基于文本挖掘的鸿蒙系统用户评论情感分析与研究
2021-09-13陈诺祎单菁王佳英石璐璐
陈诺祎 单菁 王佳英 石璐璐



摘 要:随着科技的进步,由华为自主研发的首个国产手机操作系统成功问世,引起社会强烈反响。该文以网络爬虫抓取的用户评论为例,基于评论数据建立评估模型,主要使用文本描述性分析、SnowNlp情感分析和LDA潜在主题挖掘这3种方法,从不同角度对用户评论进行文本挖掘分析,找出评论大数据背后隐含的情感倾向、用户观点等深层信息,有助于企业了解用户的使用体验与口碑动态。该文分析模型的评估效果可信度较高,所采用的研究方法也适用于电商评论分析、社会舆情分析等方面。
关键词:网络爬虫 情感分析 LDA 文本挖掘
中圖分类号:TP391.1 文献标识码:A文章编号:1672-3791(2021)06(a)-0026-04
Sentiment Analysis and Research of User Comments on Hongmeng System Based on Text Mining
CHEN Nuoyi SHAN Jing* WANG Jiaying SHI Lulu
(School of Information and Control Engineering, Shenyang Jianzhu University, Shenyang, Liaoning Province, 110168 China)
Abstract: With the progress of science and technology, the first domestic mobile phone operating system developed by Huawei has been successfully launched, which has aroused strong social response. Based on web crawler fetching user comments as an example, the evaluation model is established based on the review data. Three methods are mainly used: text descriptive analysis, SnowNlp sentiment analysis and LDA potential topic mining method, conducting text mining analysis on user comments from different perspectives to find out the emotional tendency, user opinions and other deep information hidden behind thebig data of comments. It is helpful for enterprises to understand the user experience and word of mouth dynamic. The evaluation effect of the analysis model in this paper has a high credibility, and the research method adopted is also applicable to e-commerce comment analysis, social public opinion analysis and other aspects.
Key Words: Web crawler; Sentiment analysis; LDA; Text mining
鸿蒙OS是一款“面向未来”的操作系统,一款基于微内核的面向全场景的分布式操作系统,是由华为开发人员打造的国产手机系统。对于鸿蒙系统的正式发布,国内用户纷纷通过网络平台发布对该系统的评论看法,用户的评论信息包括了客观评论与主观评论,而主观评论占比极高,该文通过采集B站(某自媒体视频网站)关于鸿蒙系统视频底部的用户评论信息,建立相关的文本挖掘模型对评论信息进行深度挖掘。
1 研究方法
该文对选取的关于鸿蒙系统在线评论利用八爪鱼采集工具,设定采集规则。主要抓取的内容为用户评论信息。对于抓取的数据进行数据清洗,包括文本去重、去空值、剔除广告信息与无意义评论、中文文本分词等操作。通过……
