基于情感分析的网络舆情主题演化分析 ——以双减政策为例

Internet Public Opinion Theme Evolution Analysis Based on Emotion Analysis—Taking the Double Reduction Policy as an Example

摘要: 本文通过python编写爬虫程序，收集有关双减政策的微博评论并通过snowNLP对其情感分析，在积极与消极情感分别通过LDA模型做主题分析，得出了“政策实施效果”、“时间分配”、“教培老师转行”、“疫情开学”、“教育公平”、“教师待遇”、“家长陪伴”、“学业减负”、“兴趣培养”、“人才培养”、“教育改革”等13个主题关键词。通过分析得知，双减政策对于处于义务教育阶段学生的学习模式有较大的改变，也对课外教育机构进行了有效的打压。可以看出双减政策的强势性。从长远的角度看，该项政策有效的推动了人才强国战略的实施，有助于教育的良好发展。

Abstract: In this paper, a crawler program was written by Python to collect Weibo comments about the double-reduction policy and analyze their emotions through snowNLP. In terms of positive and negative emotions, subject analysis was conducted by LDA model. Thirteen key words were obtained, including “policy implementation effect”, “time distribution”, “teachers changing careers”, “epidemic reopening”, “education equity”, “teachers’ treatment”, “parents’ companionship”, “academic burden reduction”, “interest cultivation”, “talent cultivation” and “education reform”. Through the analysis, the double-reduction policy has greatly changed the learning mode of students in the compulsory education stage, and implemented a strong pressure on extracurricular education institutions. We can see the strength of the double-reduction policy. However, in the long run, this policy effectively promotes the implementation of the strategy of reinvigorating the country with talents and con-tributes to the sound development of education.