Communities in Networks

Mason A. Porter, Jukka-Pekka Onnela, and Peter J. Mucha

Introduction: Networks and Communities

But although, as a matter of history, statistical
mechanics owes its origin to investigations in
thermodynamics, it seems eminently worthy of
an independent development, both on account of
the elegance and simplicity of its principles, and
because it yields new results and places old truths
in a new light in departments quite outside of
thermodynamics.
Josiah Willard Gibbs,
Elementary Principles in Statistical Mechanics,
1902 [47]

rom an abstract perspective, the term

network is used as a synonym for a mathematical graph. However, to scientists
across a variety of fields, this label means
so much more [13, 20, 44, 83, 88, 120, 124].
In sociology, each node (or vertex) of a network
represents an agent, and a pair of nodes can be
connected by a link (or edge) that signifies some
social interaction or tie between them (see Figure 1
Mason A. Porter, Oxford Centre for Industrial and Applied Mathematics, Mathematical Institute, University of
Oxford, and CABDyN Complexity Centre, University of
Oxford. His email address is [email protected].
Jukka-Pekka Onnela, Harvard Kennedy School, Harvard
University; Department of Physics, University of Oxford;
CABDyN Complexity Centre, University of Oxford; and
Department of Biomedical Engineering and Computational Science, Helsinki University of Technology. His email
address is [email protected].
Peter J. Mucha, Carolina Center for Interdisciplinary
Applied Mathematics, Department of Mathematics,
University of North Carolina, and Institute for Advanced
Materials, Nanoscience and Technology, University of
North Carolina. His email address is [email protected].

1082

for an example). Each node has a degree given by

the number of edges connected to it and a strength
given by the total weight of those edges. Graphs
can represent either man-made or natural constructs, such as the World Wide Web or neuronal
synaptic networks in the brain. Agents in such
networked systems are like particles in traditional
statistical mechanics that we all know and (presumably) love, and the structure of interactions
between agents reflects the microscopic rules that
govern their behavior. The simplest types of links
are binary pairwise connections, in which one only
cares about the presence or absence of a tie. However, in many situations, links can also be assigned
a direction and a (positive or negative) weight to
designate different interaction strengths.
Traditional statistical physics is concerned with
the dynamics of ensembles of interacting and
noninteracting particles. Rather than tracking the
motion of all of the particles simultaneously, which
is an impossible task due to their tremendous number, one averages (in some appropriate manner)
the microscopic rules that govern the dynamics
of individual particles to make precise statements
of macroscopic observables such as temperature
and density [112]. It is also sometimes possible
to make comments about intermediate (mesoscopic) structures, which lie between the microscopic
and macroscopic worlds; they are large enough
that it is reasonable to discuss their collective
properties but small enough so that those properties are obtained through averaging over smaller
numbers of constituent items. One can similarly
take a collection of interacting agents, such as
the nodes of a network, with some set of microscopic interaction rules and attempt to derive the
resulting mesoscopic and macroscopic structures.

Notices of the AMS

Volume 56, Number 9

Figure 1. The largest connected component of

a network of network scientists. This network
was constructed based on the coauthorship of
papers listed in two well-known review
articles [13, 83] and a small number of
additional papers that were added
manually [86]. Each node is colored according
to community membership, which was
determined using a leading-eigenvector
spectral method followed by Kernighan-Lin
node-swapping steps [64, 86, 107]. To
determine community placement, we used the
Fruchterman-Reingold graph
visualization [45], a force-directed layout
method that is related to maximizing a quality
function known as modularity [92]. To apply
this method, we treated the communities as if
they were themselves the nodes of a
(significantly smaller) network with
connections rescaled by inter-community
links. We then used the Kamada-Kawaii
spring-embedding graph visualization
algorithm [62] to place the nodes of each
individual community (ignoring
inter-community links) and then to rotate and
flip the communities for optimal placement
(including inter-community links). See the
main text for further details on some of the
ideas in this caption. (We gratefully
acknowledge Amanda Traud for preparing this
figure.)

October 2009

One mesoscopic structure, called a community,

consists of a group of nodes that are relatively
densely connected to each other but sparsely connected to other dense groups in the network [39].
We illustrate this idea in Figure 2 using a wellknown benchmark network from the sociology
literature [131].
The existence of social communities is intuitively clear, and the grouping patterns of humans
have been studied for a long time in both sociology [25, 44, 79] and social anthropology [66, 113].
For example, Stuart Rice clustered data by hand
to investigate political blocs in the 1920s [106],
and George Homans illustrated the usefulness
of rearranging the rows and columns of data
matrices to reveal their underlying structure in
1950 [60]. Robert Weiss and Eugene Jacobson performed (using organizational data) what might
have been the first analysis of network community structure in 1955 [126], and Herbert Simon
espoused surprisingly modern views on community structure and complex systems in general in the
1960s [117]. Social communities are ubiquitous,
arising in the flocking of animals and in social organizations in every type of human society:
groups of hunter-gatherers, feudal structures, royal families, political and business organizations,
families, villages, cities, states, nations, continents,
and even virtual communities such as Facebook
groups [39, 88]. Indeed, the concept of community is one of everyday familiarity. We are all
connected to relatives, friends, colleagues, and
acquaintances who are in turn connected to each
other in groups of different sizes and cohesions.
The goals of studying social communities have
aligned unknowingly with the statistical physics
paradigm. As sociologist Mark Granovetter wrote
in his seminal 1973 paper [51] on weak ties, Largescale statistical, as well as qualitative, studies offer
a good deal of insight into such macro phenomena
as social mobility, community organization, and
political structure... But how interaction in small
groups aggregates to form large-scale patterns
eludes us in most cases.
Sociologists recognized early that they needed powerful mathematical tools and large-scale
data manipulation to address this challenging
problem. An important step was taken in 2002,
when Michelle Girvan and Mark Newman brought
graph-partitioning problems to the broader attention of the statistical physics and mathematics
communities [48]. Suddenly, community detection in networks became hip among physicists
and applied mathematicians all over the world,
and numerous new methods were developed to
try to attack this problem. The amount of research
in this area has become massive over the past
seven years (with new discussions or algorithms
posted on the arXiv preprint server almost every
day), and the study of what has become known

Notices of the AMS

1083

11

29

32

28

17

26

25

24

34

33

31

12

30

13

27
14

23

10

16

15

20

19

21

18

22

Figure 2. (Top) The Zachary Karate Club

network [131], visualized using the
Fruchterman-Reingold method [45]. Nodes are
colored black or white depending on their
later club affiliation (after a disagreement
prompted the organizations breakup). The
dashed lines separate different communities,
which were determined using a
leading-eigenvector spectral maximization of
modularity [86] with subsequent
Kernighan-Lin node-swapping steps (see the
discussion in the main text). (Bottom) Polar
coordinate dendrogram representing the
results of applying this community-detection
algorithm to the network. Nodes are grouped
into the communities indicated in the top
panel. One can see the initial split of the
network into two groups (identical to the
actual membership of the new clubs) by
moving outward from the center of the ring.
Moving farther outward culminates in the final
partition of the network into four
communities.

as community structure is now one of the most

prominent areas of network science [26, 39, 110].
Although a rigorous notion of community structure remains elusive, one can clarify some issues
through the notions of modules and hierarchies. In
general, a networks community structure encompasses a potentially complicated set of hierarchical
and modular components [39, 48, 117]. In this

1084

context, the term module is typically used to refer to a single cluster of nodes. Given a network
that has been partitioned into nonoverlapping
modules in some fashion (although some methods also allow for overlapping communities), one
can continue dividing each module in an iterative
fashion until each node is in its own singleton
community. This hierarchical partitioning process
can then be represented by a tree, or dendrogram (see Figure 2). Such processes can yield
a hierarchy of nested modules (see Figure 3),
or a collection of modules at one mesoscopic
level might be obtained in an algorithm independently from those at another level. However
obtained, the community structure of a network
refers to the set of graph partitions obtained at
each reasonable step of such procedures. Note
that community structure investigations rely implicitly on using connected network components.
(We will assume such connectedness in our discussion of community-detection algorithms below.)
Community detection can be applied individually
to separate components of networks that are not
connected.
Many real-world networks possess a natural hierarchy. For example, the committee assignment
network of the U.S. House of Representatives includes the House floor, groups of committees, committees, groups of subcommittees within larger
committees, and individual subcommittees [100,
101]. As shown in Figure 4, different House committees are resolved into distinct modules within
this network. At a different hierarchical level, small
groups of committees belong to larger but less
densely connected modules. To give an example
closer to home, lets consider the departmental
organization at a university and suppose that
the network in Figure 3 represents collaborations
among professors. (It actually represents grassland species interactions [23].) At one level of
inspection, everybody in the mathematics department might show up in one community, such
as the large one in the upper left. Zooming in,
however, reveals smaller communities that might
represent the departments subfields.
Although network community structure is almost always fairly complicated, several forms of
it have nonetheless been observed and shown
to be insightful in applications. The structures
of communities and between communities are
important for the demographic identification of
network components and the function of dynamical processes that operate on networks (such as the
spread of opinions and diseases) [39]. A community in a social network might indicate a circle of
friends, a community in the World Wide Web might
indicate a group of pages on closely related topics,
and a community in a cellular or genetic network
might be related to a functional module. In some

Notices of the AMS

Volume 56, Number 9

APPROPRIATIONS
BUDGET

ENERGY/COMMERCE

VETERANS AFFAIRS
INTELLIGENCE

ARMED SERVICES
AGRICULTURE

HOUSE ADMINISTRATION
OFFICIAL CONDUCT
RULES

TRANSPORTATION

HOMELAND SECURITY
SMALL BUSINESS

RESOURCES

EDUCATION

GOVERNMENT REFORM
WAYS AND MEANS
SCIENCE
INTERNATIONAL RELATIONS
JUDICIARY

FINANCIAL SERVICES

APPROPRIATIONS

Figure 3. A network with both hierarchical and

modular structure. This image, courtesy of
Aaron Clauset, is an adaptation of a figure
from [23].

