人工智能技术在外卖舆情分析中的应用  

Research on the Application of Artificial Intelligence Technology in Takeaway Public Opinion Analysis

作　　者：文培煊 叶紫 潘笑天 章国道 谷山川 WEN Peixuan;YE Zi;PAN Xiaotian;ZHANG Guodao;GU Shanchuan(School of Media and Design,Hangzhou Dianzi University,Hangzhou,China,310018;School of Marxism,Zhejiang Institute of Economics and Trade,Hangzhou,China,310018)

机构地区：[1]杭州电子科技大学人文艺术与数字媒体学院,杭州310018 [2]浙江经贸职业技术学院马克思主义学院,杭州310018

出　　处：《福建电脑》2025年第4期26-30,共5页Journal of Fujian Computer

基　　金：“突发应急事件网络舆情传播机制研究”(No.Y202352761)资助。

摘　　要：为解决评论数据量和用户情感复杂性给用户情感分析带来的困难,实现快速有效地提取用户情感信息,本文设计了一种基于人工智能技术的外卖平台舆情分析应用系统。系统采用基于BERT的情感分析方法,通过融合多头注意力机制增强模型对文本关键信息的捕捉能力,针对外卖平台用户评论进行情感分类。实验的结果显示,BERT模型在情感分类任务中的表现优于传统模型。在真实的外卖评论数据集上,该方法的准确率可达91.12%以上,可帮助外卖平台及时掌握用户反馈,优化服务策略并提升商家决策效率。To solve the difficulties brought by the large amount of comment data and the complexity of user emotions to user sentiment analysis,and to achieve fast and effective extraction of user sentiment information,this paper designs an AI based public opinion analysis application system for food delivery platforms.The system adopts a sentiment analysis method based on BERT,which enhances the model's ability to capture key information in text by integrating multi head attention mechanism,and performs sentiment classification for user comments on food delivery platforms.The experimental results show that the BERT model performs better than traditional models in sentiment classification tasks.On a real food delivery review dataset,the accuracy of this method can reach over 91.12%,which can help food delivery platforms timely grasp user feedback,optimize service strategies,and improve merchant decision-making efficiency.

关 键 词：外卖平台 人工智能 情感分析 舆情分析 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]