arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2603.11955cs.CL
收藏

PersonaTrace: 基于LLM代理的数字足迹合成

PersonaTrace: Synthesizing Realistic Digital Footprints with LLM Agents

Minjia Wang, Yunfeng Wang, Xiao Ma, Dexin Lv, Qifan Guo, Lynn Zheng, Benliang Wang, Lei Wang, Jiannan Li, Yongwei Xing, David Xu, Zheng Sun

AI总结:

PersonaTrace通过LLM代理生成多样且逼真的数字足迹,提升数据集的多样性和真实性,从而增强模型在真实任务中的表现。

摘要:

数字足迹(个体与数字系统交互的记录)对于研究行为、开发个性化应用和训练机器学习模型至关重要。然而,该领域的研究常常受到多样化和可访问数据稀缺的阻碍。为了解决这一限制,我们提出了一种新的方法,利用大型语言模型(LLM)代理来合成逼真的数字足迹。从结构化的用户档案开始,我们的方法生成多样且合理的用户事件序列,最终生成相应的数字制品,如电子邮件、信息、日历条目、提醒等。内在评估结果表明,生成的数据集比现有基线更加多样和真实。此外,在真实世界非分布任务上评估时,基于我们合成数据微调的模型优于其他合成数据集训练的模型。

英文摘要:

Digital footprints (records of individuals' interactions with digital systems) are essential for studying behavior, developing personalized applications, and training machine learning models. However, research in this area is often hindered by the scarcity of diverse and accessible data. To address this limitation, we propose a novel method for synthesizing realistic digital footprints using large language model (LLM) agents. Starting from a structured user profile, our approach generates diverse and plausible sequences of user events, ultimately producing corresponding digital artifacts such as emails, messages, calendar entries, reminders, etc. Intrinsic evaluation results demonstrate that the generated dataset is more diverse and realistic than existing baselines. Moreover, models fine-tuned on our synthetic data outperform those trained on other synthetic datasets when evaluated on real-world out-of-distribution tasks.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