ENERGY/COMMERCE
ARMED SERVICES
BUDGET
INTELLIGENCE
VETERANS AFFAIRS

RULES

TRANSPORTATION

OFFICIAL CONDUCT
HOUSE ADMINISTRATIONAGRICULTURE
HOMELAND SECURITY
SMALL BUSINESS

cases, a network can contain several identical replicas of small communities known as motifs [75].
Consider a transcription network that controls
gene expression in bacteria or yeast. The nodes
represent genes or operons, and the edges represent direct transcriptional regulation. A simple
motif called a feed-forward loop has been shown
both theoretically and experimentally to perform
signal-processing tasks such as pulse generation.
Naturally, the situation becomes much more complicated in the case of people (doesnt it always?).
However, monitoring electronically recorded behavioral data, such as mobile phone calls, allows
one to study underlying social structures [49, 95].
Although these pairwise interactions (phone calls)
are short in duration, they are able to uncover
social groups that are persistent over time [97].
One interesting empirical finding, hypothesized by
Granovetter [51], is that links within communities
tend to be strong and links between them tend
to be weak [95]. This structural configuration has
important consequences for information flow in
social systems [95] and thus affects how the underlying network channels the circulation of social
and cultural resources. (See below for additional
discussion.)
With methods and algorithms drawn from statistical physics, computer science, discrete mathematics, nonlinear dynamics, sociology, and other
subjects, the investigation of network community
structure (and more general forms of data clustering) has captured the attention of a diverse group
of scientists [39, 54, 88, 110]. This breadth of interest has arisen partly because the development

October 2009

RESOURCES

GOVERNMENT REFORM

EDUCATION
WAYS AND MEANS
SCIENCE
INTERNATIONAL RELATIONS

JUDICIARY

FINANCIAL SERVICES

Figure 4. (Top) The network of committees

(squares) and subcommittees (circles) in the
108th U.S. House of Representatives
(200304), color-coded by the parent standing
and select committees and visualized using
the Kamada-Kawai method [62]. The darkness
of each weighted edge between
committees/subcommittees indicates how
strongly they are connected. Observe that
subcommittees of the same parent committee
are closely connected to each other. (Bottom)
Coarse-grained plot of the communities in this
network. Here one can see some close
connections between different committees,
such as Veterans Affairs/ Transportation and
Rules/Homeland Security.

of community-detection methods is an inherently

interdisciplinary endeavor and partly because interpreting the structure of a community and its
function often requires application-specific knowledge. In fact, one aspect that makes the problem
of detecting communities so challenging is that
the preferred formulation of communities is often domain-specific. Moreover, after choosing a
formulation, one still has to construct the desired

Notices of the AMS

1085

1086

communities by solving an optimization problem that is known to be NP-complete in at least

one family of formulations [17]. This has necessitated the adaptation of classical computationaloptimization heuristics and the development of
new ones.

Karate Club memberships) by drawing a pie chart

for each community, indicating the composition
of node characteristics in that community, and
showing the strength of inter-community connections as ties between the pies (as in Figure 4 for
Congressional committees).

A Simple Example

Identifying Communities

To set the stage for our survey of communitydetection algorithms below, consider the ubiquitous but illustrative example of the Zachary Karate
Club, in which an internal dispute led to the schism
of a karate club into two smaller clubs [131]. We
show a visualization of the friendships between
members of the original club in Figure 2. When the
club split in two, its members chose preferentially
to be in the one with most of their friends. Sociologist Wayne Zachary, who was already studying the
clubs friendships when the schism occurred, realized that he might have been able to predict the
split in advance. This makes the Zachary Karate
Club a useful benchmark for community-detection
algorithms, as one expects any algorithmically produced division of the network into communities
to include groups that are similar to the actual
memberships of the two smaller clubs.
In Figure 2, we show the communities that we obtained using a spectral-partitioning optimization
of a quality function known as modularity [86].
(This method is described below.) Keeping in mind
the hierarchical organization that often occurs as
part of network community structure, we visualize
the identified divisions using a polar coordinate
dendrogram and enumerate the networks nodes
around its exterior. Each distinct radius of the
dendrogram corresponds to a partition of the
original graph into multiple groups. That is, the
community assignments at a selected level of the
dendrogram are indicated by a radial cut in the
bottom panel of Figure 2; one keeps only connections (of nodes to groups) that occur outside this
cut. The success of the community identification
is apparent in the Zachary Karate Club example, as
the two main branches in the dendrogram reflect
the actual memberships of the new clubs.
As shown in Figure 2, this community-detection
method subsequently splits each of the two main
branches. Hence, we see that the Zachary Karate
Club network has a natural hierarchy of decompositions: a coarse pair of communities that correspond precisely to the observed membership
split and a finer partition into four communities. In larger networks, for which algorithmic
methods of investigation are especially important,
the presence of multiple such partitions indicates
network structures at different mesoscopic resolution levels. At each level, one can easily compare
the set of communities with identifying characteristics of the nodes (e.g., the post-split Zachary

Intuitively, a community is a cohesive group of

nodes that are connected more densely" to each
other than to the nodes in other communities. The
differences between many community-detection
methods ultimately come down to the precise
definition of more densely" and the algorithmic
heuristic employed to identify such sets. As different scientific fields have different needs, it is
not surprising that a wide variety of communitydetection methods have been developed to serve
those needs [39]. These differing needs have also
resulted in the deployment of different real and
computer-generated benchmark networks to test
community-finding algorithms [39, 70]. A 2005 review article [26] compared the performance of
several of the (then-)available methods in terms of
both computation time and output. A thorough,
more recent discussion is available in [39].
Rather than attempting a similar comparison
using every available algorithm, our aim is instead to expose a larger readership to many of
the most popular methods (as well as a few of
our personal favorites), while contrasting their
different perspectives and revealing a few important (and sometimes surprising) similarities.
Although we will highlight an extensive suite
of techniques in our survey below, there are of
course numerous other methodsincluding ones
based on maximum likelihood [23], mathematical
programming [1], block modeling [105, 127], link
partitioning [2, 33], inference [56, 91], latent space
clustering [55], and morethat we unfortunately neglect here because of space considerations.
Many of them are discussed in other review articles
[26, 39, 110].
Traditional Clustering Techniques
The idea of organizing data by coarse graining according to common features is a very old
one [39,119]. The original computational attempts
to find clusters of similar objects are rooted
in statistics and data mining. Important methods include partitional clustering techniques such
as k-means clustering, neural network clustering
techniques such as self-organizing maps, and
multidimensional scaling (MDS) techniques such
as singular value decomposition (SVD) and principal component analysis (PCA) [46]. For example,
MDS algorithms of various levels of sophistication have proven to be amazingly successful at
finding clusters of similar data points in myriad

Notices of the AMS

Volume 56, Number 9

applications, such as voting patterns of legislators

and Supreme Court justices [99101, 118]. Such
techniques start with a matrix that indicates similarities (e.g., a tabulation of how every legislator
voted on every bill) and return a coordinate matrix that minimizes an appropriate loss function.
In the U.S. Congress, this allows one to see that
the most important dimensions correspond to the
liberal-conservative axis (partisanship") and how
well a given legislator plays with others (bipartisanship"). During periods of heightened racial
tension, such analyses have also revealed a third
dimension corresponding to the division between
North and South [99].
Another prominent set of classical techniques
to detect cohesive groups in graphs are hierarchical
clustering algorithms such as the linkage clustering methods used in phylogenetic biology [46, 61].
One starts with the complete set of N individual
nodes in a weighted network, represented by an
adjacency matrix A whose elements (links) Aij
indicate how closely nodes i and j are related to
each other. For the purpose of our presentation,
we will only consider undirected networks, which
implies that A is symmetric (a few algorithms can
also handle directed networks [39, 53, 72]). Linkage clustering is an example of an agglomerative
method, as it starts from individual nodes and
ultimately connects the entire graph. The nodes
are conjoined sequentially into larger clusters,
starting with the pair with maximal Aij (i.e., the
most strongly connected pair). At each step, one
recomputes the similarities between the new cluster and each of the old clusters and again joins
the two maximally similar clusters, and one continues iteratively until all clusters with nonzero
similarity are connected. Different linkage clustering methods utilize different measures of the
similarity between clusters. For instance, in single
linkage clustering, the similarity of two clusters
X and Y is defined as the greatest similarity between any pair of nodes x X and y Y . Joining
nodes using single linkage clustering essentially
mirrors Joseph Kruskals algorithm for computing minimum spanning trees (MSTs) [31, 50]. With
clustering, however, the order of cluster formation is important and can be represented as a
dendrogram, whose depths indicate the steps at
which two clusters have been combined. More
sophisticated techniques that build on these ideas
are discussed in [46].
There are also a few classical divisive techniques, in which one starts with the full graph
and breaks it up to find communities [39, 46, 110].
(As with agglomerative techniques, one can visualize the results using dendrograms.) The most
prominent examples are spectral methods, which
we discuss in detail below. New data clustering
methods, which are applicable both to networks
and to more general data structures, continue

October 2009

to be developed very actively [46, 110]. Scientists

studying community detection and those studying
data clustering are obviously looking at the same
coin. The two fields are advancing in parallel, and
there are numerous deep connections between
them (including, we suspect, far more than are
already known).
The Kernighan-Lin Algorithm
An algorithm from computer science, which can
be combined with other methods, was proposed
by Brian Kernighan and Shen Lin (KL) in 1970 in
their study of how to partition electric circuits into
boards so that the nodes in different boards can
be linked to each other using the fewest number
of connections [64]. To do this, they maximized a
that relates the number of edges
quality function Q
inside each group of nodes to the number between
different groups. Starting with an initial partition
of a graph into two groups of predefined size, KL
steps swap subsets containing equal numbers of
vertices between the two groups. To reduce the
chance of getting stuck at a local maximum, the
After
KL method permits swaps that decrease Q.
a specified number of swaps, one keeps the parti to use as the initial condition
tion with maximal Q
for a new set of KL steps. When the number and
sizes of communities are not specified, a natural
generalization of the KL method is to move a single
node at a time [12, 86, 87, 107]. Unsurprisingly, the
partitions of networks into communities that are
obtained using the KL algorithm depend strongly
on ones initial partition and, therefore, it is best
used as a supplement to high-quality partitions
obtained using other methods [39, 86, 87]. In typical situations, both the KL swaps and the other

method would seek to optimize the same Q.

