SA-LSPL:Sequence-Aware Long- and Short- Term Preference Learning for next POI recommendation

TL;DR

The paper tackles the next POI recommendation problem by addressing long sequence dynamics, sparse data, and the need to model both consecutive and non-consecutive user movements. It introduces SA-LSPL, a end-to-end framework that fuses multi-modal embeddings with a hierarchy of long- and short-term preference models, including Bi-LSTM historical trajectory encoding, inter-trajectory spatio-temporal attention, personalized social context, inter-level self-attention, and STC-dilated LSTM for non-consecutive visits, culminating in a robust prediction layer. The key contributions are explicit sequence-level spatio-temporal modeling, integration of social context for data-sparse users, and a novel adaptive weighting mechanism balancing continuous and non-continuous check-ins, all validated on NYC and TKY datasets with clear performance gains over state-of-the-art baselines. The findings demonstrate that incorporating sequence-aware spatio-temporal correlations and category transitions yields meaningful improvements in predictive accuracy, with practical implications for location-based services and personalized mobility recommendations. Future work includes extending the framework with graph neural networks to further leverage contextual information in next POI prediction.

Abstract

The next Point of Interest (POI) recommendation aims to recommend the next POI for users at a specific time. As users' check-in records can be viewed as a long sequence, methods based on Recurrent Neural Networks (RNNs) have recently shown good applicability to this task. However, existing methods often struggle to fully explore the spatio-temporal correlations and dependencies at the sequence level, and don't take full consideration for various factors influencing users' preferences. To address these issues, we propose a novel approach called Sequence-Aware Long- and Short-Term Preference Learning (SA-LSPL) for next-POI recommendation. We combine various information features to effectively model users' long-term preferences. Specifically, our proposed model uses a multi-modal embedding module to embed diverse check-in details, taking into account both user's personalized preferences and social influences comprehensively. Additionally, we consider explicit spatio-temporal correlations at the sequence level and implicit sequence dependencies. Furthermore, SA-LSPL learns the spatio-temporal correlations of consecutive and non-consecutive visits in the current check-in sequence, as well as transition dependencies between categories, providing a comprehensive capture of user's short-term preferences. Extensive experiments on two real-world datasets demonstrate the superiority of SA-LSPL over state-of-the-art baseline methods.

Figures (9)

• Figure 1: The overall framework of our model.
• Figure 2: Explanation of spatio-temporal correlation between different trajectories.
• Figure 3: Time correlation matrix.
• Figure 4: The similarity in check-in behaviors among users.
• Figure 5: Category Transition Matrix and Generation Process.