A new algorithm for positive influence maximization in signed networks.

•We formally define the problem of positive influence maximization problem in signed networks under IC model. We also propose a set of propagation rules to model the competitive influence spreading.•We present an algorithm to compute the positive activation probability between each node pair using the independent paths.•To avoid the time-consuming process of simulating the influence propagation, we define a propagation function to estimate the influenced range of a seed set.•An algorithm named PIMSN (positive influence maximization in signed networks) is presented for detecting optimal seed set using the propagation function.