A certain topic in Microblogging arouses strong attention, along with the development of the situation, the information in regard to this topic begins to deviate from the subject, the wrong viewpoints and the untrue reports permeate the Microblogging information platform through forwarding, and the topic evolves into rumours in the process of dissemination. Microblogging rumour has the characteristics of increasing dissemination speed and scope, and the growing difficulty in dissemination controlling process. In the paper we propose a method of detecting the evolution of Microblogging topic to rumours during the dissemination process, which is based on BP neural network and by improving its excitation function and introducing the impulse item at the same time. Experimental results prove that the algorithm has fairly good test effect.%微博中某一话题引起强烈关注,随着事态的发展,关于此话题的信息也开始偏轨,错误观点、失实报道经转发充斥了微博信息平台,话题在传播过程中演变出谣言.微博谣言在传播过程中具有传播速度越来越快、传播范围越来越广、传播过程越来越难以控制等特点.提出基于BP神经网络模型及改进其激发函数,同时引入冲量项,对微博话题在传播过程中演变为谣言进行检测.实验结果表明,该算法具有较好的检测效果.
展开▼
