I don't know why physicists are interested in 

complex network theory, but well, whenever

you can create a physical model describing some behavior you could call it "physics"
5down
(econophysics, sociophysics, etc), and this is likely the reason why they study complex
voteaccept
network. I will just answer the 2nd part of the question — the concept of small-world
ed network.
Short average distance
The small-world phenomena is best known as the six degree of separation, i.e. two
persons are related to each other by at most 6 steps in a network of human relationship.
In this network, people are represented by nodes, and if two people knows each other
directly we create alink between the two nodes. If there are two nodes without a direct
link, we could possibly take other routes. The smallest number of links need to be
traversed is called the distance (a.k.a. path length) between the two nodes.
A small-world network (with size N) needs to satisfy a condition, that the average
distance ⟨ℓ⟩ over all possible node pairs is "short", i.e. ⟨ℓ⟩∼logN.
Many natural and artificial complex networks, like the human relationship network
above, neural network, metabolic networks etc. have short average distance, which is why
this property is being studied, because physicists wanted to have a model that describes
most of these complex networks. The Erdős-Rényi model — which is essentially a
network that the links are constructed with some constant probability — has the short
average distance property.
High clustering coefficient
The ER model was one of the first models used to describe real-world complex networks,
but was soon found to be insufficient because it doesn't have another important property
— a "high" clustering coefficient.
The (local) clustering coefficient was introduced by Duncan J. Watts and Steven Strogatz
in 1998 as a measure of how well nodes are "clustered" together locally. It is defined as
cv=number of triangles containing vkv(kv−1)2
where kv is the number of links connected to v, i.e. its degree. The local clustering
coefficient is actually a ratio of "neighbors of v that knows each other", and "maximum
number of potential neighbors of v that can know each other". For example, in this graph:

the clustering coefficient of the central green node is

c=number of trianglesk(k−1)2=45(5−1)2=0.4.
 If the local clustering coefficient is 0, then all your friends don't know each other, which
is not true in the social network. The expected behavior is a large clustering coefficient,
where groups of your friends know each other (thus forming triangles, i.e. A knows B, B
knows C, C knows A). This property is where the ER model breaks down — its
clustering coefficient is close to zero with the same number of nodes and links with a real
network.

When a network has both short average distance and high average local clustering
coefficient, we call it asmall-world network.
And beyond
The Watts-Strogatz model was invented to address the small-world property. However,
it was soon determined that even the WS model is not good enough as it is not scale-free.
And then the Barabási-Albert model was created to describe why real-world networks
both small-world and scale-free, although it also cannot explain other properties like the
clustering coefficient distribution, hierarchical structure, etc, and of course more and
more sophisticated models are proposed as well.
In the end, these properties are studied to construct a universal model that describes all (if
not possible, "most") real-world complex networks, and use it to test and improve
behaviors such as error and attack tolerance, evolution dynamics etc.

If you don't fear a lot of mathematics, the 2002 review paper Statistical mechanics of
complex networks by Réka Albert and Albert-László Barabási is a (IMO) must-read
classic for anyone beginning to study complex networks. All of the above can be found in
this paper.
shareimprove this answer edited Dec 13 '10 at 21:54 answered

Marek KennyTM
13.7k2682 1,691

+1 Although there is no physics in this I like the answer very much. Also, I took a liberty of fixing some typos
    link; I hope you don't mind. –  Marek Dec 13 '10 at 21:55

By the way, do you know whether there is any connection to things like k-SAT and graph coloring problems?
some of these problems (which seem to be purely in the domain of complexity theory) were studied by metho
close in spirit to statistical physics. In particular, random k-SAT problems can be shown to have different pha
characterized as "clustered" and "frozen" (among others). See e.g. this paper. Of course, this was made quite f
    recently because of an P≠NP attack of Deolalikar. –  Marek Dec 13 '10 at 22:00

    This is a very nice answer +1 –  user346 Dec 13 '10 at 22:50

@Marek: Thanks. I don't think it is related to algorithmic/combinatorics stuff like k-SAT and graph
    coloring. – KennyTM Dec 14 '10 at 15:02
add a comment
up Richard Feynman reformulated quantum mechanics (and quantum field theory) in terms of
7
vote down a path integral, meaning that in order to find the likelihood of some process occurring, you
 vote take a kind of weighted average over all potential trajectories. The weighting function is the
exponentiated "action," exp(iS/ℏ). and the dominant contribution comes from paths
which extremize this function, i.e. classical trajectories.
Typically -- almost always -- this integral is too hard for anyone to do (let alone define
rigorously), so Feynman developed a perturbation theory, an expansion in terms of graphs.
The nature of the graphs depends on the interactions and coupling constants of your model
-- that is, on the action.

