中文摘要：

微博作为企业网络舆情传播的主要载体和平台，使信息传播更迅速，也使舆情事件的影响力剧增。提出基于微博的企业网络舆情热度的计算公式。针对企业网络舆情热度的波动范围较大，并会呈现某种趋势的特点，提出了一种基于马尔可夫链的舆情热度趋势分析模型。该模型采集相关舆情的指标数据，得到热度的时间序列值；划分舆情热度趋势的状态空间；构建状态转移矩阵；预测热度的趋势变化区间。实验表明，该方法能有效地预测企业网络舆情的走势，进而为企业进行网络舆情管理和网络危机公关活动，以及管理效果评估提供理论依据。

英文摘要：

Enterprise Interact public opinion often has a hidden trend behind its characteristics shown online. A Markov chain-basext Inter- net public opinion analysis method is proposed. We collect microblog data about enterprise Interact public opinion to get the value of time series for the trend. Then the trend changes and the state spaces are analyzed, and the state transition matrix for the final trend is built. Experiments demonstrate that this method can effectively predict the trend of enterprise Intemet public opinion, and is significant for the enterprise to guide and controU its Interact public opinion.