The lock navigation scheduling problem involves multiple stakeholder groups. The game relationship between stakeholders directly affects the sustainable development of regional society and economy. Considering the objectivity and accuracy of social network analysis (SNA) for relationship identification, indicator screening, and system construction, it completes the stakeholder extraction and evaluation index system construction. Considering that the conditional probability of nodes in Bayesian networks (BN) can flexibly and intuitively characterize the direction and strength of the factors’ roles in a complex scenario, this study proposes an index model for the evaluation of the performance of lock navigation scheduling rules. Firstly, build the BN topology under the efficiency, safety, and fairness criteria. Subsequently, an improved composite index method is combined with the BN to determine the evaluation index from the stakeholder perspective. This study takes the Three Gorges locks maintenance scenario as an example. The index evaluation model can accurately identify the direction, trend, and magnitude of the performance of the scheduling rules on the efficiency, safety, and fairness criteria under navigational constraints, realizing the dynamic and quantitative evaluation of the performance of lock navigation scheduling rules during the observation period.