摘要
微博是现下热门的短文本社交软件之一,笔者以疫情相关的微博内容评论作为研究切入点,将机器学习运用在情感分析构建分类器的过程中,对比不同条件下分类器的准确率,得出在特征维度为2 500时,运用结巴分词和多项式朴素贝叶斯方法的组合构建分类器,分类效果最优,对舆情检测与控制提供了参考意义。
Weibo is one of the essay our pay now popular software,based on the epidemic situation of related weibo content comments as a research starting point,applying machine learning in sentiment analysis in the process of constructing classifier,compared the accuracy in classifier under different conditions,it is concluded that when the characteristic dimension for 2 500,using the stuttering participle and polynomial naive bayesian method is a combination of constructing classifier,the optimal classification effect,provides reference to inspect and control the public opinion.
作者
胡梦雅
樊重俊
朱玥
Hu Mengya;Fan Chongjun;Zhu Yue(Business School,University of Shanghai for Science and Technology,Shanghai 200093,China)
出处
《信息与电脑》
2020年第12期71-73,共3页
Information & Computer
关键词
文本情感分析
机器学习
朴素贝叶斯
支持向量机
逻辑回归
text emotion analysis
machine learning
naive bayes
support vector machine
logistic regression
