发表机构
State Key Laboratory of Complex & Critical Software Environment; Beihang University（复杂与关键软件环境国家重点实验室; 北京航空航天大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
SemPOI-RL框架通过微调LLM、引入SPAM模块并结合推荐导向强化学习，实现LLM语义推理与结构化序列生成的对齐，在两个真实数据集上优于传统推荐器和LLM基线，可生成可解释的异地POI序列。
AI 中文摘要
大语言模型（LLM）具备强大的语义推理与开放生成能力，但将这些能力与结构化序列生成对齐仍具挑战性，这在异地兴趣点（POI）序列生成任务中尤为突出——模型需从用户家乡行为推断可迁移的旅行意图，适配跨城市兴趣漂移，并在结构约束下生成连贯的目的地轨迹。现有方法要么依赖可解释性有限的潜在ID迁移，要么直接使用LLM生成序列，未将推断的语义明确落地到位置感知预测中。为解决这一缺口，本文提出SemPOI-RL框架，将LLM语义推理与结构化序列生成对齐，以实现可解释的异地推荐。具体而言，首先微调LLM，以自然语言作为可解释的语义中间表示，从用户家乡轨迹推断面向目的地的旅行风格；随后引入语义POI对齐模块（SPAM），将这些推断的风格落地到风格条件掩码自动编码器中，用于位置感知轨迹生成；最后应用带推荐导向奖励的强化学习，将LLM生成的风格与下游序列质量对齐。在两个真实世界数据集上的实验表明，SemPOI-RL始终优于传统推荐器与直接LLM基线，同时能在旅行的不同阶段提供可解释的风格归因。代码可在this https URL获取。
英文摘要
Large language models (LLMs) exhibit strong semantic reasoning and open-ended generation abilities, but aligning these abilities with structured sequential generation remains challenging. This challenge is particularly evident in out-of-town (OOT) POI sequence generation, where a model must infer transferable travel intent from a user's hometown behaviors, adapt to cross-city interest drift, and generate a coherent destination trajectory under structural constraints. Existing approaches either rely on latent ID-based transfer with limited interpretability or directly use LLMs for sequence generation without explicitly grounding inferred semantics into position-aware predictions. To address this gap, we propose SemPOI-RL, a framework that aligns LLM semantic reasoning with structured sequence generation for interpretable OOT recommendation. Specifically, we first fine-tune an LLM to infer destination-oriented travel styles from users' hometown trajectories, using natural language as an interpretable semantic intermediate. We then introduce a Semantic POI Alignment Module (SPAM) to ground these inferred styles into a style-conditioned masked autoencoder for position-aware trajectory generation. Finally, we apply reinforcement learning with recommendation-oriented rewards to align LLM-generated styles with downstream sequence quality. Experiments on two real-world datasets show that SemPOI-RL consistently outperforms both traditional recommenders and direct LLM baselines, while providing interpretable style attribution across different phases of a trip. The code is available at https://github.com/Wind-Flipped/SemPOI-RL .
Comments19 pages in total, including 9 pages of main text and 4 figures
