【摘要】 信息技术的加速迭代，形成了以人或虚拟数字人为中心，融合视频、声音、文字等多种模态信号的主持传播新模式，专业主持传播与社会化传播之间的边界变得逐渐模糊，主持传播内容同质化和传播穿透力弱化的特征日渐突出。因此，本文研究提升主持传播效能和引导力的方法。具体而言，基于计算社会科学的理论，首先提出了计算主持传播范式，即以主持传播多模态自洽呈现为关键核心，实现多模态信息的有效协同；其次，以主持传播多时空多尺度精准分析和主持传播效能有效度量为重要支撑，研究多模态信息融合方法，实现传播内容和效能的差异化、细粒度感知。这一研究为数智化时代的主持传播提供了定量化的理论方法、框架和指导准则。更多还原

【Abstract】 The accelerated iteration of information technology has formed a new mode of hosted communication. The mode is centered on human and virtual digital human, and integrates multiple modal signals such as video, sound, and text. The boundary between professional and general hosted communication has gradually become blurred. The content homogenization and weakening of communication penetration are becoming more and more prominent. Therefore, this paper studies the methods to improve the effectiveness of hosted communication. Specifically, based on the theory of computational social science, this study firstly proposes the paradigm of computational hosted communication. The multimodal self-consistent presentation of hosted communication is the key to realize the effective coordination of multimodal information. Multi-temporal and multiscale accurate analysis and effective measurement of hosted communication effectiveness are important supports. Multi-modal information fusion methods are studied to realize the differentiation and fine-grained perception of communication content and effectiveness. This research provides quantitative theoretical methods, frameworks, and guiding principles for host communication in the era of digital intelligence.更多还原