Centrality-Based Community Detection
Michelle Girvan and Mark Newman generated a
great deal of attention in mathematics and statistical physics for network community structure
in [48] when they devised a community-finding
algorithm based on the sociological notion of betweenness centrality [4, 42, 124]. An edge has high
betweenness if it lies on a large number of short
paths between vertices. (Note that betweenness
can also be defined for nodes.) If one starts at a
node and wants to go to some other node in the
network, it is clear that some edges will experience
a lot more traffic than others. The betweenness
of an edge quantifies such traffic by considering
strictly shortest paths (geodesic betweenness) or
densities of random walks (random walk betweenness) [85] between each pair of nodes and taking
into account all possible pairs. One can identify
communities through a process of ranking all of
the edges based on their betweenness, removing
the edge with the largest value, and recalculating

Notices of the AMS

1087

1088

betweenness for the remaining edges. The recalculation step is important because the removal of
an edge can cause a previously low-traffic edge to
have much higher traffic. An iterative implementation of these steps gives a divisive algorithm for
detecting community structure, as it deconstructs
the initial graph into progressively smaller connected chunks until one obtains a set of isolated
nodes.
Betweenness-based methods have been generalized to use network components other than
edges, to bipartite networks [100], and to use
other sociological notions of centrality [39]. Note,
however, that although centrality-based community detection is intuitively appealing, it can be
too slow for many large networks (unless they are
very sparse), and it tends to give relatively poor
results for dense networks.

that they have to date provided some of the most

successful ways to consider community overlap.
Allowing the detection of network communities that overlap is especially appealing in the
social sciences, as people belong simultaneously to
several communities (constructed via colleagues,
family, hobbies, etc.) [78,79]. Purely agglomerative
or divisive techniques do not allow communities to overlap, so it is important to consider
local methods as well. Several such methods have
now been developed [2, 7, 8, 14, 22, 28, 33, 39, 74,
114, 129], including one that enables the consideration of overlapping communities at multiple
resolution levels [69]. We believe that further development of global clustering algorithms that
take community overlap explicitly into account is
essential to complement the insights from these
local approaches.

k-Clique Percolation and Other Local Methods

Modularity Optimization

The method of k-clique percolation [97] is based

on the concept of a k-clique, which is a complete
subgraph of k nodes that are connected with all
k(k 1)/2 possible links. The method relies on
the observation that communities seem to consist
of several small cliques that share many of their
nodes with other cliques in the same community. A
k-clique community is then defined as the union of
all adjacent k-cliques, which by definition share
k 1 nodes. One can also think about rolling a
k-clique template from any k-clique in the graph
to any adjacent k-clique by relocating one of its
nodes and keeping the other k 1 nodes fixed [28].
A community, defined through the percolation of
such a template, then consists of the union of all
subgraphs that can be fully explored by rolling a
k-clique template. As k becomes larger, the notion
of a community becomes more stringent. Values of
k = 3, . . . , 6 tend to be most appropriate because
larger values become unwieldy. The special case
of k = 2 reduces to bond (link) percolation and
k = 1 reduces to site (node) percolation.
The k-clique percolation algorithm is an example of a local community-finding method. One
obtains a networks global community structure
by considering the ensemble of communities obtained by looping over all of its k-cliques. Some
nodes might not belong to any community (because they are never part of any clique), and
others can belong to several communities (if they
are located at the interface between two or more
communities). The nested nature of communities
is recovered by considering different values of k,
although k-clique percolation can be too rigid because focusing on cliques typically causes one to
overlook other dense modules that arent quite as
well connected. On the other hand, the advantage
of k-clique percolation and other local methods is

One of the most popular quality functions is modularity, which attempts to measure how well a
given partition of a network compartmentalizes
its communities [84, 86, 87, 89, 90]. The problem
of optimizing modularity is equivalent to an instance of the famous MAX-CUT problem [86], so
it is not surprising that it has been proven to
be NP-complete [17]. There are now numerous
community-finding algorithms that try to optimize modularity or similarly constructed quality
functions in various ways [6, 12, 26, 39]. In the
original definition of modularity, an unweighted
and undirected network that has been partitioned
into communities has modularity [84, 90]
X
(1)
Q = (eii bi2 ),
i

where eij denotes the fraction of ends of edges in

group i for which the
P other end of the edge lies in
group j, and bi = j eij is the fraction of all ends
of edges that lie in group i. Modularity is closely
related to the Freeman segregation index [43]; a
key difference is that Q = 0 when all nodes are
assigned to the same community, which enforces
the existence of a nontrivial partition with Q > 0.
Modularity explicitly takes degree heterogeneity
into account, as it measures the difference between
the total fraction of edges that fall within groups
versus the fraction one would expect if edges were
placed at random (respecting vertex degrees).1
Thus, high values of Q indicate network partitions
in which more of the edges fall within groups than
expected by chance (under a specified null model,
as discussed further below). This, in turn, has been
1

Interest in degree heterogeneity exploded in the late

1990s with the sudden wealth of empirical data and
the seemingly ubiquitous manifestation of heavy-tailed
degree distributions such as power laws [3, 83].

Notices of the AMS

Volume 56, Number 9

found to be a good indicator of functional network

divisions in many cases [86, 87].
For weighted networks, one counts the sums
of the weights of edges instead of the number
of edges, so heavily weighted edges contribute
more than lightly weighted ones. Both eij and bi
are thus straightforwardly generalized, and then
the modularity is again calculated from (1). The
meaning of modularity remains essentially the
same: It measures when a particular division of
the network has more edge weight within groups
than one would expect by chance.
Quality functions such as modularity provide
precise measures of how to count the total strength
of connections within communities versus those
between communities [87, 89]. Modularity is a
scaled assortativity measure based on whether
high-strength edges are more or less likely to be
adjacent to other high-strength edges [84, 89, 90].
Because communities are supposed to have high
edge density relative to other parts of the graph,
a high-modularity partition tends to have high
edge-strength assortativity by construction. More
generally, assortativity notions can be used to
partition a graph into groups according to any
characteristic by examining whether nodes are
more likely (in assortative graphs) or less likely (in
disassortative graphs) to be connected to nodes of
the same type [83].
Interestingly, maximizing modularity is closely
related to the energy models of pairwise attraction,
such as the Fruchterman-Reingold method, that
are commonly used for graph visualization [92].
While this isnt necessarily surprising given the
clusters that one can typically observe with good
graph visualization tools, this recent insight does
suggest that such tools may also help lead to the
development of better community-detection methods. Conversely, the analysis and construction of
algorithms to find network communities might
help lead to better graph-visualization techniques.
It is typically impossible computationally to
sample a desired quality function by exhaustively enumerating the nonpolynomial number
of possible partitions of a network into communities [17]. A number of different methods
have thus been proposed to balance the typical quality of their identified optima with the
computational costs. Some methods, such as the
greedy algorithms in [24, 84], are fast heuristics
intended to be applied to networks with millions of nodes or more. Other methodssuch
as spectral partitioning [86, 87] (discussed below), refined greedy algorithms [111], simulated
annealing [52], extremal optimization [30], and
others [93]provide more sophisticated but slow-

October 2009

er means to identify high-modularity partitions.2

We discuss the spectral partitioning method below,
in part because of its interesting reformulation of
the modularity scalar as a matrix, but we note
that other algorithmic choices might be superior in many situations. We believe that there is
considerable value in having multiple computational heuristics available, as this provides greater
flexibility to compare and contrast the identified
communities.
Importantly, many modularity-maximization
techniques are easy to generalize for use with
other related quality functions because it is far
from clear that modularity is the best function
to optimize. For example, modularity has a
known resolution limit (see below) that might
cause one to miss important communities [38].
A few alternatives to modularity have been
considered [6, 39, 57, 69, 121], and it is ultimately
desirable to optimize a quality function that
includes not only information about the network
structure but also other information (such as
node characteristics or relevant time-dependence)
that would allow one to incorporate functionality
directly [116]. Such consideration of additional
information is one of the most important open
issues in community detection [88, 122].
Spectral Partitioning
The method of spectral partitioning arose most
prominently in the development of algorithms for
parallel computation [35, 102]. In traditional spectral partitioning, network properties are related to
the spectrum of the graphs Laplacian matrix L,
which has components Lij = ki (i, j) Aij , where
ki is the degree of node i (or, in a weighted network, its strength), and (i, j) is the Kronecker
delta (i.e., (i, j) = 1 if i = j and 0 otherwise).
The simplest such method starts by splitting a
network into two components. One then applies
two-group partitioning recursively to the smaller
networks one obtains as long as it is desirable to
do so. (One can also partition networks into more
than two groups during each step [21,29,86,107].)
For a single partitioning step, one defines an index
vector s whose components take the value +1 if
they belong to group 1 and 1 if they belong
to group 2. The total edge weight between the
two groups of nodes can then be expressed as
1
R = 4 s T Ls. The best partition of the network
seemingly results from choosing s to minimize R
(called the cut size") and hence the total strength
of inter-group edges. (Recall the max-flow min-cut
theorem, which states that the minimum cut between any two vertices of a graphthat is, the
minimum set of edges whose deletion places the
2

As we have discussed, one can also supplement any of

these methods with KL swapping steps [12, 64, 86, 87, 93,
107].

