IntroductionTraditional villages are precious historical and cultural heritage sites. The selection of post-earthquake recovery and reconstruction (PERR) mode directly affects the village cultural heritage protection and the development direction of post-disaster reconstruction. A scientific and comprehensive feasibility evaluation for selecting the PERR mode of traditional villages can provide sufficient evidence for the recovery efforts in earthquake-stricken villages.MethodThe author summarizes three PERR modes and constructs an evaluation index system for the selection of PERR modes of traditional villages. Based on the interrelationship of the indicators, the author has preliminarily established the Dependency Analytic Process (DAP), Based on this method, a model of traditional village PERR mode selection is constructed, and an empirical analysis is carried out in the case of the earthquake-stricken area of Xieluo-buzi Village in 2022 M6.8 Luding earthquake, to discuss the selection of PERR modes of traditional villages.ResultsThe authors have explored the application of the DAP in the selection of PERR modes for traditional villages and verified the effectiveness of the method. Since a large amount of actual research work is required to conduct an assessment, it is believed that with the widespread applications of the DAP, its superiority and practicality will be further demonstrated.ConclusionThe protection of traditional villages is a dynamic protection process, in which the will of the indigenous people is respected, the social network of the indigenous people is maintained, and the fair rights of the indigenous people to participate in the implementation of the project and to enjoy the preferential policies and resource benefits are guaranteed, as they are the real main body of the heritage protection, so that the traditional village ethnic heritage can be inherited and developed permanently in the protection. DAP is applicable to the comprehensive evaluation of multiple factors, particularly in situations where the importance of the indicators is difficult to be distinguished from each other. This is a new method to determine the weight vector, which has a broad application prospect.