arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.06326cs.AIcs.LGcs.SI
收藏

时序交互图上的提示学习

Prompt Learning on Temporal Interaction Graphs

Fudan University（复旦大学）；Shanghai Key Laboratory of Data Science（上海市数据科学重点实验室）；School of Computer Science（计算机学院）；IFM Lab（IFM实验室）；The Chinese University of Hong Kong（香港中文大学）；Ant Group（蚂蚁集团）

Xi Chen, Siwei Zhang, Yun Xiong, Xixi Wu, Jiawei Zhang, Xiangguo Sun, Yao Zhang, Feng Zhao, Yulin Kang

备注：
11 pages, 8 figures
阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对时序交互图现有“预训练-预测”范式存在的时序差异与语义分歧问题，提出通用框架TIGPrompt，通过时序提示生成器弥合两类差距，支持扩展范式，实现了当前最优性能与效率优势。

摘要：

时序交互图（Temporal Interaction Graphs, TIGs）被广泛用于表征现实世界系统。为推进TIG上的表征学习，研究人员已提出一系列TIG模型。然而，这些模型在其“预训练-预测”训练范式下，预训练与下游预测之间仍存在两个难以弥合的差距。其一，预训练与推理数据之间的时序差异严重削弱了模型在动态演化数据上进行远期未来预测的适用性。其二，前置任务与下游任务之间的语义分歧阻碍了其实际应用，因为这些模型难以在不同应用场景下对齐自身的学习与预测能力。\n 近年来，“预训练-提示”范式作为一种轻量化的模型泛化机制应运而生。应用该范式是解决上述挑战的一种潜在方案。然而，将这一范式适配到TIG上并非易事。静态图场景下的提示应用未考虑时间敏感的动态性且表达能力不足，无法满足时序场景的需求。为解决这一问题，我们提出时序交互图提示框架（Temporal Interaction Graph Prompting, TIGPrompt），这是一个可与各类TIG模型无缝集成的通用框架，能够同时弥合时序差距与语义差距。具体而言，我们提出了一种时序提示生成器，可为不同任务提供具备时序感知能力的提示。这些提示的突出特点是设计极简，仅需通过极少量监督数据微调提示生成器即可实现。为满足不同的计算资源需求，我们提出了扩展的“预训练-基于提示的微调”范式，具备更高的灵活性。通过大量实验，TIGPrompt展现出了SOTA（state-of-the-art，当前最优）性能与显著的效率优势。

英文摘要：

Temporal Interaction Graphs (TIGs) are widely utilized to represent real-world systems. To facilitate representation learning on TIGs, researchers have proposed a series of TIG models. However, these models are still facing two tough gaps between the pre-training and downstream predictions in their ``pre-train, predict'' training paradigm. First, the temporal discrepancy between the pre-training and inference data severely undermines the models' applicability in distant future predictions on the dynamically evolving data. Second, the semantic divergence between pretext and downstream tasks hinders their practical applications, as they struggle to align with their learning and prediction capabilities across application scenarios. Recently, the ``pre-train, prompt'' paradigm has emerged as a lightweight mechanism for model generalization. Applying this paradigm is a potential solution to solve the aforementioned challenges. However, the adaptation of this paradigm to TIGs is not straightforward. The application of prompting in static graph contexts falls short in temporal settings due to a lack of consideration for time-sensitive dynamics and a deficiency in expressive power. To address this issue, we introduce Temporal Interaction Graph Prompting (TIGPrompt), a versatile framework that seamlessly integrates with TIG models, bridging both the temporal and semantic gaps. In detail, we propose a temporal prompt generator to offer temporally-aware prompts for different tasks. These prompts stand out for their minimalistic design, relying solely on the tuning of the prompt generator with very little supervision data. To cater to varying computational resource demands, we propose an extended ``pre-train, prompt-based fine-tune'' paradigm, offering greater flexibility. Through extensive experiments, the TIGPrompt demonstrates the SOTA performance and remarkable efficiency advantages.

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。

↑