社会大众如何看待生成式人工智能在教育中的应用?——对B站ChatGPT话题弹幕文本的舆情主题与情感分析

Public Perspective of Generative AI in Education——Opinion Topic and Sentiment Analysis of Bilibili’s Bullet Comments on ChatGPT

文章从社会大众角度出发,通过网络舆情主题与情感分析等文本分析方法,对2023年1月-2024年3月Bilibili平台中“生成式人工智能/ChatGPT教育应用”的话题弹幕文本进行数据挖掘,以剖析社会大众对ChatGPT类生成式人工智能在教育应用中的情感态度、关注热点与看法意见,审视生成式人工智能在教育中的应用成效和潜在的伴生性风险。研究发现:社会大众对生成式人工智能教育应用总体持积极态度且整体情感占比趋于稳定;关注热点聚焦于学生学习、教师教学、教育公平、教育应用、态度观点,具有多极化倾向;看法意见集中在准确性、普适性、伦理性、异化性等方面,呈现多元化特点;可从学生、教师、企业、政府等层面进行优化。

Generative artificial intelligence(AI)is leading large-scale AI models towards practical applications,which has a profound impact on education and other fields.From the perspective of the general public,this paper analyzes the bullet comment texts of the topic"Generative AI/ChatGPT education application"on the Bilibili platform from January 2023 to March 2024,using text analysis methods such as thematic and sentiment analysis of online public opinions.It aims to thoroughly anatomize the public’s emotions,attitudes,concerns,and opinions about the application of ChatGPT-like generative AI in education,and to examine its effectiveness and potential associated risks,thus providing new ideas and perspectives for its development.This study suggests that the general public has a positive sentiment toward generative AI educational applications and the overall sentiment ratio is stable;the main concerns are students’learning,teachers’teaching,educational fairness,educational application,attitudes and viewpoints with a multipolar tendency;public opinions are diversified,focusing on accuracy,universality,ethics and alienation.