基于成人学习者在线文本数据的挖掘和情感分析

Online Text Data Mining and Emotion Analysis Based on Adult Learners

通过采集成人学习者在线聊天文本数据,利用python中的jieba分词对在线聊天记录文本进行中文分词,得到分词结果、生成词频表,实现高频词可视化;采用朴素贝叶斯分类器进行文本情感倾向分析,并生成正面、负面、中立三个情感文本文档;以LDA主题模型对情感文本进行主题聚类,得到的正面、负面、中立的主题情感词。情感词分析结果显示,聊天文本多为正常积极的情感交流。针对负面情感集中的问题原因,分别从平台操作、系统使用等方面提出了整改建议。

The paper collects online chat text data of adult learners,using Jiebaword segmentation in Python to segment Chinese words in online chat record text to get word segmentation results,generate word frequency table,and visualize high-frequency words.It uses Naive Bayes classifier for text sentiment tendency analysis,and generates positive,negative,and neutral emotional text documents.Then it uses the LDA topic model to cluster the emotional texts,and gets the positive,negative and neutral topic emotional words.The analysis of emotional words shows that the chat texts are mostly normal and positive emotional communication.For the causes of the problem of negative emotion concentration,it gives rectification suggestions on the platform operation and system use.