Opinion-aware Information Diffusion Model Based on Multivariate Marked Hawkes Process

Haoming Zhang,Yiping Yao,Wenjie Tang,Jiefan Zhu,Yonghua Zhang
DOI: https://doi.org/10.1016/j.knosys.2023.110883
IF: 8.139
2023-01-01
Knowledge-Based Systems
Abstract:Currently, information diffusion on online platforms plays an important role in social governance. Opinions contained in information have a significant influence on information diffusion. However, both classic and extended diffusion models lack consideration of the influence of opinions. Therefore, in this study, we propose an opinion-aware information diffusion model based on the multivariate marked Hawkes process to simulate the information diffusion process with effect of opinions. In addition, to simulate this process more accurately, we consider various endogenous and exogenous stimuli apart from the opinion. Simulation results show that the diffusion trend prediction evaluation metrics of our model are 31.4% (on average) better than the contemporary models tested. Moreover, the effects of internal and external factors are found to be qualitatively similar to the real information diffusion process; hence, we gain useful inspiration for social network research and public opinion guidance.
What problem does this paper attempt to address?