Beyond Leakage and Complexity: Towards Realistic and Efficient Information Cascade Prediction

ArXi:2510.25348v2 Announce Type: replace Information cascade popularity prediction is a key problem in analyzing content diffusion in social networks. However, current related works suffer from three critical limitations: (1) temporal leakage in current evaluation--random cascade-based splits allow models to access future information, yielding unrealistic results; (2) feature-poor datasets that lack downstream conversion signals (e.g., likes, comments, or purchases), which limits practical applications; (3) computational inefficiency of complex graph-based methods that require days of