Inductive Transfer Learning for Graph-Based Recommenders

2025-10-26

AI Summary

Transductive graph recommendation systems suffer from poor generalization to unseen users, items, or entirely new datasets. To address this, we propose NBF-Rec—the first graph recommendation framework enabling **disjoint cross-dataset inductive transfer**, wherein no user or item overlaps between source and target domains. NBF-Rec dynamically generates node embeddings via interaction-level message passing, eliminating the need for retraining on the target domain and supporting zero-shot cross-domain recommendation and lightweight fine-tuning. By unifying inductive learning and transfer learning principles, it operates without requiring shared entities across domains. Extensive evaluation across seven real-world datasets demonstrates that NBF-Rec significantly outperforms baselines under zero-shot settings; further, performance improves substantially with only minimal target-domain data for fine-tuning. This work establishes, for the first time, the feasibility and effectiveness of cross-dataset inductive transfer in graph-based recommendation.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Data Mining & Knowledge Management: Recommender Systems
• Application Domains: Humanities & Computational Social Science

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Search and Retrieval-Augmented AI: Vertical and domain-specific search

Abstract

Graph-based recommender systems are commonly trained in transductive settings, which limits their applicability to new users, items, or datasets. We propose NBF-Rec, a graph-based recommendation model that supports inductive transfer learning across datasets with disjoint user and item sets. Unlike conventional embedding-based methods that require retraining for each domain, NBF-Rec computes node embeddings dynamically at inference time. We evaluate the method on seven real-world datasets spanning movies, music, e-commerce, and location check-ins. NBF-Rec achieves competitive performance in zero-shot settings, where no target domain data is used for training, and demonstrates further improvements through lightweight fine-tuning. These results show that inductive transfer is feasible in graph-based recommendation and that interaction-level message passing supports generalization across datasets without requiring aligned users or items.

Problem

• Enables graph-based recommenders to generalize across datasets with disjoint entities
• Eliminates retraining requirements through dynamic embedding computation at inference
• Achieves competitive zero-shot performance and improves with lightweight fine-tuning

Innovation

• Graph-based recommender with inductive transfer learning
• Dynamic node embedding computation at inference
• Zero-shot generalization via interaction-level message passing