Notices of the AMS

1089

two vertices in disconnected components of the

graphcarries the maximum flow between the two
vertices [32,37].) Unfortunately, this minimization
is easily accomplished by choosing the trivial (and
useless) partition of a single group containing
every node. The most common solution to this
situation is to fix the sizes of the two groups in
advance and incorporate this information in the
partitioning procedure (as described in, e.g., [86]).
This solution is perfectly reasonable for some
applications, such as load balancing in parallel
computing. However, this approach is neither appropriate nor realistic for community detection in
most other contexts because one typically does
not know the number or sizes of communities in
advance, so choosing arbitrary sizes at the outset
precludes attacking the main problem of interest.
Fortunately, one can use the idea of modularity
to obtain spectral partitioning algorithms that are
appropriate for a broader class of problems [87]
(see also the earlier publication [29] and a spiritually similar approach based on peer influences
in the sociology literature [77]). By reformulating
the scalar quantity of modularity in terms of a
modularity matrix B, with components
(2)

Bij = Aij Pij ,

spectral partitioning can be directly applied [87] as

a means of heuristically optimizing the modularity
1 X
(3)
Q=
Bij (Ci , Cj ) ,
2W i,j
where (Ci , Cj ) indicates that the Bij components
are only summed over cases in which nodes i
and j are assigned
to the same community. The
P
factor W = 12 ij Aij is the total edge strength in
the network (equal to the total number of edges
for unweighted networks), where ki denotes the
strength of node i. In (2), Pij denotes the components of a null model matrix, which specifies
the relative value of intra-community edges in
assessing when communities are tightly connected [9, 86]. In general, one is free to specify any
reasonable null model. The most popular choice,
proposed by Newman and Girvan [86,87,89,90], is
ki kj
.
2W
This recovers the definition of modularity in (1),
specified in terms of edge-weight deviations from
a network chosen randomly from the set of all
graphs with the same expected strength distribution as the actual network. This null model is
closely related to the configuration model [76],
which (as with Erds-Renyi random graphs) yields
networks that arent expected to have a natural
hierarchy [76, 83, 90]. The difference is that (4) is
conditioned on the expected degree (or strength)
sequence, whereas the configuration model is
conditioned on the actual observed sequence.

(4)

1090

Pij =

In spectral partitioning, one can use as many

eigenvectors of B as there are positive eigenvalues, but it is effective (and simplest) to recursively
subdivide a network using only the leading eigenvector v, which is paired with the largest positive
eigenvalue of B. One can then separate the network into two communities according to the signs
si = sgn(vi ). The magnitude of vi indicates the
strength to which the ith node belongs to its
assigned community [86]. For vi = 0, one can
assign node i to a community based on which
choice produces the higher modularity, changing
si = 0 to +1 or 1 as appropriate. The modularity of the resulting two-group partition of the
1 T
network is Q = 4W
s Bs. After this bipartition,
one then repeats the partitioning procedure for
each graph component, keeping track of the fact
that they are actually part of a larger network.
One continues recursively until the modularity
can no longer be increased with additional subdivisions [86, 87]. The final network partition gives
the community structure at a specific resolution
level (e.g., committees in the U.S. House of Representatives committee assignment network). This
method can be generalized by considering different quality functions [39, 107], allowing steps
that decrease global quality in order to further
subdivide the communities [107, 133], using more
eigenvectors [107, 123], or including a resolution
parameter [6, 104] that allows one to examine
the networks community structure at different
mesoscopic scales.
The Potts Method
Particles that possess a magnetic moment are
often called spins [112, 130]. Such spins interact
with other spins either ferromagnetically (they
seek to align) or antiferromagnetically (they seek
to have different orientations). A spin glass is
a system that encompasses both disorder and
competing ferromagnetic and antiferromagnetic
interactions. This leads to a very large number
of metastable spin configurations separated by
energy barriers with long, glass-like characteristic
relaxation times [36, 130]. An important recent
insight, inspired by earlier work on data clustering
based on the physical properties of an inhomogeneous ferromagnetic model [11], is that optimizing
modularity is mathematically equivalent to minimizing the energy (i.e., finding the ground state of
the Hamiltonian) of an infinite-range q-state Potts
model [103, 104].
In a q-state Potts spin glass, each spin can be
in one of q states. The interaction energy between
spins i and j is given by Jij if the spins are in
the same state and zero if they are not [104, 130].
The Hamiltonian of the system is given by the sum

Notices of the AMS

Volume 56, Number 9

into equations like (6) to obtain [104]

over all of the pairwise interaction energies:

X
(5)
H({ }) = Jij (i , j ) ,
ij

where l indicates the state of spin l and { }

denotes the configuration of spins (i.e., the state
of each of the systems N spins). There are a total
of q N such configurations.
We map the problem of minimizing (5) to a network community-detection problem by assigning
a spin to each node and letting q = N. In this
language, one adds the interaction energy Jij if
and only if nodes i and j are placed in the same
community. Two nodes that are connected by an
edge interact ferromagnetically (Jij > 0) when the
weight of the edge is greater than expected (given
a specific null model) and interact antiferromagnetically (Jij < 0) when it is less than expected. If
Jij = 0, spins i and j do not interact with each
other. Hence, two nodes want to be in the same
community if they interact ferromagnetically and
in different ones if they interact antiferromagnetically. One cannot in general find any arrangement
of spins (i.e., any partition of a graph into communities) that simultaneously minimizes all of the
pairwise interaction energies. Regardless of this
inability to satisfy all of the bonds simultaneously,
a phenomenon termed frustration [36], one can
still try to minimize H globally to find the ground
state of the system. The choice of interaction
matrix elements given by
(6)

Jij =

Aij Pij
W

implies that H = Q and recovers modularity maximization [104]. (Division by W is a normalization

and does not affect the optimization algorithms.)
Alternative interaction models can also be used to
partition networks (see, e.g., [57]).
Resolution Parameters
In 2007, Santo Fortunato and Marc Barthlemy
demonstrated, using both real and computergenerated networks, that modularity suffers from
a resolution limit in its original formulation [38],
as it misses communities that are smaller than a
certain threshold size that depends on the size of
the network and the extent of interconnectedness
of its communities. Communities smaller than
the threshold tend to be merged into larger communities, thereby missing important structures.
We have seen this in our own work on the U.S.
House committee assignment network, as detecting communities by maximizing modularity can
group multiple standing committees (with their
subcommittees) into a single community [100].
One way to address this resolution limit is to incorporate an explicit resolution parameter directly

October 2009

Aij Pij
.
W
One can alternatively incorporate a resolution parameter into Aij or elsewhere in the definition
of a quality function (see, e.g., [6]). This allows
one to zoom in and out in order to find communities of different sizes and thereby explore
both the modular and the hierarchical structures
of a graph. Fixing in (7) corresponds to setting the scale at which one is examining the
network: Larger values of yield smaller communities (and vice versa). Resolution parameters
have now been incorporated (both explicitly and
implicitly) into several methods that use modularity [12, 104], other quality functions [6, 69], and
other perspectives [8, 97].
Although introducing a resolution parameter
using equations like (7) seems ad hoc at first, it
can yield very interesting insights. For example,
Jij = (Aij )/W gives a uniform null model in
which a given fixed average edge weight occurs
between each node. This can be useful for correlation and similarity networks, such as those
produced from matrices of yea and nay votes.
Nodes i and j want to be in the same community
if and only if they voted the same way more than
some threshold fraction of times that is specified
by the value of .
Even more exciting, one can relate resolution
parameters to the time scales of dynamical processes unfolding on a network [27,68,98,108]. Just
as we can learn about the behavior of a dynamical
system by studying the structural properties of
the network on which it is occurring, we can also
learn about the networks structural properties
by studying the behavior of a given dynamical
process. This suggests the intuitive result that the
choice of quality function should also be guided by
the nature of the dynamical process of interest. In
addition to revealing that resolution parameters
arise naturally, this perspective shows that the
Potts method arises as a special case of placing
a continuous time random walk with Poissondistributed steps on a network [27]. Freezing the
dynamics at a particular point in time yields the
modularity-maximizing partition. Freezing at earlier times yields smaller communities (because
the random walker hasnt explored as much of
the graph), and waiting until later times results in
larger communities. The t limit reproduces
the partitioning from Miroslav Fiedlers original
spectral method [35].

(7)

Jij =

Applications
Armed with the above ideas and algorithms, we
turn to selected demonstrations of their efficacy.
The increasing rapidity of developments in network community detection has resulted in part

Notices of the AMS

1091

at the selection of topics and citations in this section.) In this spirit, we use scientific coauthorship
networks as our first example.
A bipartite (two-mode) coauthorship network
with scientists linked to papers that they authored
p
or coauthoredcan be defined by letting i = 1
if scientist i was a coauthor on paper p and
0 otherwise. Such a network was collected and
examined from different databases of research
papers in [8082]. To represent the collaboration
strength between scientists i and j, one can define
(8)

Aij =

X pi pj
p

Figure 5. The largest connected component

(379 nodes) of the network of network
scientists (1,589 total nodes), determined by
coauthorship of papers listed in two
well-known review articles [13, 83] and a small
number of papers added manually [86]. Each
of the nodes in the network, which we depict
using a Kamada-Kawai visualization [62], is
colored according to its community
assignment using the leading-eigenvector
spectral method [86].

from the ever-increasing abundance of data sets

(and the ability to extract them, with user cleverness). This newfound wealthwhich includes
large, time-dependent data setshas, in turn,
arisen from the massive amount of information
that is now routinely collected on websites and by
communication companies, governmental agencies, and others. Electronic databases now provide
detailed records of human activity patterns, offering novel avenues to map and explore the
structure of social, communication, and collaboration networks. Biologists also have extensive
data on numerous systems that can be cast into network form and which beg for additional
quantitative analyses.
Because of space limitations, we restrict our
discussion to five example applications in which
community detection has played a prominent role:
scientific coauthorship, mobile phone communication, online social networking sites, biological
systems, and legislatures. We make no attempt
to be exhaustive for any of these examples; we
merely survey research (both by others and by
ourselves) that we particularly like.
Scientific Collaboration Networks
We know from the obsessive computation of Erds
numbers that scientists can be quite narcissistic.
(If you want any further evidence, just take a look

