Information Diffusion Prediction via Cascade-Retrieved In-context Learning

Conference Paper (2024)
Author(s)

Ting Zhong (Kash Institute of Electronics and Information Industry, University of Electronic Science and Technology of China)

Jienan Zhang (University of Electronic Science and Technology of China)

Zhangtao Cheng (University of Electronic Science and Technology of China)

Fan Zhou (Intelligent Terminal Key Laboratory of Sichuan Province, University of Electronic Science and Technology of China)

X. Chen (TU Delft - Sanitary Engineering)

Research Group
Sanitary Engineering
DOI related publication
https://doi.org/10.1145/3626772.3657909
More Info
expand_more
Publication Year
2024
Language
English
Research Group
Sanitary Engineering
Pages (from-to)
2472-2476
ISBN (electronic)
9798400704314
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

Information diffusion prediction, which aims to infer the infected behavior of individual users during information spread, is critical for understanding the dynamics of information propagation and users' influence on online social media. To date, existing methods either focus on capturing limited contextual information from a single cascade, overlooking the potentially complex dependencies across different cascades, or they are committed to improving model performance by using intricate technologies to extract additional features as supplements to user representations, neglecting the drift of model performance across different platforms. To address these limitations, we propose a novel framework called CARE (CAscade-REtrieved In-Context Learning) inspired by the concept of in-context learning in LLMs. Specifically, CARE first constructs a prompts pool derived from historical cascades, then utilizes ranking-based search engine techniques to retrieve prompts with similar patterns based on the query. Moreover, CARE also introduces two augmentation strategies alongside social relationship enhancement to enrich the input context. Finally, the transformed query-cascade representation from a GPT-type architecture is projected to obtain the prediction. Experiments on real-world datasets from various platforms show that CARE outperforms state-of-the-art baselines in terms of effectiveness and robustness in information diffusion prediction.

Files

3626772.3657909.pdf
(pdf | 1.71 Mb)
- Embargo expired in 11-01-2025
License info not available