Ulrik Brandes

A Faster Algorithm for Betweenness Centrality

Motivated by the fast‐growing need to compute centrality indices on large, yet very sparse, networks, new algorithms for betweenness are introduced in this paper. They require O(n + m) space and run in O(nm) and O(nm + n2 log n) time on unweighted and weighted networks, respectively, where m is the number of links. Experimental evidence is provided that this substantially increases the range of networks for which centrality analysis is feasible. The betweenness centrality index is essential in the analysis of social networks, but costly to compute. Currently, the fastest known algorithms require ?(n 3) time and ?(n 2) space, where n is the number of actors in the network.

On Modularity Clustering

Modularity is a recently introduced quality measure for graph clusterings. It has immediately received considerable attention in several disciplines, particularly in the complex systems literature, although its properties are not well understood. We study the problem of finding clusterings with maximum modularity, thus providing theoretical foundations for past and present work based on this measure. More precisely, we prove the conjectured hardness of maximizing modularity both in the general case and with the restriction to cuts and give an Integer Linear Programming formulation. This is complemented by first insights into the behavior and performance of the commonly applied greedy agglomerative approach.

On variants of shortest-path betweenness centrality and their generic computation

Betweenness centrality based on shortest paths is a standard measure of control utilized in numerous studies and implemented in all relevant software tools for network analysis. In this paper, a number of variants are reviewed, placed into context, and shown to be computable with simple variants of the algorithm commonly used for the standard case.

Experiments on Graph Clustering Algorithms

A promising approach to graph clustering is based on the intuitive notion of intra-cluster density vs. inter-cluster sparsity. While both formalizations and algorithms focusing on particular aspects of this rather vague concept have been proposed no conclusive argument on their appropriateness has been given. As a first step towards understanding the consequences of particular con- ceptions, we conducted an experimental evaluation of graph clustering approaches. By combining proven techniques from graph partitioning and geometric clustering, we also introduce a new approach that compares favorably.

CENTRALITY ESTIMATION IN LARGE NETWORKS

Centrality indices are an essential concept in network analysis. For those based on shortest-path distances the computation is at least quadratic in the number of nodes, since it usually involves solving the single-source shortest-paths (SSSP) problem from every node. Therefore, exact computation is infeasible for many large networks of interest today. Centrality scores can be estimated, however, from a limited number of SSSP computations. We present results from an experimental study of the quality of such estimates under various selection strategies for the source vertices.

GraphML Progress Report Structural Layer Proposal

Following a workshop on graph data formats held with the 8th Symposium on Graph Drawing (GD 2000), a task group was formed to propose a format for graphs and graph drawings that meets current and projected requirements. On behalf of this task group, we here present GraphML (Graph Markup Language), an XML format for graph structures, as an initial step towards this goal. Its main characteristic is a unique mechanism that allows to de.ne extension modules for additional data, such as graph drawing information or data specific to a particular application. These modules can freely be combined or stripped without affecting the graph structure, so that information can be added (or omitted) in a well-defined way.

Centrality measures based on current flow

We consider variations of two well-known centrality measures, betweenness and closeness, with a different model of information spread. Rather than along shortest paths only, it is assumed that information spreads efficiently like an electrical current. We prove that the current-flow variant of closeness centrality is identical with another known measure, information centrality, and give improved algorithms for computing both measures exactly. Since running times and space requirements are prohibitive for large networks, we also present a randomized approximation scheme for current-flow betweenness.

Eigensolver methods for progressive multidimensional scaling of large data

We present a novel sampling-based approximation technique for classical multidimensional scaling that yields an extremely fast layout algorithm suitable even for very large graphs. It produces layouts that compare favorably with other methods for drawing large graphs, and it is among the fastest methods available. In addition, our approach allows for progressive computation, i.e. a rough approximation of the layout can be produced even faster, and then be refined until satisfaction.

On finding graph clusterings with maximum modularity

Modularity is a recently introduced quality measure for graph clusterings. It has immediately received considerable attention in several disciplines, and in particular in the complex systems literature, although its properties are not well understood. We study the problem of finding clusterings with maximum modularity, thus providing theoretical foundations for past and present work based on this measure. More precisely, we prove the conjectured hardness of maximizing modularity both in the general case and with the restriction to cuts, and give an Integer Linear Programming formulation. This is complemented by first insights into the behavior and performance of the commonly applied greedy agglomaration approach.

Longitudinal analysis of personal networks : the case of Argentinean migrants in Spain

This paper discusses and illustrates various approaches for the longitudinal analysis of personal networks (multilevel analysis, regression analysis, and SIENA). We combined the different types of analyses in a study of the changing personal networks of immigrants. Data were obtained from 25 Argentineans in Spain, who were interviewed twice in a 2-year interval. Qualitative interviews were used to estimate the amount of measurement error and to isolate important predictors. Quantitative analyses showed that the persistence of ties was explained by tie strength, network density, and alters’ country of origin and residence. Furthermore, transitivity appeared to be an important tendency, both for acquiring new contacts and for the relationships among alters. At the network level, immigrants’ networks were remarkably stable in composition and structure despite the high turnover. Clustered graphs have been used to illustrate the results. The results are discussed in light of adaptation to the host society.