Efficient Model-Agnostic Continual Learning for Next POI Recommendation

2025-11-12

AI Summary

Existing POI recommendation methods struggle to adapt to the dynamic evolution of user interests. To address this, we propose a novel task—*continual next-POI recommendation*—which requires models to incrementally update on streaming data while avoiding catastrophic forgetting, all under constraints of inference efficiency and memory overhead. We introduce GIRAM, the first model-agnostic continual learning framework for POI recommendation. Its core innovation lies in a generative key-value retrieval mechanism that decouples long-term memory from recent behavioral patterns. Specifically, GIRAM incorporates an interest memory module, context-aware key encoding, and an adaptive fusion strategy to ensure knowledge retention and efficient incremental updates. The framework seamlessly integrates with any POI recommendation backbone. Extensive experiments on three real-world datasets demonstrate that GIRAM significantly improves recommendation accuracy (average +12.7% HR@10) while reducing trainable parameter updates by over 90%, confirming its strong practical deployability.

Technology Category

• Machine Learning: Life-Long and Continual Learning
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Next point-of-interest (POI) recommendation improves personalized location-based services by predicting users'next destinations based on their historical check-ins. However, most existing methods rely on static datasets and fixed models, limiting their ability to adapt to changes in user behavior over time. To address this limitation, we explore a novel task termed continual next POI recommendation, where models dynamically adapt to evolving user interests through continual updates. This task is particularly challenging, as it requires capturing shifting user behaviors while retaining previously learned knowledge. Moreover, it is essential to ensure efficiency in update time and memory usage for real-world deployment. To this end, we propose GIRAM (Generative Key-based Interest Retrieval and Adaptive Modeling), an efficient, model-agnostic framework that integrates context-aware sustained interests with recent interests. GIRAM comprises four components: (1) an interest memory to preserve historical preferences; (2) a context-aware key encoding module for unified interest key representation; (3) a generative key-based retrieval module to identify diverse and relevant sustained interests; and (4) an adaptive interest update and fusion module to update the interest memory and balance sustained and recent interests. In particular, GIRAM can be seamlessly integrated with existing next POI recommendation models. Experiments on three real-world datasets demonstrate that GIRAM consistently outperforms state-of-the-art methods while maintaining high efficiency in both update time and memory consumption.

Problem

• Adapting recommendation models to evolving user interests over time
• Balancing new behavior learning with retention of historical knowledge
• Ensuring efficient model updates for real-world deployment constraints

Innovation

• Generative key-based retrieval for diverse interests
• Adaptive fusion of sustained and recent interests
• Model-agnostic framework for existing recommendation systems