# LLM-Aligned Geographic Item Tokenization for Local-Life Recommendation

arXiv: https://arxiv.org/abs/2511.14221 · HTML: https://mlsanity.com/paper/2511.14221/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Preference alignment for recommenders and RAG, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 15 of its 39 references that are in the map.
- First posted on arXiv 18 Nov 2025.

## Placed in

- LLM alignment / RLHF / preference optimisation / [Preference alignment for recommenders and RAG](https://mlsanity.com/alignment/preference-alignment-for-recommenders-and-rag/)

## Builds on (most cited first, 15 of 15)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [C-Pack: Packed Resources For General Chinese Embeddings](https://mlsanity.com/paper/2309.07597/) (2023)
- [OneRec: Unifying Retrieve and Rank with Generative Recommender and Iterative Preference Alignment](https://mlsanity.com/paper/2502.18965/) (2025)
- [End-to-End User Behavior Retrieval in Click-Through RatePrediction Model](https://mlsanity.com/paper/2108.04468/) (2021)
- [NoteLLM-2: Multimodal Large Representation Models for Recommendation](https://mlsanity.com/paper/2405.16789/) (2024)
- [QARM: Quantitative Alignment Multi-Modal Recommendation at Kuaishou](https://mlsanity.com/paper/2411.11739/) (2024)
- [Aligning Large Language Models for Controllable Recommendations](https://mlsanity.com/paper/2403.05063/) (2024)
- [Enhancing Embedding Representation Stability in Recommendation Systems with Semantic ID](https://mlsanity.com/paper/2504.02137/) (2025)
- [Generative Next POI Recommendation with Semantic ID](https://mlsanity.com/paper/2506.01375/) (2025)
- [Enhancing Item Tokenization for Generative Recommendation through Self-Improvement](https://mlsanity.com/paper/2412.17171/) (2024)
- [Unified Semantic and ID Representation Learning for Deep Recommenders](https://mlsanity.com/paper/2502.16474/) (2025)
- [Universal Item Tokenization for Transferable Generative Recommendation](https://mlsanity.com/paper/2504.04405/) (2025)
- [LLM-CoT Enhanced Graph Neural Recommendation with Harmonized Group Policy Optimization](https://mlsanity.com/paper/2505.12396/) (2025)
- [LLM-Alignment Live-Streaming Recommendation](https://mlsanity.com/paper/2504.05217/) (2025)

## Built on by (most cited first, 3 of 3)

- [Reasoning Over Space: Enabling Geographic Reasoning for LLM-Based Generative Next POI Recommendation](https://mlsanity.com/paper/2601.04562/) (2026)
- [Why Users Go There: World Knowledge-Augmented Generative Next POI Recommendation](https://mlsanity.com/paper/2605.11807/) (2026)
- [Interpretable Representation via LLM-Driven Generative Disentanglement for Local-Life Service Recommendation](https://mlsanity.com/paper/2607.27944/) (2026)

## Abstract (arXiv)

Recent advances in Large Language Models (LLMs) have enhanced text-based recommendation by enriching traditional ID-based methods with semantic generalization capabilities. Text-based methods typically encode item textual information via prompt design and generate discrete semantic IDs through item tokenization. However, in domain-specific tasks such as local-life services, simply injecting location information into prompts fails to capture fine-grained spatial characteristics and real-world distance awareness among items. To address this, we propose LGSID, an LLM-Aligned Geographic Item Tokenization Framework for Local-life Recommendation. This framework consists of two key components: (1) RL-based Geographic LLM Alignment, and (2) Hierarchical Geographic Item Tokenization. In the RL-based alignment module, we initially train a list-wise reward model to capture real-world spatial relationships among items. We then introduce a novel G-DPO algorithm that uses pre-trained reward model to inject generalized spatial knowledge and collaborative signals into LLMs while preserving their semantic understanding. Furthermore, we propose a hierarchical geographic item tokenization strategy, where primary tokens are derived from discrete spatial and content attributes, and residual tokens are refined using the aligned LLM's geographic representation vectors. Extensive experiments on real-world Kuaishou industry datasets show that LGSID consistently outperforms state-of-the-art discriminative and generative recommendation models. Ablation studies, visualizations, and case studies further validate its effectiveness.