摘要
为了分析疫情期间网民情绪变化趋势,笔者爬取新冠疫情爆发初期的微博相关讨论话题,应用百度PaddlePaddle深度学习平台,通过预训练模型ERNIE结合fine-tune实现迁移学习,并构建基于Paddlepaddle的文本情感分析模型,有助于政府部门做好科学防控和舆情引导工作。
In order to analyze the trend of netizens’emotions during the epidemic,the author crawled Weibo related discussion topics in the early stage of the outbreak of the new crown epidemic,applied about:blank PaddlePaddle deep learning platform,realized migration learning through pre-training model ERNIE combined with fine-tune,and constructed text emotion based on Paddlepaddle The analysis model helps government departments to do a good job in scientific prevention and control and public opinion guidance.
作者
郑耀东
李旭峰
ZHENG Yaodong;LI Xufeng(Computer Science Department,Software Engineering Institute of Guangzhou,Guangzhou Guangdong 510990,China)
出处
《信息与电脑》
2021年第7期189-191,共3页
Information & Computer
