HSUGA: LLM-Enhanced Recommendation with Hierarchical Semantic Understanding and Group-Aware Alignment

2026-05-12

AI Summary

This work addresses two key challenges in current large language model–enhanced sequential recommendation methods: the unreliability of user semantic embeddings derived from long interaction sequences and the oversimplified assumption of a uniform semantic strategy that ignores variations in user activity levels. To overcome these limitations, we propose HSUGA, a novel framework comprising a Hierarchical Semantic Understanding (HSU) module and a Group-Aware Alignment (GAA) module. HSU improves embedding reliability by modeling preference evolution in a staged manner, while GAA dynamically adjusts the strength of semantic alignment based on individual user activity, enabling personalized utilization of semantic information. Extensive experiments demonstrate that HSUGA significantly outperforms existing approaches across three benchmark datasets, exhibiting both strong effectiveness and high compatibility with diverse recommendation architectures.

Abstract

Large language model (LLM)-enhanced sequential recommendation typically aims to improve two core components: user semantic embedding extraction and utilization. Despite promising results, existing methods still have two limitations: 1) In the extraction stage, most methods directly input long interaction sequence fragments into LLM for preference summarization. However, excessively long sequences increase inference difficulty, making it challenging to reliably infer accurate user embeddings. 2) In the utilization stage, most methods employ the same semantic embedding utilization strategy for all users, neglecting the differences caused by user activity levels, leading to suboptimal performance. To address these issues, we propose HSUGA, which introduces a simple yet effective plugin for each of the two core components: Hierarchical Semantic Understanding (HSU) and Group-Aware Alignment (GAA). HSU performs a staged two-phase preference mining and models preference evolution through constrained editing operations, thereby improving the reliability of user semantic extraction. GAA adjusts the intensity of semantic utilization based on user activity levels, providing weaker alignment for active users and stronger guidance for users with sparse historical data. Finally, extensive experiments on three benchmark datasets demonstrate the effectiveness and compatibility of HSUGA.

Problem

• LLM-enhanced recommendation
• user semantic embedding
• long interaction sequences
• user activity levels
• sequential recommendation

Innovation

• Hierarchical Semantic Understanding
• Group-Aware Alignment
• LLM-enhanced Recommendation
• Preference Evolution
• User Activity-aware Utilization