{"ID":2849116,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.24430","arxiv_id":"2510.24430","title":"From Time and Place to Preference: LLM-Driven Geo-Temporal Context in Recommendations","abstract":"Most recommender systems treat timestamps as numeric or cyclical values, overlooking real-world context such as holidays, events, and seasonal patterns. We propose a scalable framework that uses large language models (LLMs) to generate geo-temporal embeddings from only a timestamp and coarse location, capturing holidays, seasonal trends, and local/global events. We then introduce a geo-temporal embedding informativeness test as a lightweight diagnostic, demonstrating on MovieLens, LastFM, and a production dataset that these embeddings provide predictive signal consistent with the outcomes of full model integrations. Geo-temporal embeddings are incorporated into sequential models through (1) direct feature fusion with metadata embeddings or (2) an auxiliary loss that enforces semantic and geo-temporal alignment. Our findings highlight the need for adaptive or hybrid recommendation strategies, and we release a context-enriched MovieLens dataset to support future research.","short_abstract":"Most recommender systems treat timestamps as numeric or cyclical values, overlooking real-world context such as holidays, events, and seasonal patterns. We propose a scalable framework that uses large language models (LLMs) to generate geo-temporal embeddings from only a timestamp and coarse location, capturing holiday...","url_abs":"https://arxiv.org/abs/2510.24430","url_pdf":"https://arxiv.org/pdf/2510.24430v1","authors":"[\"Yejin Kim\",\"Shaghayegh Agah\",\"Mayur Nankani\",\"Neeraj Sharma\",\"Feifei Peng\",\"Maria Peifer\",\"Sardar Hamidian\",\"H Howie Huang\"]","published":"2025-10-28T13:57:23Z","proceeding":"cs.IR","tasks":"[\"cs.IR\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false}