Abstract
For the purpose of detecting communities in social networks, a triad percolation method is proposed, which first locates all close-triads and open-triads from a social network, then a specified close-triad or open-triad is selected as the seed to expand by utilizing the triad percolation method, such that a community is found when this expanding process meet a particular threshold. This approach can efficiently detect communities not only from a densely social network, but also from the sparsely one. Experimental results performing on real-world social benchmark networks and artificially simulated networks give a satisfactory correspondence.