System and method for grouping of users into overlapping clusters in social networks
Members of a social network user's social graph are automatically segregated into overlapping clusters according to patterns of their past communications. Each cluster within the social graph represents a group of members having a high degree of intra-cluster communication or other connection with one another. The clustering is performed according to a sorting or ranking in accordance with non-principal eigenvectors of connectivity matrices describing the intra-cluster communications/connections. The overlapping clusters exhibit maximum internal density and minimum external sparsity.
1. A method, comprising automatically clustering members of a social network user's social graph, in which graph individuals are represented as nodes and connections between individuals are represented as edges between the nodes, the social graph thereby being a defined set of relationships among the members, the clustering being performed using patterns of past communications among the members of the social graph as a basis for said clustering by partitioning said members of the social graph into respective, overlapping clusters according to a defined optimization function, with each respective cluster of the overlapping clusters representing a group of said members having a high degree of intra-cluster communication, said clustering performed (i) by a computer system according to a sorting in accordance with non-principal eigenvectors of connectivity matrices describing the intra-cluster communications, and (ii) according to the optimization function, which comprises a rule-based quality function which favors selection of clusters which have minimum similarity among their respective memberships, for which eigenvectors of the connectivity matrices exhibit maximum internal density and minimum external sparsity, and which imposes a penalty based on a size of a cluster.
2. The method of claim 1 , wherein the connectivity matrices comprise an adjacency matrix and a Laplacian, the adjacency matrix having elements
a
i
,
j
=
{
1
,
(
v
i
,
v
j
)
∈
E
0
,
(
v
i
,
v
j
)
∉
E
and the Laplacian having elements
I
i
,
j
=
{
a
i
,
j
,
i
≠
j
∑
k
a
i
,
k
,
i
=
j
where vertices v represent the members of the user's social graph and E represents connections between said members.