An (oversimplified) example: Suppose you only had one degree of freedom, x, and the
action is S0(x)=iℏx2/2, so that exp(iS0/ℏ)=exp(−x2/2). (You can ignore ℏ in
this example.) Then the path integral is ∫exp(−x2/2)dx and equals π√. However, if
we add a cubic "interaction" term, so S=S0−iℏax3 then we can expand ∫exp(iS/ℏ) in
powers of a, the first nonzero contribution being a2∫(x3)2exp(−x2/2)dx, which you
can do exactly. (The term linear in a is zero because the three powers of x can't be paired
up ["contracted"].) The graph for this term has two vertices (the two powers of a), each
with three edges attached (the three powers of x in the interaction term).
So graphs are ubiquitous in QFT!

shareimprove this answer edited Dec 13 '10 at 16:33 answered

Eric Zaslow
1,736616
add a comment
up Not sure this is what you are aiming at, but graphs are ubiquitous in statistical physics!
vote4do They really crop up all over the place. To give you some ideas from the top of my head:
wn vote
Probability
Classical statistical physics is built on the concept of microstates. Each microstate has a certain
energy Egets a probability P=1Zexp(−βE) where β is a parameter depending on the
temperature of the system and
Z=∑microstatesexp(−βE)

is the partition function.
Once you make this assignment of probabilities you can ask for things like average energy of
the system and fluctuations and lots of other interesting stuff. And all of that stuff can be
computed easily if you can somehow carry out the summation and determine the partition
function. That is, the problem can often (at least in the discrete case) be reduced to counting,
which means combinatorics. And often the combinatorial problem has to do with graph
theory (either directly or by means of some clever duality).
Crystals
Statistical physics often investigates lattices. This is because they model crystals or other
ordered forms of matter. These are very special graphs that possess translational symmetries
(and more generally also some rotational and reflectional symmetries; think about hexagonal
lattice). Once again you can define energy for microstates as in the first example and proceed to
formulate the probabilistic problem. But it can be observed that lots of methods that work for
investigation of the system on the lattice can actually be generalized to arbitrary graphs.

Arbitrary graphs
Perhaps nicest connection (and one positively stunning if you haven't heard about it before) is
between correlation function of the Gaussian free field on the graph and random walks on the
same graph (see e.g.this recent blog on the topic). This is a discrete version of the Feynman path
integral which gives you probability amplitudes of particle getting from one place to another in
terms of summing over every path between the two points.
Polymer model
One more model I'd like to mention is the Polymer model. The idea is that you have some
objects, calledpolymers, that usually live on some kind of lattice (imagine e.g. cycles on the
edges of the hexagonal lattice). Now the requirement is that these objects do not occupy the
same space of the lattice (i.e. they don't intersect). This idea can be rigorously captured by the
means of a huge infinite-dimensional graph where vertices are all of the possible polymers with
edges between any two of them that are not compatible (that is, when they intersect on the
original lattice).
This looks like a hard problem but actually it can be investigated by the means of cluster
expansion. To give (an oversimplified) idea: it all stems from the simple but very useful
combinatorial identity
exp(∑ixi)=∑N1N!(∑ixi)N=∑N∑i1+⋯+iN=N∏kxikkik!

Using this one can transform the partition function Z (which is an ugly sum over all possible
subgraphs of the original huge graph) into an exponential of sum over just connected subgraphs
of the polymer graph. Now connected subgraphs are much nicer objects than arbitrary
subgraphs and one indeed obtains nice results by expanding the logarithm of the partition
function in this way.
shareimprove this answer edited Dec 13 '10 at 16:41 answered Dec 13 '10 at 16:34

Marek
13.7k2682

    If you care to add spectral graph theory and its role to that list, you'll get my vote. :) –  Noldorin Dec 13 '10 at 20:50

@Noldorin: I will have to think about it. Most of the uses of spectral graph theory I know of are pure mathematics o
computer science. But thanks to your comment I recalled other places graphs pop up: relation between Potts model a
number of graph colorings; relation between electrical resistance on the graph and its spanning trees; and also Marko
chains on graphs. Latter two cases can be studied by the means of some associated matrix. But it's not quite an
adjacency matrix and so not quite a spectral theory. If you have anything concrete in mind, please let me
    know :-) –  Marek Dec 13 '10 at 21:38
 @Marek: That is in fact true nowadays, but spectral graph theory (hinted by its name) actually originated in quantum
    chemistry... Not that I know much about it - was hoping you would know more heh. –  Noldorin Dec 13 '10 at 22:15

@Noldorin: ah, if the quantum chemistry origin is indeed correct, I had no idea about that at all! What I always
implicitly assumed is that spectral refers to the spectrum of the (adjacency) matrix. And this is quite general term
pertaining to many areas of mathematics (as in spectrum of some operator). I'll try to find out which version is the
    correct one. –  Marek Dec 13 '10 at 22:25 

@Marek: You could be right, though I do remember some relation with quantum chemistry.... –  Noldorin
    at 23:09
show 2 more comments
up One context in which graphs can be useful in physics is in the discrete representation of
2
vote do spacetime in quantum gravity, where events are represented by the nodes of a type of poset
wn vote (partially ordered set) called a causet and causal relationships are represented by the edges. This

is particularly suited to a graph-theoretic interpretation, since posets can be intuitively

visualized as DAGs.
shareimprove this answer edited Dec 13 '10 at 14:47 answered Dec 13 '10 at 14:22

Will Vousden
18812
add a comment
up Graph theory is very useful in design and analysis of electronic circuits.
0
vote do
wn vote It is very useful in designing various control systems.

E.g. Signal Flow Graphs and Meson's Rule make your life a lot easier while trying to find
transfer functions.
Also, while solving differential equations numerically Graph Theory is used for mesh
generation.

shareimprove this answer answered Dec 13 '10 at 19:52

Pratik Deoghare
3,7481535
add a comment