摘要
针对传统旅游业对旅游景点评论的主题挖掘深度不足,舆情分析实时追踪能力较差的问题,本文提出了一种基于LDA模型和情感分析的舆情分析方法。以贵州省旅游景点评论为研究对象,通过Python采集和数据清洗后形成2010年至2020年共计33666条评论,利用领域词典和LDA模型实现细粒度情感分析和主题挖掘,再采用共词分析和社交网络方法挖掘积极评论和消极评论的关联关系。实验结果表明,通过融合领域词典的LDA模型能够成功识别出积极情感的著名景点、民族风俗、旅游体验三大类主题,消极情感的收费问题、交通环境、游客拥挤三大类主题,有效挖掘出贵州旅游景点主题的关联关系和情感诉求,为景区优质服务和决策提供数据支持和理论支撑。
In view of the problems of the traditional tourism industry's lack of depth in exploring the theme of tourist attractions'comments and poor real-time tracking ability of public opinion analysis,this paper proposes a public opinion analysis method based on the LDA model and sentiment analysis.Taking comments on tourist attractions in Guizhou Province as the research object,a total of 33666 comments from 2010 to 2020 are formed through Python collection and data cleaning.The domain dictionary and LDA model are used to achieve fine-grained sentiment analysis and topic mining,and then co-word analysis and social interaction are used.The network method explores the relationship between positive comments and negative comments.The experimental results show that the LDA model of the fusion domain dictionary successfully identified three categories of positive emotions:famous scenic spots,ethnic customs,and tourism experience,and three categories of negative emotions:charging issues,traffic environment,and tourist congestion,effectively mining Guizhou The relationship and emotional statement of the theme of tourist attractions provide data support and theoretical support for the scenic area to provide high-quality services and decision-making.
作者
杨秀璋
宋卓远
赵凯
陈镱尹
杨鑫
杨云帆
赵小明
周既松
罗子江
Yang Xiuzhang;Song Zhuoyuan;Zhao Kai;Chen Yiyin;Yang Xin;Yang Yunfan;Zhao Xiaoming;Zhou Jisong;Luo Zijiang(School of Information of Guizhou University of Finance and Economics,Guiyang 550025)
出处
《现代计算机》
2021年第25期36-43,共8页
Modern Computer
基金
贵州省科技计划项目(黔科合基础[2019]1041、黔科合基础[2020]1Y279、黔科合基础[2020]1Y021)
贵州省教育厅青年科技人才成长项目(黔教合KY字[2021]135、黔教合KY字[2018]166)
贵州财经大学2019年度校级项目(2019XQN01)
贵州省研究生教育创新计划项目(黔教合YJSCXJH[2019]066)。
关键词
贵州旅游
舆情分析
LDA模型
社交网络
主题挖掘
Guizhou tourism
public opinion analysis
LDA model
social network
theme mining