فهرست نشریات علمی دسترسی آزاد ایران

This paper addresses the critical challenge of identifying influential users in online social networks, with a specific focus on X (formerly Twitter). Existing methods for influence detection often rely on simplistic metrics, such as follower counts or retweet ratios, which fail to account for the nuanced dynamics of information spread. Alternatively, some approaches employ overly complex machine learning or graph-theoretical models that, while theoretically sound, suffer from high computational costs and limited interpretability. To bridge this gap, we proposed a novel and scalable approach based on linear systems theory to model information propagation within social networks. By representing the network as a linear dynamical system, we derived influence scores by analyzing the system's state and transition matrices, which captured the temporal evolution of user interactions. This method not only offers computational efficiency but also provides a theoretically grounded framework for understanding influence by considering the broader systemic context of information flow. Unlike traditional methods, our approach considers indirect influence pathways and the cumulative effects of user interactions over time. To validate our model, we conduct extensive experiments using a real-world X dataset, comparing our results against established baselines. Our findings demonstrate that the proposed method effectively identifies influential users while maintaining interpretability and scalability. This work contributes to the growing body of research on social network analysis by offering a balanced solution that strikes a balance between accuracy and computational feasibility, with potential applications in viral marketing, misinformation detection, and community engagement strategies.