Resumen de Temporal analysis of large dynamic graphs
The objective of this thesis is to provide a temporal analysis of the structural and interaction dynamics of large evolving graphs. In this work we propose new definitions of important graph metrics in order to include the temporal dimension of dynamic graphs. We further extend three important problems of graph mining in the temporal setting. The three problems that we study are temporal graph summarization, temporal community search, and temporal betweenness centrality.
We start with the high level analysis of the dynamic graph and with the problem of temporal graph summarization. Our approach is based on a modification of graph clustering in temporal graphs. Then in a mid-level approach, we continue with the problem of community search.
Our solution is based on extracting a temporal selective connector according to our definition of shortest-fastest paths.
Finally, analyzing the graph at the vertex level, we propose a new metric for temporal betweenness centrality, based on shortest-fastest paths, and we provide an algorithm for quickly computing it.
Additionally, we propose a distributed version of all our algorithms, that help our techniques to scale up to million vertices.
We finally evaluate the validity of our methods in terms of efficiency and effectiveness with extensive experimentation on large-scale real-world graphs.
