Leveraging Topic Models to Extract Tourist Preference from Points of Interest Descriptions
摘要
Tourist preferences play a crucial role in designing personalized travel recommendations and enhancing destination marketing strategies. This paper presents a comparative analysis and a novel approach for modelling tourist preferences by leveraging topic modeling. The main goal of the paper is to extract Topics Of Interest (TOIs) from textual descriptions of points of interest (POIs) to be subsequently used to express tourist preferences when adopting a tourism recommender system. Descriptions of POIs, which encapsulate their key features and attractions, serve as input to advanced topic models to uncover latent topics that represent key themes or categories of interest. The inferred topics can be used to characterize the preferences of tourists to subsequently recommend relevant itineraries.