社交网络中基于情感模型的用户转发行为预测

摘要

Retweeting behavior on Twitter is the behavior that user reposts comments from their friends. Few studies have investigated in combining a user's subjectivity motivation with his conformity to environment to predict a user's retweeting behavior. Based on the sentiment analy-sis, combines a user's subjectivity motivation with a designed adoption model which measures a user's neighbors' influences, and then es-tablishes a mixture model with these two factors to do prediction. Evaluates the model in Twitter dataset to verify its prediction perfor-mance.%基于Twitter 的用户社会影响力的分析一直是社交网络分析的热点。然而很少有研究针对用户对微博的主题、情感、内容等语意信息的主观看法来预测用户的转贴行为。实现基于周围邻居对目标用户转发行为影响的用户适应性模型。并且实现基于语义分析的用户主观模型结合用户适应性模型的混合模型。设计实验证实用户对微博内容的主观看法极大影响预测结果。