Incremental local community identification in dynamic social networks

Abstract

Abstract-Social networks are usually drawn from the interactions between individuals, and therefore are temporal and dynamic in essence. Examining how the structure of these networks changes over time provides insights into their evolution patterns, factors that trigger the changes, and ultimately predict the future structure of these networks. One of the key structural characteristics of networks is their community structure -groups of densely interconnected nodes. Communities in a dynamic social network span over periods of time and are affected by changes in the underlying population, i.e. they have fluctuating members and can grow and shrink over time. In this paper, we introduce a new incremental community mining approach, in which communities in the current time are obtained based on the communities from the past time frame. Compared to previous independent approaches, this incremental approach is more effective at detecting stable communities over time. Extensive experimental studies on real datasets, demonstrate the applicability, effectiveness, and soundness of our proposed framework.

Detecting Community Structure in Dynamic Social Networks Using the Concept of Leadership

“…Although their method is faster than the method which clusters each snapshot independently [4], ), running their method for a long time on a dynamic network will end up in poor quality results. The method proposed by Takaffoli et al [37] intends to find the community structure at any time based on the extracted clusters from the previous timestep. They introduced an adaptive algorithm which like two-stage methods, employs an event-tracking framework.…”

Local Community Detection in Dynamic Networks

“…The persistence requirements could be considered in conjunction with conductance to further rank candidates of interest. Evolutionary clustering for dynamic networks is another related and very active area of research [21,18,6,29]. The goal in evolutionary clustering is to track the changes in the global network partitions over time, where partitions are allowed to vary from one time slice to the next by incorporating temporal smoothness of partition membership.…”

“…Unlike evolutionary clustering [21,18,6], whose goal is to partition all vertices at every timestamp, our goal is to identify the most cohesive communities and their interval of activity without clustering all nodes in time. Hence, our problem is more similar to local community detection [3,13,29] than partitioning. In addition, evolutionary clustering methods are often concerned with the long-term group membership evolution: how partitions appear, grow, shrink and disappear [6].…”

Clustering Evolving Networks

“…This is also the case for static data sets; however, due to even more degrees of freedom, for example the frequency of time steps or the question whether relations age and disappear over time, this is even more immanent in the case of dynamic data. Takaffoli et al [142] choose monthly snapshots over one year, which considerably decreases the number of vertices and edges in the network. As a kind of ground truth clustering, they identify "persisting topics" based on keyword extraction.…”