文章摘要
黎巎,谢宗彦,张公鹏,郝志成,向征.基于LDA 的游客网络评论主题分类:以故宫为例[J].情报工程,2017,3(3):055-063
基于LDA 的游客网络评论主题分类:以故宫为例
Topic Classification of Tourist Online Reviews Based on LDA: The Case of the Forbidden City, Beijing
  
DOI:10.3772/j.issn.2095-915X.2017.03.008
中文关键词: LDA, 游客, 网络评论, 情感分析, 故宫
英文关键词: LDA, tourist, online review, sentimental analysis, the Forbidden City
基金项目:本文受国家自然科学基金青年项目“基于Agent的景区游客游憩行为仿真建模研究”(41101111),北京联合大学学术(科研)创新团队资助项目“旅游大数据研究方法、关键技术与应用研究”(Rk100201509)的资助。
作者单位
黎巎 北京联合大学旅游信息化协同创新中心 
谢宗彦 北京联合大学北京市信息服务工程重点实验室 
张公鹏 北京联合大学旅游信息化协同创新中心 
郝志成 北京联合大学旅游信息化协同创新中心 
向征 美国弗吉尼亚理工大学 
摘要点击次数: 2363
全文下载次数: 1897
中文摘要:
      游客的网络评论由于能够真实反映对旅游服务的真实体验及感受,正在逐渐影响旅游者对旅游目的地的看法甚至旅游消费行为。如何将碎片化的旅游评论转化为对其他用户和旅游经营者有价值的且直观的信息,成为旅游信息挖掘的热点。本文提出了基于LDA(Latent Dirichlet Allocation)主题发现模型的游客评论挖掘方法,以大众点评、携程及马蜂窝中关于故宫的用户在线评论为例,挖掘游客关于故宫的关注主题并分析其情感倾向。实验结果表明,故宫的游客网络评论主题包含入口服务、历史文化、体验感受以及遗址文物四个方面,游客对该四个主题的情感倾向均为正向;其中,大众点评和马蜂窝在体验感受方面的情感极性值较高。该方法对定位旅游目的地游客关注点具有实践意义。
英文摘要:
      Tourist online reviews can reflect their real experiences and feelings on travel service, which gradually influence other tourists’ opinions on tourist destinations and even consumption behaviors. How to convert the fragmented travel comment data into useful information for other users and tour operators becomes a hotspot of tourism information mining. This paper proposed an information mining method based on the LDA (Latent Dirichlet Allocation) topic discovery model, and carried on a topic classification and sentimental analysis of the online reviews of the Forbidden City in dianping.com, Ctrip.com and mafengwo. cn. The results showed that there were four topics of online reviews of the Forbidden City, these are entrance service, history and culture, experience and feeling, heritage and cultural relic, and the sentimental on these four topics is positive. Wherein, dianping.com and mafengwo.cn have high sentimental polarity values in terms of experience and feeling. This method has practice meaning for addressing what tourist pay attention to on destinations.
查看全文   查看/发表评论  下载PDF阅读器
关闭

分享按钮