A study on the formation mechanism of ancient transportation routes based on explainable machine learning

Linear cultural heritage, as a cross-regional, multi-factor collaborative cultural carrier, plays a crucial role in heritage protection. This study takes the Shu Road in China as an example, combining machine learning analysis to examine the nonlinear relationships and synergistic effects in route selection. The results indicate that (1) Socio-economic factors, natural environment, and military-political considerations are the primary determinants of route selection. (2) There are significant nonlinear and threshold effects among the various factor variables influencing the choice of Shu Road routes, and there are obvious synergistic effects between certain variables. (3) The route selection strategy for the Shu Road exhibited obvious dynamic adjustment characteristics in different historical periods. This study demonstrates the wisdom and experience of ancient transportation systems in adapting to complex environments, emphasizing the multidimensional attributes of transportation routes as critical carriers for military and economic functions, thereby providing support for cultural heritage protection and utilization.