Mixture-of-Experts for Personalized and Semantic-Aware Next Location Prediction

2025-05-30

AI Summary

Existing location prediction methods struggle to model the multifunctional semantics of places and the heterogeneity of user mobility behaviors. To address this, we propose NextLocMoE, a novel Mixture-of-Experts (MoE) framework featuring a hierarchical two-tier architecture: an outer layer leverages large language models (LLMs) to partition experts at the location semantic level, while an inner layer embeds trajectory-driven, personalized expert routing within a Transformer backbone to capture user-specific mobility patterns. We further introduce a history-aware routing mechanism to enhance routing stability and interpretability. Evaluated on multiple real-world urban trajectory datasets, NextLocMoE achieves substantial improvements in prediction accuracy (average +12.7%), cross-domain generalization, and semantic interpretability. Our approach establishes a new paradigm for human mobility modeling that jointly advances semantic depth and individual adaptability.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Planning, Routing, and Scheduling: Planning with Language Models
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Systems and Infrastructure for Web, Mobile and WoT: Location- and context-aware Web and WoT applications and services

Abstract

Next location prediction plays a critical role in understanding human mobility patterns. However, existing approaches face two core limitations: (1) they fall short in capturing the complex, multi-functional semantics of real-world locations; and (2) they lack the capacity to model heterogeneous behavioral dynamics across diverse user groups. To tackle these challenges, we introduce NextLocMoE, a novel framework built upon large language models (LLMs) and structured around a dual-level Mixture-of-Experts (MoE) design. Our architecture comprises two specialized modules: a Location Semantics MoE that operates at the embedding level to encode rich functional semantics of locations, and a Personalized MoE embedded within the Transformer backbone to dynamically adapt to individual user mobility patterns. In addition, we incorporate a history-aware routing mechanism that leverages long-term trajectory data to enhance expert selection and ensure prediction stability. Empirical evaluations across several real-world urban datasets show that NextLocMoE achieves superior performance in terms of predictive accuracy, cross-domain generalization, and interpretability

Problem

• Capturing multi-functional semantics of real-world locations
• Modeling heterogeneous behavioral dynamics across user groups
• Enhancing next location prediction accuracy and interpretability

Innovation

• Dual-level Mixture-of-Experts design
• Location Semantics MoE module
• History-aware routing mechanism