POIBERT: A Transformer-based Model for the Tour Recommendation Problem
This work addresses personalized tour planning for tourists by leveraging past trajectories, though it is incremental as it adapts an existing NLP model to a new domain.
The authors tackled the tour recommendation problem by proposing POIBERT, a transformer-based model that adapts BERT to generate personalized itineraries based on user preferences and time constraints, achieving improved performance over existing sequence prediction algorithms on a Flickr dataset across seven cities with higher recall, precision, and F1-scores.
Tour itinerary planning and recommendation are challenging problems for tourists visiting unfamiliar cities. Many tour recommendation algorithms only consider factors such as the location and popularity of Points of Interest (POIs) but their solutions may not align well with the user's own preferences and other location constraints. Additionally, these solutions do not take into consideration of the users' preference based on their past POIs selection. In this paper, we propose POIBERT, an algorithm for recommending personalized itineraries using the BERT language model on POIs. POIBERT builds upon the highly successful BERT language model with the novel adaptation of a language model to our itinerary recommendation task, alongside an iterative approach to generate consecutive POIs. Our recommendation algorithm is able to generate a sequence of POIs that optimizes time and users' preference in POI categories based on past trajectories from similar tourists. Our tour recommendation algorithm is modeled by adapting the itinerary recommendation problem to the sentence completion problem in natural language processing (NLP). We also innovate an iterative algorithm to generate travel itineraries that satisfies the time constraints which is most likely from past trajectories. Using a Flickr dataset of seven cities, experimental results show that our algorithm out-performs many sequence prediction algorithms based on measures in recall, precision and F1-scores.