OneLoc: Geo-Aware Generative Recommender Systems for Local Life Service

2025-08-20

AI Summary

Addressing the challenge of jointly modeling user interests and real-time geographic locations for video recommendation in local life services, this paper proposes the first end-to-end generative geographically-aware recommendation framework. To overcome insufficient geographic signal utilization and difficulties in harmonizing multiple objectives—user interest, spatial proximity, and GMV—we introduce four key innovations: (1) a geographically-aware semantic ID encoding scheme to capture spatial semantics; (2) a geography-aware self-attention mechanism to model location-sensitive user-item interactions; (3) a neighborhood-aware prompting strategy to enhance contextual relevance; and (4) a reinforcement learning–based dual reward function jointly optimizing for geographic proximity and GMV. Deployed on the Kuaishou App, the model serves 400 million daily active users, achieving statistically significant improvements of +21.02% in GMV and +17.89% in order volume over strong baselines.

Technology Category

• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Data Mining & Knowledge Management: Recommender Systems
• Computer Vision: Remote Sensing / Geospatial AI

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Local life service is a vital scenario in Kuaishou App, where video recommendation is intrinsically linked with store's location information. Thus, recommendation in our scenario is challenging because we should take into account user's interest and real-time location at the same time. In the face of such complex scenarios, end-to-end generative recommendation has emerged as a new paradigm, such as OneRec in the short video scenario, OneSug in the search scenario, and EGA in the advertising scenario. However, in local life service, an end-to-end generative recommendation model has not yet been developed as there are some key challenges to be solved. The first challenge is how to make full use of geographic information. The second challenge is how to balance multiple objectives, including user interests, the distance between user and stores, and some other business objectives. To address the challenges, we propose OneLoc. Specifically, we leverage geographic information from different perspectives: (1) geo-aware semantic ID incorporates both video and geographic information for tokenization, (2) geo-aware self-attention in the encoder leverages both video location similarity and user's real-time location, and (3) neighbor-aware prompt captures rich context information surrounding users for generation. To balance multiple objectives, we use reinforcement learning and propose two reward functions, i.e., geographic reward and GMV reward. With the above design, OneLoc achieves outstanding offline and online performance. In fact, OneLoc has been deployed in local life service of Kuaishou App. It serves 400 million active users daily, achieving 21.016% and 17.891% improvements in terms of gross merchandise value (GMV) and orders numbers.

Problem

• Developing geo-aware generative recommender for local services
• Balancing user interests, location, and business objectives
• Integrating geographic information into end-to-end recommendation model

Innovation

• Geo-aware semantic ID for tokenization
• Geo-aware self-attention in encoder
• Reinforcement learning with dual rewards