1092

np 1

as the components of a weighted unipartite (onemode) network, where np is the number of authors
of paper p and the sum runs over multiple-author
papers only. Applying betweenness-based community detection to a network derived from Santa Fe
Institute working papers using (8) yields communities that correspond to different disciplines [48].
The statistical physics community can then be further subdivided into three smaller modules that
are each centered around the research interests
of one dominant member. Similar results have
been found using various community-finding algorithms and numerous coauthorship networks,
such as the network of network scientists [86]
(see Figures 1 and 5), which has become one of
the standard benchmark examples in communitydetection papers.
Mobile Phone Networks
Several recent papers have attempted to uncover
the large-scale characteristics of communication
and social networks using mobile phone data
sets [49, 95, 96]. Like many of the coauthorship
data sets studied recently [96], mobile phone networks are longitudinal (time-dependent). However,
in contrast to the ties in the coauthorship networks
above, links in phone networks arise from instant
communication events and capture relationships
as they happen. This means that at any given instant, the network consists of the collection of ties
connecting the people who are currently having a
phone conversation. To probe longer-term social
structures, one needs to aggregate the data over a
time window.
In 2007 one research group used a society-wide
communication network containing the mobile
phone interaction patterns of millions of individuals in an anonymous European country to explore
the relationship of microscopic, mesoscopic, and
macroscopic network structures to the strength
of ties between individuals on a societal level [95].
They obtained very interesting insights into Mark
Granovetters famous weak tie hypothesis, which
states that the relative overlap of the friendship circles of two individuals increases with the strength

Notices of the AMS

Volume 56, Number 9

of the tie connecting them [51]. At the mesoscopic

level, this leads to a structure in which individuals
within communities tend to be linked via strong
ties, whereas communities tend to be connected
to other communities via weak ties. Because of
this coupling between link strength and function,
the weak ties are responsible for the structural
integrity of the communication network: It is robust to the removal of the strong ties but breaks
up if the weak ties are removed (see Figure 6). In
fact, one can even show that the removal of weak
ties leads to a (phase) transition from a regime in
which the network remains globally connected to
one in which the network essentially consists of
insular communities. However, there is no phase
transition if the strong ties are removed, so the
network remains globally connected. The location
of the transition also suggests a natural quantitative demarcation between weak and strong ties.
This mesoscopic organization of social networks
has important consequences for the flow of information. If one assumes that every tie (regardless
of strength) is equally efficient in transferring
information, one recovers the classical result of
Granovetter that weak ties are mostly responsible for information diffusion [51]. However, if
one assumes that the rate of information transfer
is proportional to the strength of the tie, then
neither weak nor strong ties are as effective as
intermediate ties for information diffusion [95].
To help develop methods that can be applied to
time-dependent networks, another research group
has recently applied k-clique percolation to a large
mobile phone data set to investigate community
formation, evolution, and destruction [96]. They
found that large communities persist longer if
they are capable of dynamically altering their
membership (suggesting that an ability to change
the group composition results in better adaptability), whereas small groups last longer if they stay
virtually unchanged. Weve all seen examples of
such dynamics in real life: A mathematics department at a university will last a long time and will
still be identified as fundamentally the same community even though its membership can change
quite drastically over time. On the other hand, an
individual research group might rely on only one
or two of its members for its existence.
Online Social Networks
Social networking sites (SNSs) have become a pervasive part of everyday life. They allow users to
construct a public or semi-public online profile
within a bounded system, articulate a list of other
users (called friends) with whom they share a
connection, and view and traverse their network
of connections [16, 122]. Since their introduction,
SNSs such as Facebook, LinkedIn, MySpace, and
hundreds of others have collectively attracted over

October 2009

Figure 6. (Left) A sample of a mobile phone

network studied in [94, 95]. After the strong
ties have been removed, the network still
retains its global connectivity. (Right) Removal
of weak ties leads, through a phase transition,
to a disintegration of the network. (Figure
adapted from [94].)

one billion users [128]. People have integrated SNSs

into their daily lives, using them to communicate
with friends, send emails, solicit opinions or votes,
organize events, spread ideas, find jobs, and more.
The emergence of SNSs has also revolutionized
the availability of quantitative social and demographic data, which has in turn transformed the
study of social networks [16]. This has obviously been a big deal for social scientists (see [73]
for an excellent recent example), but numerous
mathematicians, computer scientists, physicists,
and more have also had a lot of fun with this new
wealth of data. This has included investigations
of attachment mechanisms to determine how SNS
network structure might develop [67], the formation of friends, allies, and nemeses [18, 59], and
much more [16, 122].
In a recent paper [122], we used anonymous
Facebook data from September 2005 to compare
the friendship networks of several American universities. This data yields networks in which each
node represents a person and each edge represents
a self-identified, reciprocal online friendship. The
institutions we considered ranged from small technical universities such as Caltech (1,089 nodes) to
large state universities such as the University of
Oklahoma (about 24,000 nodes). Our data also includes limited demographic information for each
user: gender, high school, class year, major, and
dormitory residence. Using permutation tests, we
developed tools that allow one to quantitatively
compare different network partitions, which can
be obtained from any desired methodincluding,
in particular, community-detection algorithms and
user demographics. This enables one to investigate the demographic organization of different
universities and compare the results of different
community-detection algorithms. We found, for

Notices of the AMS

1093

opens up significant privacy issues when data is

not fully anonymous.
Biological Networks

Figure 7. (Top) Caltech friendship network

based on Facebook data from September 2005
using the Fruchterman-Reingold visualization
method [45]. The colors (and shapes)
correspond to self-identified House
(dormitory) affiliation. (Bottom) Communities
in the Caltech network. We obtained this
community structure, which accurately
matches the Caltech House structure, using a
slight modification of Newmans
leading-eigenvector algorithm [86] and
subsequent KL node-swapping
steps [107, 122, 133]. It also gives some
indication of the most probable House
affiliations of the people in white, who did not
identify their House [122].

example, that communities at Princeton, Georgetown, and the University of North Carolina at
Chapel Hill are organized predominantly by class
year, whereas those at Caltech are based almost
exclusively on House (dormitory) affiliation. As
we illustrate in Figure 7, community structure can
also be used to make simple yet intelligent guesses
about withheld user characteristics. Naturally, this

1094

One of the paramount goals of studying biological

networks is to determine the principles governing
their evolution. It is hoped that finding important network structures might give some insights
into the mechanisms (and, ideally, the natural
design principles) that control the flow of biological information. Thus, it is unsurprising that
clustering methods form a core part of bioinformatics [15, 63], and there is simply no way to do
justice to this vast literature here. Accordingly, we
only present a couple of our favorite examples.
In 2002 Ron Milo et al. investigated a plethora
of directed networks to develop the idea of miniature communities known as motifs [75], which are
used to describe patterns of nodes and edges that
occur significantly more prevalently than expected
in a network. Motifs can be interpreted as basic
building blocks of complex networks, perhaps representing small functional modules that arose via
evolutionary mechanisms. The simplest motifs
triangles (3-cliques), in which three nodes are all
mutually interconnectedhave long been studied
in social networks [44,124]. The amazing discovery
of Milo et al. is that different types of motifs are, in
fact, prevalent universally in many different types
of networks. Among the omnipresent motifs they
observed are 3-chains in ecological networks (in
which a prey node is connected by a directed edge
to a predator node, which is in turn connected by
a directed edge to another predator); diamonds
in ecological networks, neural networks, and logic
chips; feed-forward loops in gene regulation networks, neural networks, and logic chips; triangles
in the World Wide Web; and more.
Numerous scientists have built on this foundation of motifs, and several investigations have
provided fascinating connections between motifs
and larger mesoscopic structures. For example,
one team investigated three-node and four-node
motifs in an integrated Saccharomyces cerevisiae network, which they constructed using information from protein interactions, genetic interactions, transcriptional regulation, sequence
homology, and expression correlation [132]. Their
primary finding was that most motifs form larger network themes" of recurring interconnection
patterns that incorporate multiple motif occurrences. They were also able to tie some of these
mesoscopic themes to specific biological phenomena, such as a pair of protein complexes with many
genetic interactions between them. The notion of
motifs has also recently been used to develop
generalizations of graph modularity [5].
One can imagine constructing a course-grained
network consisting of interconnected network

Notices of the AMS

Volume 56, Number 9

themes. For example, in a seminal 2005 paper,

Roger Guimer and Lus Amaral used communities
to try to construct a functional cartography
of biological networks in order to employ nodes
and modules with known roles to try to obtain
interesting insights on nodes and modules with
unknown functionality [52]. To understand their
perspective, consider the prototypical maps of
countries in which important cities are marked by
special symbols, other cities are marked with dots,
and smaller towns are not shown. In the network
context, there is a one-to-one correspondence
between markers and communities, and the
symbols are determined according to structural
and/or functional roles. The connections between
communities are analogous to major highways. In
fact, our coloring of community ties according to
the demographic composition of their nodes (see
Figures 4 and 7) was originally inspired by [52].
To illustrate their idea, Guimer and Amaral considered the metabolic networks of twelve
different organisms. They started by detecting
communities by maximizing modularity using simulated annealing and then (following a suggestion
from social scientist Brian Uzzi) calculated appropriate properties of each node to determine their
role in their assigned community. After finding
communities, they calculated for each node i the
within-module degree, given by the number of its
edges that connect to other nodes in the same
community, and a participation ratio Pi , which
measures the extent to which links are distributed among all network communities. Guimer and
Amaral then interpreted the role of each node
based on its location in the plane determined by
Pi and the z-score zi of the within-module degree. They thereby found that non-hub connector
nodes (which have low zi and moderately high Pi ,
indicating a preferential connectivity to a subset
of the networks communities) are systematically more conserved across species than provincial
hubs (which have high zi and low Pi ). This appears
to be related to the role of non-hub connectors in
describing the global structure of fluxes between
different network modules. Importantly, one can
follow a similar procedure using other measures,
such as betweenness centrality [34], as the essential insightwhich, we stress, was borrowed from
ideas in the social sciencesis to calculate network
quantities relative to community assignment.
Legislative Networks
Advances in network science have also begun to
uncover the ways in which social relationships
shape political outcomes [40, 41, 101]. In this section, we describe our own work on legislative
networks [100,101,125,133], in which community
detection has played a central role.

