[目的/意義]建立一套行之有效的網絡輿情預測和分級管理系統對提高政府網絡監管效率具有重大意義。[方法/過程]分時段記錄網絡輿情事件的新浪微指數、百度指數、頭條指數作為事件熱度的衡量指標,運用EGM(1.1)模型對輿情事件的發展趨勢進行預測,并在預測數據基礎上,運用灰色關聯分析方法,提出網絡輿情事件分級方案。綜合運用上述兩模型確立的輿情監管體系對"江歌案宣判""上海攜程親子園虐童案""莫煥晶案二審宣判"等事件進行分析評價。[結果/結論]模型在處理輿情管理問題方面具有良好的適用性和精確性。結合輿情事件特點與分析結果,向政府及公司危機公關部門等提出了相應建議。

[Purpose/significance]Establishing an effective system to predict and grade network public opinion has great significance in improving the efficiency of governmental network supervision.[Method/process]First,record Sina micro index,Baidu index,Toutiao index of network public opinion events hour by hour which will be used as the measurement indices of event concern degree.Then EGM(1.1) Model is used to predict the development trend of public opinion events and grey relational analysis is applied to propose a grading scheme for network public opinion events based on the predicted data.The paper analyzes and evaluates typical events such as the"Jiang Ge case""Shanghai Xiecheng child abusing case""Mo Huanjing case"with the public opinion supervision system established by the two models above.[Result/conclusion]Established models have great applicability and precision in dealing with the issues of public opinion management.Combined with the characteristics of public opinion events and the system analysis results,suggestions oriented the government and the company’s public relation departments have been made in the paper.