工业推荐系统已经变得越来越复杂，可能涉及诸如电子商务产品和用户生成的内容的\emph{diverse domains}，并且可以包括诸如检索、排名、解释生成甚至AI辅助的内容制作的\emph{a myriad of tasks}。到目前为止，主流的方法是为每个领域和每个任务开发单独的算法。在本文中，我们探讨了开发一个统一的基础模型，以支持\emph{open-ended domains and tasks}在工业推荐系统，这可能会减少对下游设置的数据的需求，并可以通过避免训练一个单独的模型从头开始为每一个任务的碳足迹最小化的可能性。由于（i）潜在的无限下游域和任务集，以及（ii）现实世界系统对计算效率的强调，导出统一的基础是具有挑战性的。因此，我们以M6为基础，M6是一种现有的大规模工业预训练语言模型，类似于GPT-3和T5，并利用M6的预训练能力进行样本高效的下游适应，将用户行为数据表示为纯文本，并将任务转换为语言理解或生成。为了应对紧张的硬件预算，我们提出了一个改进版本的提示微调，优于微调与可忽略的1\%的任务特定的参数，并采用技术，如后期交互，提前退出，参数共享，修剪，以进一步减少推理时间和模型的大小。我们展示了基础模型的多功能性，广泛的任务，如检索，排名，zero-shot推荐，解释生成，个性化内容创建和会话推荐，并设法将其部署在云服务器和移动设备。
英文摘要：
Industrial recommender systems have been growing increasingly complex, may involve \emph{diverse domains} such as e-commerce products and user-generated contents, and can comprise \emph{a myriad of tasks} such as retrieval, ranking, explanation generation, and even AI-assisted content production. The mainstream approach so far is to develop individual algorithms for each domain and each task. In this paper, we explore the possibility of developing a unified foundation model to support \emph{open-ended domains and tasks} in an industrial recommender system, which may reduce the demand on downstream settings' data and can minimize the carbon footprint by avoiding training a separate model from scratch for every task. Deriving a unified foundation is challenging due to (i) the potentially unlimited set of downstream domains and tasks, and (ii) the real-world systems' emphasis on computational efficiency. We thus build our foundation upon M6, an existing large-scale industrial pretrained language model similar to GPT-3 and T5, and leverage M6's pretrained ability for sample-efficient downstream adaptation, by representing user behavior data as plain texts and converting the tasks to either language understanding or generation. To deal with a tight hardware budget, we propose an improved version of prompt tuning that outperforms fine-tuning with negligible 1\% task-specific parameters, and employ techniques such as late interaction, early exiting, parameter sharing, and pruning to further reduce the inference time and the model size. We demonstrate the foundation model's versatility on a wide range of tasks such as retrieval, ranking, zero-shot recommendation, explanation generation, personalized content creation, and conversational recommendation, and manage to deploy it on both cloud servers and mobile devices.
Toward Holistic Evaluation of Recommender Systems Powered by Generative Models
Polytechnic University of Bari（巴里理工大学）；Duke University（杜克大学）；Bespoke Labs（Bespoke实验室）；ETH Zurich（苏黎世联邦理工学院）；Autónoma University of Madrid（马德里自治大学）；University of California, San Diego（加州大学圣地亚哥分校）
Do recommender systems function in the health domain: a system review
School of Computer Science and Technology, Harbin Institute of Technology（哈尔滨工业大学计算机科学与技术学院）；The 1st Affiliated Hospital of Harbin Medical University（哈尔滨医科大学附属第一医院）
