社会化媒体让消费者拥有了更多自我表达的主权，也迁移并拓展了现实世界的社交关系，让他们能够自由地加入各种基于人际、兴趣和归属等多重关系的网络之中，组成形形色色的圈子和群体进行社会互动，产生了丰富的群体创造内容。群体中的内容和网络是共同演化的，具有内容时序性、用户自演性、关联多重性和结构动态性等特征。已有文献侧重于分析个体层面的用户创造内容，较少关注群体创造内容。综合动态话题模型、情感分析、超图分析、计量经济学模型和行为实验等方法，本项目（1）将单一关系的群体检测扩展至多重动态关系的群体检测；（2）将用户偏向性内容的形成延伸至群体多样性内容的形成；（3）将内容的传播过程扩充至内容的收敛过程；（4）将创造者的影响拓展至沉默者的影响。研究结论有助于认识群体创造内容，有助于理解群体的行为规律和分析范式，有助于提升社会化营销的效果。

Social media enables consumers to get more empowerment on self-expression, and also migrates and extends offline social relations. Consumers could feel free to join in multi-mode networks based on social relations, interests and affiliations to organize cliques and groups. The social interactions in groups lead to group-generated content (GGC). The content and network of GGC are co-evolution with content sequence, user self-evolution, multiple relations, and dynamic structure. The existing literature focus on user-generated content, and pay less attention on GGC. Based on dynamic topic model, sentiment analysis, hyper-network analysis, econometric modeling, and behavioral experiment, this proposal will (1) enrich group detection from signal relationship to multiple dynamic relationships; (2) extend generation of content valence to the formation of diversified content; (3) expand content communication process to convergence process; and (4) enlarge content effects of contributors to silence users. The research findings could perfect the theories of GGC, understand the behavior rules and analysis patterns of group, and enhance the effectiveness of group marketing.