An identification method for node parameters and topologies of complex networks is investigated with network heterogeneously splitting in a driving-response model.Modeling of heterogeneously associated networks is presented.Two heterogeneously associated complex networks are disassembled and modeled.With LaSalle invariance theory,Gram matrix and driving-response model,adaptive observers for identifying node dynamical parameters and topological parameters of heterogeneously associated complex networks are obtained.Node parameters and coupling topology parameters can be identified on line using the proposed observers.Numerical simulations indicate that the proposed methods are effective.