发表机构
Spotify（Spotify）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文提出文本化用户品味系统，从收听行为生成自然语言品味画像并部署于Spotify，集成行为嵌入后提升MRR 0.6%和NDCG@7 2.2%，作为可解释、可引导的用户上下文接口。
AI 中文摘要
基础模型推荐系统需要能够被大型语言模型消费、推理，并通过自然语言交互进行优化的用户上下文。传统的行为嵌入向量在检索和排序中仍然非常有效，但它们对用户不透明，且并非原生适用于语言模型工作流。我们提出了文本化用户品味（Textual User Taste），一个从收听行为、交互信号、内容元数据和可选的用户反馈中生成结构化自然语言品味画像的系统，并将其部署给数百万Spotify用户。我们描述了在工业规模下生成、评估、优化和维护这些表示所需的端到端生产生命周期，包括提示词开发与压缩、用户引导以及与下游个性化系统的集成。由于不存在唯一的真实品味画像，我们引入了一个多方面的评估框架，将品味画像作为生产表示进行评估：它们独立携带用户特定的预测信号，并且在与行为嵌入集成时，在未来曲目预测中将MRR提高0.6%，在搜索排序中将NDCG@7提高2.2%。我们的评估还揭示，品味画像支持积极的自然语言引导，同时暴露了重要局限性，包括在否定表达和短期时间适应方面的挑战。这些发现将品味画像定位为行为嵌入的替代品，而是作为不断演化的用户上下文与基础模型推荐系统之间的可解释、可引导的接口。
英文摘要
Foundation model recommender systems require user context that can be consumed by large language models, reasoned over, and refined through natural-language interaction. Traditional behavioral embedding vectors remain highly effective for retrieval and ranking, but they are opaque to users and not natively expressed for language model workflows. We present Textual User Taste, a system that generates structured natural-language taste profiles from listening behavior, interaction signals, content metadata, and optional user feedback, and deploys them to millions of Spotify users. We describe the end-to-end production lifecycle required to generate, evaluate, optimize, and maintain these representations at industrial scale, including prompt development and compression, user steering, and integration with downstream personalization systems. Because no unique ground-truth taste profile exists, we introduce a multi-faceted evaluation framework to evaluate taste profiles as a production representation: they carry user-specific predictive signal independently, and when integrated with behavioral embeddings, improve MRR by 0.6% for future-track prediction and NDCG@7 by 2.2% for search ranking. Our evaluation also reveals that taste profiles support positive natural-language steering, while exposing important limitations, including challenges with negation and short-term temporal adaptation. These findings position taste profiles not as replacements for behavioral embeddings, but as an interpretable and steerable interface between evolving user context and foundation-model recommender systems.
