This paper studies the problem of identifying multiple information sources in networks. Assuming that the information diffusion follows a Susceptible-Infected (SI) model which allowing all nodes in the network are in the susceptible state or infected state. The number of information sources is known, we propose a simple method to identify multiple diffusion sources. After the diffusion started, any node can be infected very quickly, for an arbitrary node, it is infected by its closest source in terms of spreading time. Therefore, to identify multiple diffusion sources, we partition the information diffusion network and minimize the sum of all partition propagation times. Then in each partition we can find a node which has the minimum spreading time as diffusion source. Furthermore, we also give a new method to estimate the spreading time which can improve the proposed multiple sources identification algorithm. Simulation results show that the proposed method has distinct advantages in identifying multiple sources in various real-world networks.