Aspect-based Sentiment Analysis in Tourism Industry for Tourism Recommender System

The improvement of some aspects in tourism industry needs further study through aspect-based sentiment analysis based on tourist experience. The aim of this study is presenting the empiric results of aspect-based sentiment analysis to extract some useful aspects of services for leveraging tourism industry development based on user's feedback. The study uses customer review data from TripAdvisor website for developing classification model using machine learning algorithms including Decision Tree (DT), Random Forest Classifier (RFC), Support Vector Machine (SVM), Extreme Gradient Boosting (XGBoost) models. Each machine learning model under study is trained using K-Fold Cross Validation. The empiric results found that accuracy of RFC achieved the highest result (99.5%) followed by XGBoost (87.3%), DT (82.9%), and SVM (80.5%). The extracted meaningful aspects from customer feedback give many valuable insights for service quality improvement in tourism industry in the effort to strengthening the rising-up of sustainability economic growth through tourism industry.