Evaluating the Culture and Tourism Integration Model of the Grand Canal National Cultural Park Using Machine Learning Algorithms

The Grand Canal National Cultural Park's cultural and tourism integration approach is evaluated using machine learning techniques in this study. The goal is to enhance visitor experiences while preserving cultural heritage. It also covered visitor data collection through the use of surveys, social media analytics, and historical records to capture various dimensions of visitor behaviour and satisfaction. Extensive pre-processing of data collected includes NLP for social media text analysis, and feature engineering was carried out to bring forth relevant insights about visitors. Satisfaction and preferences of visitors are predicted using three machine-learning models: Random Forests, Decision Trees, and SVMs. The Random Forest model, however, topped with an accuracy of 88.7%, precision of 86.9%, recall of 87.3%, and an F1-Score of 87.1%. Since the model was based on ensemble learning, it was able to make strong predictions by taking in the collective intelligence of a set of multiple decision trees. In contrast, interpretability was provided by the Decision Tree model with its F1-Score of 79.8% and accuracy of 82.3%. Knowing that SVM tends to perform very well in high-dimensional spaces, the model reached an F1 score of 83.9% with an accuracy of 85.5%, hence proving to be competitive. It, therefore, underlines the potentiality of machine learning for taking cultural tourism management to the next level by allowing improved forecasts of visitors' behaviors and preferences. This, therefore, points to further research in refining predictive models with increased data while exploring adaptive management frameworks that would maintain resilient levels of cultural tourism practices at heritage sites such as the Grand Canal National Cultural Park.