Travel itinerary recommendation using interaction-based augmented data
Keisuke Otaki, Yukino Baba · Expert Systems with Applications · 2024
Itinerary planning is complicated for travelers because the traveling content, including places to visit, acceptable times, and distances, can be diverse. Travel recommender systems (TRSs) recommend the most relevant itineraries for a traveler. In this paper, we propose an interactive framework that allows users to edit itineraries directly on a map to suit their preferences better. The proposed framework collects feedback data by recording user itinerary modifications, infers positive and negative preferences, and fine-tunes the recommender models with our ranking-based loss function. This interaction-based data augmentation approach addresses data sparsity issues due to personalization by capturing a variety of travel item combinations. In our experiments, we evaluate multiple combinations of models and itinerary generation methods to show the effectiveness of integrating interaction data into TRSs. Our experimental evaluations demonstrate that our interactive TRS can provide itineraries that align with users’ preferences more in terms of point-set-wise and rank-wise accuracy; the integration consistently improves the accuracy for all combinations of the components, and particularly, the improvement is large for small backbone models.