A Multi-Source Ensemble Approach to Candidate Generation for Alternative Vacation Rental Property Recommendations

A hybrid candidate-generation architecture combining item-based collaborative filtering with graph neural network retrieval improved Recall@300 by 14.8% over the strongest baseline on a vacation rental platform with over 2M active properties, according to an arXiv paper (2609.05748v1). The study also found GNN-based embeddings alone delivered a 48-68% relative recall improvement across K over shallow Hotel2Vec embeddings, and that a stronger candidate pool carried through to higher downstream ranking quality, though the authors note the recall-conversion gap is complicated by coupling between candidate generation and ranker training.

arXiv:2609.05748v1 Announce Type: new Abstract: Alternative property recommendations play a critical role in vacation rental marketplaces, helping users discover relevant options when viewing a specific listing. However, generating high-quality candidate alternatives presents unique challenges: heterogeneous inventory, geographic constraints, rapid availability changes, and long-tail property distributions. We present a comprehensive study of candidate generation CG approaches for vacation rental alternatives, comparing collaborative filtering, shallow embeddings, and graph neural network GNN methods. Our experiments on a large-scale vacation rental platform over 2M active properties show that a hybrid architecture combining item-based collaborative filtering with GNN-based retrieval improves Recall@300 by 14.8% over the strongest baseline, by leveraging the complementary strengths of the two sources: collaborative filtering excels at early recall for properties with rich interaction history, while GNNs discover diverse, non-obvious alternatives and handle cold-start scenarios more effectively. As a component result, GNN-based embeddings alone substantially outperform shallow Hotel2Vec embeddings 48-68% relative recall improvement across K , motivating their inclusion in the ensemble. Crucially, we examine how CG-stage gains carry through to the downstream ranking stage, and find that a stronger candidate pool yields higher downstream ranking quality, though attributing this effect cleanly is complicated by the coupling between candidate generation and ranker training. This recall-conversion gap is an important consideration for practitioners deploying new retrieval methods in two-stage recommendation systems.