October 2009

Consider a bipartite graph that is constructed

using Representatives and their committee and
subcommittee (henceforth called simply committee") assignments during a single two-year
term of the U.S. House of Representatives. Each
edge represents a committee assignment and connects a Representative to a committee. We project
each such network onto a weighted unipartite
graph of committees (see Figure 4), in which the
nodes are now committees and the value of each
edge gives the normalized connection strength
between two committees. By computing the community structure of these networks and analyzing
legislator ideology, we investigated correlations
between the political and organizational structure
of House committees. This revealed close ties between the House Rules Committee and the Select
Committee on Homeland Security in the 107th
(200102) and 108th (200304) Congresses that
broke the established procedures for determining
the composition of select committees [100, 101].
(Figure 4 shows the 108th Congress.) We also
showed that the modularity of good network partitions increased following the 1994 elections, in
which the Republican party earned majority status
in the House for the first time in more than forty
years.
Studying networks constructed from legislation
cosponsorship can help uncover social connections between politicians, as legislators who work
together closely on pieces of legislation are likely to have friendly (or at least cordial) relations.
Computing centrality measures in these networks
gives a whos who list of American politics, as
it reveals important players like Bob Dole [R-KA],
John McCain [R-AZ], and Ted Kennedy [D-MA] [40].
The longitudinal study of community structure in
Congressional legislation cosponsorship [133] and
roll-call voting [125] networks shows that graph
modularity can be used to study partisan polarization and political party realignments. This reveals
patterns suggesting that political parties were not
the most significant communities in Congress for
certain periods of U.S. history and that the 1994
party-changing elections followed a rise in partisan
polarization rather than themselves leading to an
abrupt polarization in America.

Summary and Outlook

With origins in sociology, computer science, statistics, and other disciplines, the study of network
communities is in some respects quite old. Nevertheless, it has experienced incredible growth since
the seminal 2002 paper [48] that brought greater
attention to the problem, particularly among statistical physicists [39]. In this survey, we have
highlighted an extensive suite of techniques, and
there are numerous other methods that we simply

Notices of the AMS

1095

havent had space to discuss (see the review articles [26, 39, 110] for more information on many of
them). Despite this wealth of technical advances,
much work remains. As Mark Newman recently wrote [88], The development of methods for
finding communities within networks is a thriving
sub-area of the field, with an enormous number of different techniques under development.
Methods for understanding what the communities
mean after you find them are, by contrast, still
quite primitive, and much needs to be done if
we are to gain real knowledge from the output
of our computer programs. One of our primary
purposes in writing this article is as a call to
arms for the mathematics community to be a
part of this exciting endeavor. Accordingly, we
close our discussion with additional comments
about important unresolved issues.
The remarkable advances of the past few years
have been driven largely by a massive influx
of data. Many of the fascinating networks that
have been constructed using such data are enormous (with millions of nodes or more). Given
that optimization procedures, such as maximizing graph modularity, have been proven to be
NP-complete [17], much of the research drive has
been to formulate fast methods that still find
a reasonable community structure. Some of the
existing algorithms scale well enough to be used
on large networks, whereas others must be restricted to smaller ones. The wealth of data has
also led to an increasing prevalence (and, we hope,
cognizance) of privacy issues. However, although
the study of network communities has become so
prominent, this research area has serious flaws
from both theoretical and applied perspectives:
There are almost no theorems, and few methods
have been developed to use or even validate the
communities that we find.
We hope that some of the mathematicallyminded Notices readers will be sufficiently excited
by network community detection to contribute
by developing new methods that address important graph features and make existing techniques
more rigorous. When analyzing networks constructed from real-world data, the best practice
right now is to use several of the available computationally tractable algorithms and trust only
those structures that are similar across multiple
methods in order to be confident that they are
properties of the actual data rather than byproducts of the algorithms used to produce them.
Numerous heuristics and analytical arguments
are available, but there arent any theorems, and
even the notion of community structure is itself
based on the methodology selected to compute
it. There also appear to be deep but uncharacterized connections between methods that have
been developed in different fields [39, 110]. Additionally, it would be wonderful if there were a

1096

clearer understanding of which notions of community and which community-detection methods

might be especially appropriate for networks with
specific properties and for networks belonging to
a specific domain.
At the same time, the problem of how to validate
and use communities once they are identified is
almost completely open. Fortunately, recent work
offers some hope, as new methods have been
developed to infer the existence of missing edges
from network data [23] and relate the composition
of communities to intelligent guesses about the
demographic characteristics of nodes [122]. (As
with social networks more generally, sociologists
have already been considering these issues for a
long time [19, 44]. What we need are techniques
that allow us to do this even more effectively.)
In [23], Aaron Clauset, Cris Moore, and Mark Newman drew on the insight that real-world networks
should not be expected to have a unique community structure (despite the output produced by
almost all of the available methods) and formulated a new hierarchical random graph model to
develop a method for inferring hierarchical structure from network data. (A different hierarchical
random graph model was formulated for community detection in [109].) Their method, which
shows excellent promise for future development,
allowed them to make accurate predictions about
missing connections in partially known networks.
In our own work on Facebook networks [122], we
used permutation tests to advance methods for
the quantitative comparison of network partitions.
Because one can obtain such partitions either from
algorithmic community-detection methods or by
hand (from external demographics or other properties), this provides a mechanism to compare the
results of different community-finding algorithms
and to try to infer node characteristics given
partial demographic information.
It is also important to develop communitydetection techniques that can be applied to more
complicated types of graphs. As we saw in our discussion of legislative and coauthorship networks,
collaboration networks have a bipartite structure.
However, there has been seemingly only limited
work thus far on community finding that works
directly on bipartite networks rather than on their
unipartite projections [10, 53, 71, 100]. Even fewer
community-detection methods are able to handle
directed networks (whose adjacency matrices are
asymmetric) [53, 72] or signed networks (whose
connections might be construed as attracting
or repelling) [121]. Moreover, agents in social
networks are typically connected in several different mannersfor example, Representatives can
be connected using voting similarities, common
committee assignments, common financial contributors, and morebut there are presently very

Notices of the AMS

Volume 56, Number 9

few algorithms that can be applied to such multiplex situations without constructing individual
graphs for each category, and further development will likely require the application of ideas
from multilinear algebra [65,115]. It would also be
desirable to detect communities in hypergraphs
and to be able to consider connections between
agents that are given by interval ranges rather
than precise values. Finally, to be able to study
interactions between dynamical processes on networks and the structural dynamics of networks
themselves (e.g., if somebody spends a day at
home with the flu, the network structure in the
workplace is different than usual that day), a lot
more work is needed on both overlap between
communities and on the community structure
of time-dependent and parameter-dependent networks. Analyzing time- and parameter-dependent
networks currently relies on ad hoc amalgamation
of different snapshots rather than on a systematic
approach, so it is necessary to develop communitydetection methods that incorporate the network
structure at multiple times (or parameter values)
simultaneously [58, 96, 116]. More generally, this
will also have important ramifications for studies
of clustering in correlated time series.
We stress that research on network communities has focused on using exclusively structural information (i.e., node connectivity and link
weights) to deduce structural communities as imperfect proxies for functional communities [39,
116, 122]. While this seems to be sufficient for
some applications [39], in most situations it is
not at all clear that structural communities actually map well to the organization of actors in
social networks, functions in biological networks,
etc. Hence, it is necessary to develop tools for the
detection of functional communities that, whenever possible, incorporate node characteristics and
other available information along with the networks structural information. The elephant in the
literature is simply elucidated with just one question: Now that we have all these ways of detecting
communities, what do we do with them?

Acknowledgements
Our views on network community structure have
been shaped by numerous discussions with our
colleagues and students over the last several years.
We particularly acknowledge Aaron Clauset, Santo
Fortunato, Nick Jones, Eric Kelsic, Mark Newman, Stephen Reid, and Chris Wiggins. We also
thank Joe Blitzstein, Tim Elling, Santo Fortunato, James Fowler, A. J. Friend, Roger Guimer,
Nick Jones, David Kempe, Franziska Klingner, Renaud Lambiotte, David Lazer, Sune Lehmann, Jim
Moody, and David Smith for useful comments on
this manuscript, and Christina Frost and Amanda

October 2009

Traud for assistance in preparing some of the figures. We obtained data from Adam DAngelo and
Facebook, the House of Representatives Office of
the Clerk (Congressional committee assignments),
Mark Newman (network scientist coauthorship),
James Fowler (Congressional legislation cosponsorship), and Keith Poole (Congressional roll call
votes). PJM was funded by the NSF (DMS-0645369)
and by start-up funds provided by the Institute
for Advanced Materials, Nanoscience and Technology and the Department of Mathematics at
the University of North Carolina at Chapel Hill.
MAP acknowledges a research award (#220020177)
from the James S. McDonnell Foundation. JPO is
supported by the Fulbright Program.

References
[1] G. Agarwal and D. Kempe, Modularitymaximizing network communities using mathematical programming, The European Physical
Journal B, 66 (2008), 409418.
[2] Y.-Y. Ahn, J. P. Bagrow, and S. Lehmann, Communities and hierarchical organization of links in
complex networks. arXiv:0903.3178 (2009).
[3] R. Albert and A.-L. Barabsi, Statistical mechanics of complex networks, Reviews of Modern
Physics, 74 (2002), 4797.
[4] J. M. Anthonisse, The rush in a graph. Technical Report BN 9/71, Stichting Mathematische
Centrum, Amsterdam, 1971.
[5] A. Arenas, A. Fernndez, S. Fortunato, and
S. Gmez, Motif-based communities in complex networks, Journal of Physics A: Mathematical and
Theoretical, 41, 224001 (2008).
[6] A. Arenas, A. Fernndez, and S. Gmez, Analysis
of the structure of complex networks at different resolution levels, New Journal of Physics, 10,
053039 (2008).
[7] J. P. Bagrow, Evaluating local community methods in networks, Journal of Statistical Mechanics:
Theory and Experiment, P05001 (2008).
[8] J. P. Bagrow and E. M. Bollt, A local method
for detecting communities, Physical Review E, 72,
046108 (2005).
[9] S. Bansal, S. Khandelwal, and L. A. Meyers,
Evolving clustered random networks. arXiv:
0808.0509 (2008).
[10] M. J. Barber, Modularity and community detection in bipartite networks, Physical Review E, 76,
066102 (2007).
[11] M. Blatt, S. Wiseman, and E. Domany, Superparamagnetic clustering of data, Physical Review
Letters, 76 (1996), 32513254.
[12] V. D. Blondel, J.-L. Guillaume, R. Lambiotte,
and E. Lefebvre, Fast unfolding of communities
in large network, Journal of Statistical Mechanics:
Theory and Experiment, P10008 (2008).
[13] S. Boccaletti, V. Latora, Y. Moreno, M. Chavez,
and D.-U. Hwang, Complex networks: Structure and dynamics, Physics Reports, 424 (2006),
175308.
References continued on page 1164.

