An Intelligent Decision Support System for Tourists Management Via Global Tourism Analysis

One of the sectors that showed a positive growth following COVID-19 is tourism, which contributes the greatest part of the foreign exchange in developing nations, with foreign countries, particularly in Vietnam, starting to open up and the pandemic restriction of foreign travel. This paper suggests a CNN-LSTM deep learning model as the manager of tourists, relying on the global tourism analysis, with the structured tourism data being obtained at Kaggle. Similar data preprocessing methods as Standard Scaler normalization and label encoding were used. CNN module was employed to obtain useful feature extraction, whereas the LSTM component was used to obtain sequence dependencies in order to improve predictive capability. The hyperparameter tuning was conducted in order to get the best model configuration. The experimental findings indicate that the proposed model had a high generalization and low overfitting, where it had 99.5% accuracy (ACC), 99.8% precision (PRE), 99.7% recall (REC), and a 99.7% F1-score (F1) and low loss of 0.37. The comparative analysis also confirms that CNN-LSTM model is better than the conventional approaches like KNN, Bernoulli Naive Bayes, and Random Forest, and therefore it is a powerful model to use with intelligent tourism data analysis and management.