The proliferation of rumors on social media has become a critical cybersecurity issue, attracting widespread attention across various sectors. Traditional rumor detection methods often neglect the social context of rumor initiators and fail to adequately address the fine-grained temporal features inherent in the propagation process. To overcome these limitations, we propose an innovative rumor detection method, which leverages graph neural networks and temporal encoders to integrate the social information of rumor initiators with the temporal characteristics of rumor dissemination. This approach enables the simultaneous modeling of temporal attributes and the social dimensions of rumor initiators within a unified framework. Experimental results on three real-world datasets demonstrate that our method outperforms state-of-the-art models.