Tourist satisfaction is one of the most important indicators for evaluating service quality and the success of ecolodge accommodations. Given the complexity of the factors influencing this phenomenon, the application of intelligent models can provide more accurate predictions than conventional analytical approaches. This study develops an intelligent model based on an Adaptive Neuro-Fuzzy Inference System to predict tourist satisfaction with ecolodges in Zanjan province, Iran. Data were collected through web mining of online reviews for 69 ecolodges from the Jabama platform. Six key indicators—check in experience, cleanliness, information accuracy (match with online description), hosting quality, geographical location, and price value—were used as input variables, while overall satisfaction served as the output. The model was implemented in MATLAB using the Sugeno fuzzy inference system, Gaussian membership functions, and the grid partition method. The model achieved a mean RMSE of 0.4557 and converged within the second training epoch, demonstrating acceptable accuracy in simulating overall tourist satisfaction. Surface plots of variable interactions revealed nonlinear and complex relationships; notably, geographical location showed significant interactive effects with price value and cleanliness on satisfaction. These findings are consistent with previous studies and highlight the advantage of the neuro fuzzy approach over conventional linear regression. The proposed model offers a practical decision support tool for ecolodge managers, online booking platforms, and the provincial cultural heritage organization to improve service quality, standardize ecolodge ratings, and guide targeted investments. However, limitations include the use of a single data source and the exclusion of socio cultural and emotional dimensions, which can be addressed in future research.