[目的/意義]文章以微博健康信息的信息特征為依據提取預測指標,采用支持向量機方法對微博健康信息的采納水平進行預測研究。[方法/過程]根據"垂直V影響力榜"選取了10個微博賬號進行數據采集,從信息的形式、內容和來源三個方面提取了10個預測特征指標,并制定相關的編碼規則生成樣本數據集,利用支持向量機進行預測模型的訓練,并對預測模型的效果進行檢驗。[結果/結論]研究結果顯示以微博健康信息的信息特征作為預測指標能夠有效地實現其采納水平的預測,且準確率較高。同時,10個預測指標在微博健康信息點贊和轉發采納水平預測中的貢獻率各有異同。

[Purposes/significance] The article uses the support vector machine method to predict the information adoption level of the health information on Sina Micro-blog basing on the information characteristics.[Method/process] According to the "Vertical V Influence List",10 micro-blog accounts are selected for data collection.Ten predictive indicators are extracted from three aspects of the information which are form,content and source.Related coding rules are formulated to generate the sample data sets.Support Vector Machine is used to train the prediction model and the effect of the model is examined subsequently.[Result/conclusion] The results show that the information adoption level can be predicted effectively by using the information characteristics,and the accuracy is high.At the same time,the contribution rates of 10 predictive indicators in the prediction of two different information adoption levels are different.