Prompt Learning on Temporal Interaction Graphs

TL;DR

This paper introduces TIGPrompt, a novel prompt-based framework for temporal interaction graphs that addresses temporal and semantic gaps in pre-training, achieving state-of-the-art performance with high efficiency.

Contribution

The paper proposes TIGPrompt, a versatile, temporally-aware prompting framework for TIGs, with minimal supervision and flexible pre-train, prompt-based fine-tuning paradigms.

Findings

TIGPrompt achieves state-of-the-art results on multiple TIG benchmarks.

The framework demonstrates significant efficiency improvements over existing models.

Extensive experiments validate the effectiveness of temporally-aware prompts.

Abstract

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.