Kevin S. Xu, Mark Kliger, and Alfred O. Hero III	Adaptive evolutionary clustering	[9]	They recognized the problem of detecting communities in static networks explaining that various community detection methods derived from the methods of graph partitioning and data clustering which are modularity maximization and spectral clustering. They addressed the extension of community detection to dynamic networks and called it as community tracking. They proposed performing of community tracking by an adaptive evolutionary clustering framework. The adaptive Evolutionary clustering combines data from several time steps for computing the clustering results on single time step allowing the clustering results varying smoothly over time.
Nam P. Nguyen, Thang N. Dinh, Ying Xuan and My T. Thai	Quick Community Adaptation	[10]	They explained that nodes mobility and unstable links properties of the network efficient routing scheme design is a difficult issue. Due to the natural tendency of forming groups of communication a groups of nodes are densely connected inside the network than outside. They gave MANET forming community structure. They proposed QCA which is a fast and adaptive algorithm that identifies efficiently the community structure of a dynamic social network. Their approach reduced the computational cost and processing time. QCA is used to develop community identification core deploying routing strategies in MANETs. The simulation results of QCA enables applicability their proposed method in mobile computing
Andrea Lancichinetti and Santo Fortunato	Girvan and Newman (GN benchmark), LFR	[11]	They formalized the problem of using the type of algorithms which are reliable for the applications. They proposed and explained the LFR benchmark and compared partitions quantitatively. They presented analysis of the algorithms and performance on GN benchmark and then on the LFR benchmark for various versions which includes weighted and directed graphs and the graphs of overlapping communities. They explained the issue of algorithms giving a null result and how they handle networks without predictable community structure like random graphs.
Nina Mishra, Robert Schreiber, Isabelle Stanton and Robert E. Tarjan	Graph Clustering	[12]	They formulated the discovery of close-knit clusters in the networks. The clusters are important and should be deployed. The clusters that typically do not overlap have all vertices are clustered ignoring external sparsity thereby limiting to the clustering criteria. They introduced a new measure for overcoming the limitations by naturally combining internal density with external sparsity. Their proposed scheme explored combinatorial properties of internally dense and externally sparse clusters giving an algorithm for provably finding such clusters and assuming that a large gap between internal density and external sparsity is present. But still the components like external sparsity, overlapping clusters in which not every vertex is clustered and internal density lacks. They generated a new graph clustering criterion which perfectly suits for social networks.
Jie Tang, Jimeng Sun, Chi Wang and Zi Yang	Topical Affinity Propagation (TAP)	[13]	They suggested the need of methods to analyze and quantify the social influences for quantitatively measuring the strength of topic level social influence. They explained the representative nodes on a given topic, identified topic level experts and their social influence on a particular node and quick connection to a particular node via tough social ties. For performing topic level influence propagation TAP is able to accept results of any topic modeling and the existing network structure.