POI-Enhancer: An LLM-based Semantic Enhancement Framework for POI Representation Learning

TL;DR

POI-Enhancer leverages large language models with specialized prompts and novel modules to enrich POI representations with semantic knowledge, significantly improving performance in user mobility tasks.

Contribution

This paper introduces POI-Enhancer, a novel LLM-based framework with modules for extracting, aligning, and fusing semantic information into POI representations, addressing previous weak textual features.

Findings

Significant performance improvements across multiple datasets.

Effective extraction and integration of semantic knowledge from LLMs.

Enhanced POI representations lead to better task outcomes.

Abstract

POI representation learning plays a crucial role in handling tasks related to user mobility data. Recent studies have shown that enriching POI representations with multimodal information can significantly enhance their task performance. Previously, the textual information incorporated into POI representations typically involved only POI categories or check-in content, leading to relatively weak textual features in existing methods. In contrast, large language models (LLMs) trained on extensive text data have been found to possess rich textual knowledge. However leveraging such knowledge to enhance POI representation learning presents two key challenges: first, how to extract POI-related knowledge from LLMs effectively, and second, how to integrate the extracted information to enhance POI representations. To address these challenges, we propose POI-Enhancer, a portable framework that leverages LLMs to improve POI representations produced by classic POI learning models. We first design three specialized prompts to extract semantic information from LLMs efficiently. Then, the Dual Feature Alignment module enhances the quality of the extracted information, while the Semantic Feature Fusion module preserves its integrity. The Cross Attention Fusion module then fully adaptively integrates such high-quality information into POI representations and Multi-View Contrastive Learning further injects human-understandable semantic information into these representations. Extensive experiments on three real-world datasets demonstrate the effectiveness of our framework, showing significant improvements across all baseline representations.