However, comprehensive tools are lacking to turn information into practical knowledge. What is more, issues arise regarding the reliability of network data, leading to concern over the legitimacy of network research outcomes. To address these concerns, the EU-funded 'Decomposition and discovery of complex networks' (DEDINET) project provided insight into the processes that influence network structure. Overall, the aim was to develop a framework to identify block models and detect complex networks. Block models serve as good examples of interactions between nodes — individual actors, people or things within a network. The resulting framework will be used to tackle problems, particularly in systems biology. In developing the framework, project members devised methods to test and compare block models in different kinds of biological and social networks involving various nodes and interactions. Results showed that several block models are necessary for network structure and that interaction patterns are the result of multiple block models coming together. By capturing block models, the team was able to discover and predict missing or future data in networks. DEDINET succeeded in identifying several complementary block structures, leading to a better understanding of the fundamental mechanisms behind the development and makeup of networks. Outcomes are expected to have a major impact in fields such as social sciences, economics and biology.
Complex networks, network analysis, network data, network structure, nodes