%0 Journal Article
%T 基于陶瓷电商产品评论的情感分析研究<br>Research on Sentiment Analysis Based on Ceramic E-Commerce Product Reviews
%A 杨利华
%A 李鑫扬
%A 徐思雨
%A 严帅
%A 陈泽民
%J Software Engineering and Applications
%P 989-1000
%@ 2325-2278
%D 2023
%I Hans Publishing
%R 10.12677/SEA.2023.126097
%X 本文以天猫商城的景德镇红叶陶瓷青花瓷餐具盘碗碟套装为研究对象。通过爬取该研究对象的评论并对其进行文本挖掘预处理和词云绘制，采用基于BosonNLP和SnowNLP情感词典的两种情感倾向性分析方法进行情感分析和可视化展示，通过对比BosonNLP和SnowNLP两组方法得出的实验结果，从而为买家提供选购的参考标准，为卖家提供建设性的改进方向。此外，再进行相应的预处理后采用LDA主题模型进行分析。最后，利用交互可视化的方式把语句情感分数的分布情况再次直观展示，有效地验证了该产品情感偏向积极，得出了消费者可放心购买的实验结论。<br />
This article takes the Jingdezhen Red Leaf Ceramic Blue and White Porcelain Tableware, Plate, Dish Set of Tmall as the research object. By crawling through the comments on the Red Leaf Ceramic Jingdezhen Blue and White Porcelain Tableware, Dish and Dish Set on Tmall, and performing text mining preprocessing and wordle rendering on it, two sentiment analysis methods based on BosonNLP and SnowNLP sentiment dictionaries were used for sentiment analysis and visualization display. By comparing the experimental results obtained by BosonNLP and SnowNLP methods, reference standards for buyers' selection were provided, Provide constructive improvement directions for sellers. In addition, after corresponding preprocessing, the LDA topic model is adopted Conduct analysis. Finally, the distribution of sentence sentiment scores was visually displayed through interactive visualization, effectively verifying the positive emotional bias of the product and drawing experimental conclusions that consumers can purchase with confidence.
%K 情感分析，情感词典，词云，LDA主题模型，交互可视化<br>Sentiment Analysis
%K Sentiment Lexicon
%K Word Cloud
%K LDA Topic Model
%K Interactive Visualization
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=78496