Abstract: The problem of finding the minimum positive influence dominating set in a given network has been proved to be NP-hard, and there are greedy algorithms with good performance to solve it. An existing greedy approximation algorithm (Wang-Greedy) and a heuristic algorithm (Raei-Greedy) are analyzed. Accordingly, an improved greedy approximation algorithm (Hybrid-Greedy) to find the minimum positive influence dominating sets in social networks is proposed. It is shown theoretically that hybrid-Greedy remains the same approximation ratio and time complexity as Wang-Greedy. Experimental results on some large real-world social network instances show that Hybrid-Greedy yields better solutions than Wang-Greedy and Raei-Greedy for these social network instances.