Abstract – Detecting overlapping communities in social network isan important problem in the Data mining. In existence, overlapping communitydetection with swarm intelligence frequently generates overlapping communitystructures with surplus small communities. To deal with this problem, anefficient algorithm LEPSO based on line graph, ensemble learning and particleswarm optimization (PSO) proposed for overlapping community detection. Adiscrete PSO, consists of an encoding scheme with ordered neighborhood list andparticle updating tactic with ensemble clustering, is devised for improving theoptimization ability to search the hidden communities in the social networks.Then, a post processing approach is accessible for merging the finer-grainedand suboptimal overlapping communities.
Experiment on some real world datasetsby using the LEPSO is superior in terms of robustness, effectiveness andautomatically determine the number of clusters.Keywords: Particle Swarm Optimization, Overlapping Community,Line Graph, Ensemble learning, Social Network Analysis. I. INTRODUCTION In the last decade, the social network hasexperienced explosive growth. Social websites, such as YouTube and Flickr, havebillions of users sharing their information, users’ opinion and videos everyday. In order to interact users have various features like a comment,subscribe, like a reply, etc.
users easily communicate with each other. Theusers are interested to share their information around particular topics theyare represented as communities. Community detection is a great impact on understanding the organizationand the functions of the group are to be detected.
Community mining techniquesand theories have developed various applications such as photo tagging, videogloss etc.Communitymining techniques and theories have developed various applications such asphoto tagging, video gloss etc. For example 1 involved detecting communitiesbased on edge betweenness, the counts of short loops in networks and voltagedifferences in resistor networks. This method is ideal for the types ofreal-world network data with which current research is concerned. M.
Cheung,2 proposed a Genetic Algorithm is used for detecting communities in complexnetworks for finding optimizing network. The Algorithm has O (e) timecomplexity and does not need prior knowledge about the number of communities.L. Wang and X.-S. Hua 3 proposed an algorithm for network correlation basedsocial friend recommendation and correlated different “social role” networks tofind a relationship and make friend recommendations. Networks are aligned byselecting important features from each network. S.
Fortunato 4 proposed todetect network community a new model based on particle swarm optimization. Theyare extremely sensitive to the initial solution, easy to fall into the localoptimum.Asocial network can be represented as a graph. The graph consists of nodes andedges. Edges can be used to interact with the nodes. Optimizing the graph weneed some objective functions, such as modularity Q 4. The swarm intelligencealgorithms are useful for overlapping communities’ detection.
PSO does notcapture fully community structure. Fig 1. Drawbacksof traditional approaches based on swarm intelligence optimization.
(a) Real communitystructure. (b)generated community structure (Q = 0.42).