首页|基于主题情感的ChatGPT用户在线评论分析——以bilibili平台为例

基于主题情感的ChatGPT用户在线评论分析——以bilibili平台为例

扫码查看
[目的/意义]在AIGC迅速崛起的背景下,ChatGPT空前火热,引发公众的广泛讨论,用户评论可以反映出当前ChatGPT的应用现状,以便为科学合理地使用该技术提供参考.[方法/过程]通过爬取bilibili平台相关视频3568条评论内容,利用Python构建LDA模型提取文本主题,分析各主题的情感倾向,并在此基础上分别构建积极、消极倾向的语义网络,从中发现ChatGPT应用过程中的优缺点.[结果/结论]评论数据可以概括为应用领域、技术能力、科技影响、内容输出四个主题,发现ChatGPT的应用对就业竞争产生冲击,且其语义技术层面存在缺陷,输出结果的准确性和真实性有待商榷,以及用户对ChatGPT存在依赖心理的问题,并提出相应的发展建议.
Analysis of ChatGPT Users'Online Comments Based on Topic Emotion:Case Study of Bilibili Platform
[Purpose/significance]In the context of the rapid rise of AIGC,ChatGPT is unprecedentedly hot,causing extensive public discussion.User comments can reflect the current application status of ChatGPT,so as to provide references for scientific and reasonable use of the technology.[Method/process]By crawling 3568 comments from videos related to bilibili platform,this paper uses Python to build an LDA model to extract text topics,analyzes the emotional tendencies of each topic,and builds semantic networks with positive and negative tendencies respectively on this basis to find the advantages and disadvantages of ChatGPT application.[Result/conclusion]This paper summarizes the review data into four themes:application field,technical capability,scientific and technological impact,and content output.It is found that the application of ChatGPT has an impact on employment competition,and the aspect of semantic technology has defects,the accuracy and authenticity of the output results are worthy of discussion,and users are psychologically dependent on ChatGPT.Finally,it puts forward the corresponding development suggestions.

ChatGPTcomment miningLDA modelsentiment analysis

侯裕馨

展开 >

黑龙江大学信息管理学院 黑龙江哈尔滨 150080

ChatGPT 评论挖掘 LDA模型 情感分析

黑龙江省教育科学规划课题重点项目

GJB1423263

2024

情报探索
福建省科技情报学会,福建省科技信息研究所

情报探索

CHSSCD
影响因子:0.52
ISSN:1005-8095
年,卷(期):2024.(3)
  • 17