A community detection algorithm based on discrete mixed fruit fly optimization

With the increasingly close connection between complex networks and actual life, the studies on social networks are gradually deepened. This paper studies the “adopting community detection of signed complex networks by discrete fruit fly Optimization algorithm”. Community detection is aimed at dividing a clustered network with many nodes, which can make points in the same community more intense and points in different communities more sparse. Traditional fruit fly optimization algorithm has two visible disadvantages that it is easy to fall into local optimum and it works hard to deal with discrete data. Therefore, a Discrete fruit fly algorithm (DFOA) is proposed to solve community detection in this paper. Firstly, the global individual interaction strategy is cited to avoid falling into local optimum in community detection. Secondly, for the problem that the fruit fly algorithm cannot deal with discrete data even so cannot solve module function, citing discrete coding strategy and cross strategy. In addition, three real networks are comparatively analyzed in this paper and the result proves the effectiveness of the proposed algorithm.