The significance of dealing with internet rumors or false information on social media is expanding, especially when an unexpected crisis erupts. The promotion of social media's rumor refuting capacity needs more attention. Based on this, in order to systematically evaluate the rumor refuting effect of social media platforms, so that platforms can find their own weaknesses and improve their operational capacity, this paper puts forward an evaluation system and evaluation methods for anti-rumor effect of social media platforms, and applies them to five mainstream social media platforms in China for evaluation and analysis. Suggestions are also given for social media platforms with relatively poor ratings to improve the effectiveness of rumor refutation. First, build the evaluation index system, use the analytic hierarchy process (AHP) method which combines triangular fuzzy number to determine the index weight, then divide the evaluation index into qualitative and quantitative indicators. For qualitative indicators, use intuitive fuzzy number set (IFNs) method to represent the information of the two comparisons of indicators. For quantitative indicators, collect real data and conduct standardized processing. Finally use TOPSIS method to evaluate the effectiveness of social media platform rumors, and get rankings. And, a sensitivity analysis will be performed to check the robustness of the final decision. Based on the results, relevant suggestions are given.