In emergency decision-making, decision-makers often empathize with other experts' preferences due to lacking information, and preferences conflicts hinder the achievement of consensus. For group decision-making problems involving incomplete preference relations, existing methods often neglect to mine high-quality additional information as assistance, and limited information reduces the reliability of decision results. In traditional research, preference completion and consensus achievement are separate, making conflict identification in feedback hard and reducing consensus efficiency. To overcome these issues, a group decision-making framework is proposed to simultaneously achieve completion and consensus under an empathetic network. This framework first fuses indirect preferences and empathetic preferences to construct an ordinal regression consensus discrimination model. Then, we design a feedback mechanism to accurately identify conflicts and construct cost-effective consensus models. This method not only deeply mines implicit information but also promotes the integration of missing values completion, consensus achievement, and consistency realization, enhancing the credibility of estimates and the efficiency of consensus reaching. Finally, the effectiveness of the proposed framework is verified through typhoon disaster case, simulation, and comparison.