Decision-making situations have been dramatically impacted by the development of the economy and technology. Large-scale group decision making (LSGDM) in social networks has become one of the hot topics in the decision-making field. In this paper, we develop a novel framework to reach consensus in social networks for LSGDM with incomplete probabilistic hesitant fuzzy information. In this framework, an expert clustering method based on interpretive structural modeling (ISM) is designed to classify experts. The trust propagation and aggregation operators for probabilistic hesitant fuzzy are explored to achieve indirect trust assessment and weights of experts. Then, an iterative algorithm that checks the personal decision level of experts is presented to estimate the missing values in incomplete probabilistic hesitant fuzzy decision matrices. Also, a consensus process is investigated to eliminate opinion conflict using trust feedback adjustment (TFA) and weight feedback adjustment (WFA). Lastly, a case study is presented to demonstrate the application of the proposed method. The comparison analysis and discussion are established to verify the advantages and effectiveness of the proposed consensus framework for LSGDM in social networks.