Enhancing scenic recommendation and tour route personalization in tourism using UGC text mining

Tourism is vital to national economic growth and fulfilling individuals' spiritual pursuits. However, traditional scenic recommendation algorithms must improve accuracy, susceprecommended route personalizationtion of recommended tour routes, and suboptimal tour value. This study presents a UGC text mining-based travel recommendation method to address these problems. To begin with, the proposed method utilizes text mining techniques to enrich scenic tag data and introduces the recommendation algorithm based on fuzzy label-matching user-item characteristics (FLMC). Additionally, we incorporate scenic attributes and travel time between scenics to identify four key attributes that impact scenic selection in tour routes. Users' subjective preferences for these attributes are calculated, and objective weights derived from information entropy are used to determine combination weights for scenic selection. This ensures a more personalized and optimized tour route. Lastly, the proposed method features a tour route ranking method that selects the highest-scoring route as the recommended option. This approach enhances the overall quality of the recommended tour routes. Through experimentation with 240,194 reviews of 369 scenics in Beijing, in the attraction recommendation scenario, the proposed FLMC algorithm achieves up to 75.40% accuracy, which is higher than the other three compared algorithms (46.48%, 35.90%, and 32.98%). In the travel route recommendation scenario, the proposed method has a route gain value of up to 1.57 and a user experience value of recommended route personalizationd compared to the other three compared algorithms (1.03, 0.9; 0.95, 0.53; 1.12, 0.85). These results highlight the significant potential of the proposed method in improving the personalization of travel recommendations and enhancing tourist experiences.