Notices of the AMS

1097

References (Continued from page 1097)

[14] S. P. Borgatti, M. G. Everett, and P. R. Shirey, LS
Sets, Lambda Sets and other cohesive subsets, Social
Networks, 12 (1990), 337-357.
[15] J. M. Bower and H. Bolouri, eds., Computational
modeling of genetic and biochemical networks, The
MIT Press, Cambridge, Massachusetts, 2001.
[16] D. M. Boyd and N. B. Ellison, Social network
sites: Definition, history, and scholarship, Journal of
Computer-Mediated Communication, 13, 11 (2007).
[17] U. Brandes, D. Delling, M. Gaertler, R. Goerke,
M. Hoefer, Z. Nikoloski, and D. Wagner, On modularity clustering, IEEE Transactions on Knowledge
and Data Engineering, 20 (2008), 172188.
[18] M. Brzozowski, T. Hogg, and G. Szabo, Friends
and foes: Ideological social networking, in Proceedings of the SIGCHI Conference on Human Factors in
Computing, ACM Press, New York, NY, 2008.
[19] C. T. Butts, Network inference, error, and informant
(in)accuracy: a Bayesian approach, Social Networks,
25 (2003), 103140.
[20] G. Caldarelli, Scale-free networks: Complex webs
in nature and technology, Oxford University Press,
Oxford, 2007.
[21] A. Capocci, V. D. P. Servedio, G. Caldarelli,
and F. Colaiori, Detecting communities in large
networks, Physica A, 352 (2004), 669676.
[22] A. Clauset, Finding local community structure in
networks, Physical Review E, 72, 026132 (2005).
[23] A. Clauset, C. Moore, and M. E. J. Newman, Hierarchical structure and the prediction of missing links
in networks, Nature, 453 (2008), 98101.
[24] A. Clauset, M. E. J. Newman, and C. Moore, Finding community structure in very large networks,
Physical Review E, 70, 066111 (2004).
[25] J. S. Coleman, An introduction to mathematical
sociology, Collier-Macmillan, London, 1964.
[26] L. Danon, A. Diaz-Guilera, J. Duch, and A. Arenas, Comparing community structure identification,
Journal of Statistical Mechanics: Theory and Experiment, P09008 (2005).
[27] J.-C. Delvenne, S. N. Yaliraki, and M. Barahona,
Stability of graph communities across time scales.
arXiv:0812.1811 (2008).
[28] I. Dernyi, G. Palla, and T. Vicsek, Clique percolation in random networks, Physical Review Letters,
94, 160202 (2005).
[29] L. Donetti and M. A. Muoz, Detecting network
communities: A new systematic and efficient algorithm, Journal of Statistical Mechanics: Theory and
Experiment, P10012 (2004).
[30] J. Duch and A. Arenas, Community detection
in complex networks using extremal optimization,
Physical Review E, 72, 027104 (2005).
[31] J. Eisner, State-of-the-art algorithms for minimum
spanning trees: A tutorial discussion. Manuscript
available at http://cs.jhu.edu/~jason/papers/
#ms97 (1997).
[32] P. Elias, A. Feinstein, and C. E. Shannon, A
note on the maximum flow through a network, IRE
Transactions on Information Theory, 2, 4 (1956),
117119.
[33] T. S. Evans and R. Lambiotte, Line graphs, link
partitions and overlapping communities, Physical
Review E, 80, 016105 (2009).
[34] D. J. Fenn, M. A. Porter, M. McDonald,
S. Williams, N. F. Johnson, and N. S. Jones,

1164

[35]

[36]
[37]

[38]

[39]
[40]

[41]
[42]
[43]
[44]

[45]

[46]

[47]

[48]

[49]

[50]

[51]

[52]

[53]

[54]
[55]

[56]

Notices of the AMS

Dynamic communities in multichannel data: An

application to the foreign exchange market during
the 20072008 credit crisis, Chaos, 19:3, 033119
(2009).
M. Fiedler, Algebraic connectivity of graphs,
Czechoslovak Mathematical Journal, 23 (1973),
289305.
K. H. Fischer and J. A. Hertz, Spin glasses,
Cambridge University Press, Cambridge, 1993.
L. R. Ford and D. R. Fulkerson, Maximal
flow through a network, Canadian Journal of
Mathematics, 8 (1956), 399404.
S. Fortunato and M. Barthlemy, Resolution limit
in community detection, Proceedings of the National
Academy of Sciences, 104 (2007), 3641.
S. Fortunato, Community detection in graphs.
arXiv:0906.0612 (2009).
J. H. Fowler, Legislative cosponsorship networks
in the U.S. House and Senate, Social Networks, 28
(2006), 456487.
J. H. Fowler and S. Jeon, The authority of Supreme
Court precedent, Social Networks, 30 (2008), 1630.
L. C. Freeman, A set of measures of centrality based
upon betweenness, Sociometry, 40 (1977), 3541.
, Segregation in social networks, Sociological
Methods Research, 6 (1978), 411429.
, The development of social network analysis:
A study in the sociology of science, Empirical Press,
Vancouver, Canada, 2004.
T. M. J. Fruchterman and E. M. Reingold, Graph
drawing by force-directed placement, Software
Practice and Experience, 21 (1991), 11291164.
G. Gan, C. Ma, and J. Wu, Data clustering: Theory,
algorithms, and applications, Society for Industrial
and Applied Mathematics, Philadelphia, 2007.
J. W. Gibbs, Elementary principles in statistical mechanics, Charles Scribners Sons, New York, NY,
1902.
M. Girvan and M. E. J. Newman, Community structure in social and biological networks, Proceedings
of the National Academy of Sciences, 99 (2002),
78217826.
M. C. Gonzlez, C. A. Hidalgo, and A.-L.
Barabsi, Understanding individual human mobility patterns, Nature, 453 (2008), 779782.
J. C. Gower and G. J. S. Ross, Minimum spanning
trees and single linkage cluster analysis, Applied
Statistics, 18 (1969), 5464.
M. Granovetter, The strength of weak ties,
The American Journal of Sociology, 78 (1973),
13601380.
R. Guimer and L. A. N. Amaral, Functional cartography of complex metabolic networks, Nature, 433
(2005), 895900.
R. Guimer, M. Sales-Pardo, and L. A. N. Amaral, Module identification in bipartite and directed
networks, Physical Review E, 76, 036102 (2007).
N. Gulbahce and S. Lehmann, The art of
community detection, Bioessays, 30 (2008), 934938.
M. S. Handcock, A. E. Rafter, and J. M. Tantrum,
Model-based clustering for social networks, Journal of the Royal Statistical Society A, 170 (2007),
301354.
M. B. Hastings, Community detection as an inference problem, Physical Review E, 74, 035102
(2006).

Volume 56, Number 9

[57] J. M. Hofman and C. H. Wiggins, Bayesian approach to network modularity, Physical Review
Letters, 100, 258701 (2008).
[58] J. Hopcroft, O. Khan, B. Kulis, and B. Selman,
Tracking evolving communities in large linked networks, Proceedings of the National Academy of
Sciences, 101 (2004), 52495253.
[59] T. Hogg, D. Wilkinson, G. Szabo, and M. Brzozowski, Multiple relationship types in online
communities and social networks, in Proceedings of
the AAAI Spring Symposium on Social Information
Processing, AAAI Press (2008).
[60] G. C. Homans, The human group, Harcourt, Brace,
New York, 1950.
[61] S. C. Johnson, Hierarchical clustering schemes,
Psychometrica, 32 (1967), 241254.
[62] T. Kamada and S. Kawai, An algorithm for drawing
general undirected graphs, Information Processing
Letters, 31 (1989), 715.
[63] F. Kps, ed., Biological networks, Complex Systems and Interdisciplinary Science, Vol. 3, World
Scientific, Hackensack, NJ, 2007.
[64] B. W. Kernighan and S. Lin, An efficient heuristic
procedure for partitioning graphs, The Bell System
Technical Journal, 49 (1970), 291307.
[65] T. G. Kolda and B. W. Bader, Tensor decompositions and applications, SIAM Review, 51:3 (2009), in
press.
[66] C. P. Kottak, Cultural anthropology, McGraw-Hill,
New York, 5th ed., 1991.
[67] R. Kumar, J. Novak, and A. Tomkins, Structure
and evolution of online social networks. 12th International Conference on Knowledge Discovery and Data
Mining (2006).
[68] R. Lambiotte, J.-C. Delvenne, and M. Barahona, Dynamics and modular structure in networks.
arXiv:0812.1770 (2008).
[69] A. Lancichinetti, S. Fortunato, and J. Kertesz,
Detecting the overlapping and hierarchical community structure of complex networks, New Journal of
Physics, 11, 033015 (2009).
[70] A. Lancichinetti, S. Fortunato, and F. Radicchi,
Benchmark graphs for testing community detection
algorithms, Physical Review E, 93, 046110 (2008).
[71] S. Lehmann, M. Schwartz, and L. K. Hansen, Biclique communities, Physical Review E, 78, 016108
(2008).
[72] E. A. Leicht and M. E. J. Newman, Community structure in directed networks, Physical Review Letters,
100, 118703 (2008).
[73] K. Lewis, J. Kaufman, M. Gonzalez, M. Wimmer,
and N. A. Christakis, Tastes, ties, and time: A
new (cultural, multiplex, and longitudinal) social network dataset using Facebook.com, Social Networks,
30 (2008), 330342.
[74] R. Luce and A. Perry, A method of matrix analysis of group structure, Psychometrika, 14 (1949),
95116.
[75] R. Milo, S. Shen-Orr, S. Itzkovitz, N. Kashtan,
D. Chklovskii, and U. Alon, Network motifs: Simple building blocks of complex networks, Science, 298
(2002), 824827.
[76] M. Molloy and B. Reed, A critical point for random graphs with a given degree sequence, Random
Structures and Algorithms, 6 (1995), 161-180.

