TY - GEN
T1 - An energy model for network community structure detection
AU - Pang, Yin
AU - Li, Kan
PY - 2013
Y1 - 2013
N2 - Community detection problem has been studied for years, but no common definition of community has been agreed upon till now. Former modularity based methods may lose the information among communities, and blockmodel based methods arbitrarily assume the connection probability inside a community is the same. In order to solve these problems, we present an energy model for community detection, which considers the information of the whole network. It does the community detection without knowing the type of network structure in advance. The energy model defines positive energy produced by attraction between two vertices, and negative energy produced by the attraction from other vertices which weakens the attraction between the two vertices. Energy between two vertices is the sum of their positive energy and negative energy. Computing the energy of each community, we may find the community structure when maximizing the sum of these communities energy. Finally, we apply the model to find community structure in real-world networks and artificial networks. The results show that the energy model is applicable to both unipartite networks and bipartite networks, and is able to find community structure successfully without knowing the network structure type.
AB - Community detection problem has been studied for years, but no common definition of community has been agreed upon till now. Former modularity based methods may lose the information among communities, and blockmodel based methods arbitrarily assume the connection probability inside a community is the same. In order to solve these problems, we present an energy model for community detection, which considers the information of the whole network. It does the community detection without knowing the type of network structure in advance. The energy model defines positive energy produced by attraction between two vertices, and negative energy produced by the attraction from other vertices which weakens the attraction between the two vertices. Energy between two vertices is the sum of their positive energy and negative energy. Computing the energy of each community, we may find the community structure when maximizing the sum of these communities energy. Finally, we apply the model to find community structure in real-world networks and artificial networks. The results show that the energy model is applicable to both unipartite networks and bipartite networks, and is able to find community structure successfully without knowing the network structure type.
KW - Bipartite community
KW - Community detection
KW - Energy model
KW - Unipartite community
UR - http://www.scopus.com/inward/record.url?scp=84893079606&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-53914-5_35
DO - 10.1007/978-3-642-53914-5_35
M3 - Conference contribution
AN - SCOPUS:84893079606
SN - 9783642539138
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 410
EP - 421
BT - Advanced Data Mining and Applications - 9th International Conference, ADMA 2013, Proceedings
T2 - 9th International Conference on Advanced Data Mining and Applications, ADMA 2013
Y2 - 14 December 2013 through 16 December 2013
ER -
