基于用户评论情感分析的医院评分模型构建

Establishment of hospital scoring model based on emotion analysis of users' reviews

目的:构建基于用户评论情感分析的医院评分模型。方法:运用ROST EA工具对文本评论进行分析和利用相关分析对星级评分与文本评论的拟合程度进行分析,构建基于线性回归的文本评论自动打分模型,对星级评分与文本评论的相关性进行检验。结果:当情感强度的比值是“一般∶中度∶高度=1∶1∶1”时,两者具有最高的相关性,相关性系数为0.680。结论:最终回归模型为A=0.961exp(E0)+7.348,星级评分可以在一定程度上代表文本评论,说明该模型具有一定应用价值。

Objective To establish the hospital scoring model based on emotion analysis of users'reviews.Methods The linear regression-based automatic scoring model of text reviews was established by analyzing the text reviews with ROST EA and the fitting degree of star scoring and text reviews by correlation analysis.The correlation between star scoring and text reviews was verified.Results The highest correlation coefficient of star scoring and text reviews was 0.680 when the emotion intensity ratio was"generality∶moderation∶height=1∶1∶1.Conclusion The final regression model is A=0.961exp(EO)+7.348,indicating that star scoring can represent the text reviews in a certain extent and the model is of a certain application value.