Public Perception of New Generation Artificial Intelligence Technology from a Multidimensional Perspective
[Research purpose]Social media comments are a vital source for analyzing public perception towards new technological appli-cations.To overcome the randomness flaw in traditional topic mining techniques and the limitations of singular sentiment analysis methods,there is an urgent need to develop new topic model approaches and sentiment analysis tools.These advancements aim to enhance the preci-sion of text data quantitative analysis and achieve better visualization effects.[Research method]By developing a structurally integrated deep learning model,the BERT-LDA model,which combines BERT and LDA,this study focuses on ChatGPT's social media comments.It utilizes BERT and LDA to extract complex semantic information and key topics from texts,enabling the discovery of deeply hidden the-matic features.Furthermore,based on BERT sentiment analysis,this research designs a multi-dimensional visualization analysis of senti-ment evolution from overall,thematic,and attitudinal perspectives.[Research conclusion]The findings reveal that:the BERT-LDA model efficiently processes large-scale,short-text,unstructured social media comment data,successfully identifying public attitudes to-wards ChatGPT's impact across various domains such as employment,education,future development,product innovation,and technologi-cal transformation.Compared to traditional models like LDA and TF-IDF,the BERT-LDA model demonstrates superior topic identifica-tion and generalization capabilities,particularly in accurately mining key topics and important terms.Public attitudes towards ChatGPT are not uniform,exhibiting a complex mixture of praise and skepticism.
artificial intelligenceChatGPTWeibocomment texttopic miningsentiment analysispublic perceptionBERT-LDA model
万方数据
维普
