Think Before You Act: Intention-Guided Reasoning for LLM-Based Location Prediction

2026-06-06

AI Summary

This work addresses the limitations of existing large language model–based next point-of-interest (POI) prediction methods, which often directly map trajectories to locations and are thus susceptible to spurious correlations and historical frequency biases while overlooking the underlying user intent formation process. To overcome this, we propose IntentPOI, a novel framework that explicitly models user intent as an intermediate reasoning step. IntentPOI employs a two-stage mechanism that mimics human decision-making: it first infers travel intent by integrating historical behavior, similar users’ patterns, and temporal context, then selects the optimal POI from a candidate pool guided by the inferred intent. This paradigm shift—from trajectory matching to intent-driven reasoning—enables more accurate and interpretable predictions. Extensive experiments on three real-world datasets demonstrate that IntentPOI significantly outperforms eleven state-of-the-art baselines, confirming its effectiveness and generalization capability.

Abstract

Predicting a user's next Point-of-Interest (POI) based on their historical check-in records is a fundamental task in location-based services. While recent methods incorporating large language models have shown strong reasoning capabilities and promising results, they typically formulate the prediction task as a one-step trajectory-to-location mapping problem, making predictions prone to shallow trajectory correlations and historical frequency bias. We argue that users rarely choose locations directly and instead, they usually first form a traveling intention and then accordingly select specific POIs. Motivated by this insight, we propose IntentPOI, a two-stage intention-guided reasoning framework. In the thinking stage, we infer users' intermediate intentions by incorporating historical mobility patterns, similar peer behaviors, and the temporal contexts. In the acting stage, we first construct a compact candidate pool, and then perform intention-guided reasoning to identify locations that best align with the inferred intention. By explicitly decoupling intention inference from location prediction, IntentPOI transforms the next POI prediction from direct trajectory matching into intention-guided reasoning. Extensive experiments on three real-world datasets demonstrate that IntentPOI consistently outperforms eleven state-of-the-art baselines.

Problem

• Point-of-Interest prediction
• travel intention
• location-based services
• trajectory reasoning
• frequency bias

Innovation

• intention-guided reasoning
• next POI prediction
• large language models
• two-stage framework
• location-based services