The study aims to make up for the deficiency of the research on the interaction between heritage landscape and human beings from the subjective perspective of human beings in heritage protection. It takes the Ming city wall of Nanjing with a diverse landscape environment as an example and obtains the tourists’ emotions and behavior preferences by collecting the travel note data of dianping.com, using NLP algorithm, semantic analysis, text analysis, and other methods, to explore the interaction between the tourists’ behavior of the city wall and the surrounding environment. The main findings are as follows: (1) Compared with other scenic spots in Nanjing, there is still a significant gap in the number of comments on the Ming city wall of Nanjing, and the interaction between its landscape and human beings, protection and publicity still need to be strengthened. (2) According to the statistics of all tourist evaluations, the positive evaluation accounts for about 80%, and the negative evaluation accounts for about 20%. Tourists have high experience satisfaction and recognition of Nanjing Ming city wall and the surrounding overall landscape environment. (3) Because of the analysis’s positive and negative evaluation contents, combined with the site investigation, this paper summarizes the protection experience and optimization and promotion strategies of Nanjing Ming city wall and its surrounding heritage landscape from the perspective of landscape architecture and put forward reasonable suggestions. This study verifies the applicability of “social network text data” in the research field of heritage landscape protection. It provides a new solution for future heritage protection and urban green space planning.