Greenways have a series of functions such as ecological protection and promotion of economic development, and exploring the needs of greenway tourists for greenway recreation functions has become an important way to promote the comprehensive development of green ecology. To meet this demand, this paper develops an intelligent model that objectively quantifies tourist satisfaction. The TF-IDF algorithm is used to mine tourists’ comments, and the keywords extracted by TF-IDF are used to compare the similarity with the theme words of the implied themes in the LDA model, to solve the defects of the implied themes in the LDA theme model that are not clear. Create a system to evaluate satisfaction for visitor recreation functions that utilize themes extracted from the LDA model. IPA satisfaction analysis is used to process the data mined by LDA to create visitors’ ratings. Subsequently, the model was put into practice greenway facilities maintenance, rest facilities along the greenway, and greenway diversity excursion program words accounted for small weight, respectively 0.0632, 0.0765, and 0.0833, that is, the tourists’ perception of these topics is low. The average value of the sentiment scores of the 10 themes is 174.852. The score for theme 10 is 85.27, which is significantly lower than the average value, which suggests that tourists are dissatisfied with the greenway recreation facilities in the scenic area. Optimizing the quality of greenway recreation and improving the development level of greenways can be achieved by utilizing this paper as a reference.