An Empirical Evaluation of Cross-City POI Recommendation on a Large-Scale Benchmark
Artificial IntelligenceInformation Retrieval
Summary
The authors studied how well popular methods recommend places to visit in new cities using a large, worldwide dataset called Trip World. They found that current models tend to rely too much on general location popularity instead of understanding individual user tastes when suggesting new places. Simpler models work about as well or better, and adding extra information about places doesn’t help much. They also tested more advanced methods, which didn’t outperform basic popularity-based suggestions. Overall, the authors suggest new designs are needed to better transfer user preferences and use rich place data for recommending spots in cities users haven’t visited before.
Authors
Peibo Li, Yang Song, Hao Xue, Maarten de Rijke, Flora D. Salim
Abstract
Cross-city point-of-interest (POI) recommendation is crucial for navigating unfamiliar urban environments, yet its progress has historically been constrained by data limitations. Using the recently proposed large-scale benchmark Trip World, we empirically re-examine whether conclusions drawn on small prior benchmarks still hold under worldwide coverage, low home-destination region overlap, and large, semantically rich POI inventories. Our evaluation surfaces three bottlenecks of representative state-of-the-art methods: (1) hometown-aware models appear to rely more on destination-region priors than on user-specific preference transfer; (2) their accuracy-efficiency trade-off degrades at this scale, where the simplest model is among the strongest; and (3) existing mechanisms for integrating semantic metadata yield little benefit. We further include a diagnostic pilot on agentic methods adapted from next-POI recommendation, finding that naive adaptation trails a simple popularity prior even though the relevant semantic signal is present in the data. These results highlight the need for task-specific designs that support cross-city preference transfer, semantic grounding, and scalable reasoning over unseen destination inventories.