Static greedy: solving the apparent scalability-accuracy dilemma in influence maximization