基于超级网络理论的谣言检测模型研究

Research on Rumor Detection Model Based on Super Network Theory

针对社交网络中的谣言防控问题,基于超级网络理论提出了一种谣言检测模型.建立了一个由三层子网络构建的超级网络,通过融合用户社交关系、心理情感和关键词来准确刻画微博帖子的特征.提出了基于超级网络的谣言检测模型,应用主流的机器学习方法来训练分类器,以实现谣言的分类检测.从微博获取数据来构建实验数据集,并将本模型与现有的检测模型进行比较.结果表明,本模型在谣言检测方面能够表现出最好的性能.

Aiming at the problem of rumor prevention and control in social networks,a rumor detection model based on super network theory is proposed in this paper.First of all,a super network constructed by a three-layer sub network is established to acurately depict the characteristics of microblog posts by integrating users’ social relations,psychological feelings and keywords.Then,a rumor detection model based on the super network is proposed,and the mainstream machine learning methods are used to train the classifier to achieve the classification and detection of rumors.Finally,the experimental data set is constructed from the data obtained from microblog,and the model is compared with the existing detection model.The results show that this model can show the best performance in rumor detection.