October 2009

[77] J. Moody, Peer influence groups: Identifying dense

clusters in large networks, Social Networks, 23
(2001), 261283.
[78] J. Moody and D. R. White, The cohesiveness of
blocks in social networks, Sociological Methodology,
31 (2001), 305359.
, Structural cohesion and embeddedness: A
[79]
hierarchical concept of social groups, American
Sociological Review, 68 (2003), 103127.
[80] M. E. J. Newman, Scientific collaboration networks:
I. Network construction and fundamental results,
Physical Review E, 64, 016131 (2001).
[81]
, Scientific collaboration networks: II. Shortest
paths, weighted networks, and centrality, Physical
Review E, 64, 016132 (2001).
, The structure of scientific collaborations, Pro[82]
ceedings of the National Academy of Sciences, 98
(2001), 404409.
[83]
, The structure and function of complex
networks, SIAM Review, 45 (2003), 167256.
, Fast algorithm for detecting community
[84]
structure in networks, Physical Review E, 69, 066133
(2004).
[85]
, A measure of betweeenness centrality based
on random walks, Social Networks, 27 (2005), 3954.
, Finding community structure in networks us[86]
ing the eigenvectors of matrices, Physical Review E,
74, 036104 (2006).
[87]
, Modularity and community structure in networks, Proceedings of the National Academy of
Sciences, 103 (2006), 85778582.
, The physics of networks, Physics Today,
[88]
61:11 (2008), pp. 3338.
[89] M. E. J. Newman and M. Girvan, Mixing patterns
and community structure in networks, in Statistical
Mechanics of Complex Networks (R. Pastor-Satorras,
J. Rubi, and A. Diaz-Guilera, eds.), Springer-Verlag,
Berlin, 2003.
[90]
, Finding and evaluating community structure
in networks, Physical Review E, 69, 026113 (2004).
[91] M. E. J. Newman and E. A. Leicht, Mixture models
and exploratory analysis in networks, Proceedings
of the National Academy of Sciences, 104 (2007),
95649569.
[92] A. Noack, Modularity clustering is force-directed
layout, Physical Review E, 79, 026102 (2009).
[93] A. Noack and R. Rotta, Multi-level algorithms
for modularity clustering, in Experimental Algorithms: 8th International Symposium, SEA 2009,
Dortmund, Germany, June 46, 2009, Proceedings,
Springer-Verlag, New York, 2009, pp. 257268.
[94] J.-P. Onnela, J. Saramki, J. Hyvnen, G. Szab,
M. Argollo de Menezes, K. Kaski, A.-L. Barabsi,
and J. Kertsz, Analysis of a large-scale weighted
network of one-to-one human communication, New
Journal of Physics, 9 (2007), 179.
[95] J.-P. Onnela, J. Saramki, J. Hyvnen, G. Szab,
D. Lazer, K. Kaski, J. Kertsz, and A.-L. Barabsi,
Structure and tie strengths in mobile communication
networks, Proceedings of the National Academy of
Sciences, 104 (2007), 73327336.
[96] G. Palla, A.-L. Barabsi, and T. Vicsek, Quantifying social group evolution, Nature, 446 (2007),
664667.
[97] G. Palla, I. Dernyi, I. Farkas, and T. Vicsek,
Uncovering the overlapping community structure of

Notices of the AMS

1165

[98]

[99]

[100]

[101]

[102]

[103]

[104]
[105]

[106]

[107]

[108]

[109]

[110]
[111]

[112]
[113]
[114]

[115]

1166

complex networks in nature and society, Nature, 435

(2005), 814818.
P. Pons and M. Latapy, Computing communities
in large networks using random walks, Journal
of Graph Algorithms and Applications, 10 (2006),
191218.
K. T. Poole and H. Rosenthal, Congress: A
political-economic history of roll call voting, Oxford
University Press, Oxford, 1997.
M. A. Porter, P. J. Mucha, M. E. J. Newman, and A. J.
Friend, Community structure in the United States
House of Representatives, Physica A, 386 (2007),
414438.
M. A. Porter, P. J. Mucha, M. E. J. Newman, and
C. M. Warmbrand, A network analysis of committees in the United States House of Representatives,
Proceedings of the National Academy of Sciences,
102 (2005), 70577062.
A. Pothen, H. D. Simon, and K.-P. Liou, Partitioning
sparse matrices with eigenvectors of graphs, SIAM
Journal on Matrix Analysis and Applications, 11
(1990), 430452.
J. Reichardt and S. Bornholdt, Detecting fuzzy
community structures in complex networks with a
Potts model, Physical Review Letters, 93, 218701
(2004).
, Statistical mechanics of community detection, Physical Review E, 74, 016110 (2006).
J. Reichardt and D. R. White, Role models for
complex networks, European Physical Journal B, 60
(2007), 217224.
S. A. Rice, The identification of blocs in small political
bodies, American Political Science Review, 21 (1927),
619627.
T. Richardson, P. J. Mucha, and M. A. Porter,
Spectral tripartitioning of networks, Physical Review
Letters E, in press (2009), arXiv:0812.2852.
M. Rosvall and C. T. Bergstrom, Maps of random walks on complex networks reveal community
structure, Proceedings of the National Academy of
Sciences, 105 (2008), 11181123.
M. Sales-Pardo, R. Guimer, A. A. Moreira, and
L. A. N. Amaral, Extracting the hierarchical organization of complex systems, Proceedings of the
National Academy of Sciences, 104, 39 (2007),
1522415229.
S. E. Schaeffer, Graph clustering, Computer Science
Review, 1 (2007), 2764.
P. Schuetz and A. Caflisch, Efficient modularity
optimization by multistep greedy algorithm and vertex mover refinement, Physical Review E, 77, 046112
(2008).
F. Schwabl, Statistical mechanics, Springer-Verlag,
New York, NY, 2002.
R. Scupin, Cultural anthropologya global perspective, Prentice-Hall, Englewood Cliffs, NJ, 1992.
S. Seidman and B. Foster, A graph theoretic
generalization of the clique concept, Journal of
Mathematical Sociology, 6 (1978), 139154.
T. M. Selee, T. G. Kolda, W. P. Kegelmeyer, and
J. D. Griffin, Extracting clusters from large datasets
with multiple similarity measures using IMSCAND, in
CSRI Summer Proceedings 2007, Technical Report
SAND2007-7977, Sandia National Laboratories, Albuquerque, NM and Livermore, CA (M. L. Parks and
S. S. Collis, eds.), 2007, pp. 87103.

[116] C. R. Shalizi, M. F. Camperi, and K. L. Klinkner, Discovering functional communities in dynamical

networks, in Statistical Network Analysis: Models,
Issues, and New Directions, Springer-Verlag, New
York, NY (E. M. Airoldi, D. M. Blei, S. E. Feinberg,
A. Goldenberg, E. P. Xing, and A. X. Zheng, eds.),
2007, pp. 140157.
[117] H. Simon, The architecture of complexity, Proceedings of the American Philosophical Society, 106
(1962), 467482.
[118] L. Sirovich, A pattern analysis of the second Renquist U.S. Supreme Court, Proceedings of
the National Academy of Sciences, 100 (2003),
74327437.
[119] P. B. Slater, Established clustering procedures for
network analysis. arXiv:0806.4168 (2008).
[120] S. H. Strogatz, Exploring complex networks, Nature,
410 (2001), 268276.
[121] V. A. Traag and J. Bruggeman, Community detection in networks with positive and negative links.
arXiv:0811.2329 (2008).
[122] A. L. Traud, E. D. Kelsic, P. J. Mucha, and M. A.
Porter, Community structure in online collegiate
social networks. arXiv:0809.0960 (2008).
[123] G. Wang, Y. Shen, and M. Ouyang, A vector partitioning approach to detecting community structure
in complex networks, Computers & Mathematics
with Applications, 55 (2007), 27462752.
[124] S. Wasserman and K. Faust, Social network analysis: Methods and applications, Structural Analysis
in the Social Sciences, Cambridge University Press,
Cambridge, 1994.
[125] A. S. Waugh, L. Pei, J. H. Fowler, P. J. Mucha,
and M. A. Porter, Party polarization in Congress: A
social networks approach. arXiv:0907.3509 (2009).
[126] R. S. Weiss and E. Jacobson, A method for the
analysis of the structure of complex organizations,
American Sociological Review, 20 (1955), 661668.
[127] H. C. White, S. A. Boorman, and R. L. Breiger,
Social structure from multiple networks. I. Blockmodels of roles and positions, The American Journal of
Sociology, 81, 4 (1976), 730780.
[128] Wikipedia, List of social networking websites.
http://en.wikipedia.org/wiki/List_of_
social_networking_websites,
accessed
13
February 2009.
[129] F. Wu and B. A. Huberman, Finding communities
in linear time: A physics approach, The European
Physical Journal B, 38 (2004), 331338.
[130] F. Y. Wu, The Potts model, Reviews of Modern
Physics, 54 (1982), 235268.
[131] W. W. Zachary, An information flow model for
conflict and fission in small groups, Journal of
Anthropological Research, 33 (1977), 452473.
[132] L. V. Zhang, O. D. King, S. L. Wong, D. S. Goldberg,
A. H. Y. Tong, G. Lesage, B. Andrews, H. Bussey,
C. Boone, and F. P. Roth, Motifs, themes and
thematic maps of an integrated Saccharomyces cerevisiae interaction network, Journal of Biology, 4, 6
(2005).
[133] Y. Zhang, A. J. Friend, A. L. Traud, M. A. Porter,
J. H. Fowler, and P. J. Mucha, Community structure
in Congressional cosponsorship networks, Physica A,
387 (2008), 17051712.

Notices of the AMS

Volume 56, Number 9