基于LDA主题模型对AIGC的影响力分析  

Influence Analysis of AIGC Based on LDA Topic Model

作　　者：翟其玲 张佳怡 刘宝瑞 黄龙彬 陈立伟 金秀玲[1] 

机构地区：[1]闽江学院数学与数据科学学院,福建 福州

出　　处：《数据挖掘》2023年第4期366-375,共10页Hans Journal of Data Mining

摘　　要：ChatGPT的出现,引领生成式AI迅速渗透进人们生产生活的方方面面,故了解人工智能内容生成的发展趋势和未来方向非常重要。基于抓取微博平台收录的关于人工智能内容生成的文本内容,先利用jieba分词进行文本预处理,通过LDA主题模型的方法得到七个人工智能内容生成领域的热点主题,包括:人工智能概念股、智能数字内容创作、数字经济下的科技股投资趋势、自然语言技术的突破等。研究揭示了当前AIGC的舆情热点,有助于相关人员掌握生成式AI领域的发展状态,对人工智能未来的应用提供参考。The emergence of ChatGPT has led the rapid penetration of generative AI into all aspects of people’s production and life, so it is very important to understand the development trend and future direction of artificial intelligence content generation. Based on grabbing the text content of artificial intelligence content generation collected on Weibo platform, jieba segmentation is first used for text pre-processing, and seven hot topics in the field of artificial intelligence content generation are obtained by LDA topic model, including: Artificial intelligence concept stocks, intelligent digital con-tent creation, technology stock investment trends under the digital economy, breakthroughs in natural language technology, etc. The research reveals the current public opinion hot spots of AIGC, which is helpful for relevant personnel to grasp the development status of the field of generative AI and provide reference for the future application of artificial intelligence.

关 键 词：ChatGPT LDA模型 人工智能内容生成 热点分析 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]