图书情报工作 ›› 2021, Vol. 65 ›› Issue (2): 107-116.DOI: 10.13266/j.issn.0252-3116.2021.02.011

• 情报研究 • 上一篇    下一篇

共享住宿与酒店用户评论文本的跨平台比较研究:基于LDA的主题社会网络和情感分析

池毛毛1,2, 潘美钰1, 王伟军3   

  1. 1. 华中师范大学信息管理学院 武汉 430079;
    2. 华中师范大学湖北省电子商务研究中心 武汉 430079;
    3. 华中师范大学青少年网络心理与行为教育部重点实验室 武汉 430079
  • 收稿日期:2020-04-13 修回日期:2020-07-20 出版日期:2021-01-20 发布日期:2021-01-20
  • 作者简介:池毛毛(ORCID:0000-0003-2726-5933),副教授,博士,硕士生导师,E-mail:chimaomao@aliyun.com;潘美钰(ORCID:0000-0001-7591-3726),本科生;王伟军(ORCID:0000-0003-4948-0634),教授,博士生导师。
  • 基金资助:
    本文系国家自然科学基金青年项目“电商平台演化对平台绩效的影响机理研究:基于复杂适应系统的视角”(项目编号:71801104)研究成果之一。

A Cross-platform Comparative Study of Reviews on Sharing Accommodation and Hotels Reservation Platform:Combined with LDA-SNA and Sentiment Analysis

Chi Maomao1,2, Pan Meiyu1, Wang Weijun3   

  1. 1. School of Information Management, Central China Normal University, Wuhan 430079;
    2. E-commerce Research Center of Hubei Province, Central China Normal University, Wuhan 430079;
    3. Key Laboratory of Adolescent Cyberpsychology and Behavior, Central China Normal University, Wuhan 430079
  • Received:2020-04-13 Revised:2020-07-20 Online:2021-01-20 Published:2021-01-20
  • Supported by:
     

摘要: [目的/意义] 共享住宿与酒店预定平台可能同时存在替代性和互补性,但这种替代性和互补性分别体现在哪些产品和服务上当前文献还缺乏探讨,需要进一步开展跨平台的比较研究。[方法/过程] 选取携程酒店预定平台和小猪短租平台为实验对象,采集北京市相关房源的86 635条用户评论文本,结合LDA模型、主题社会网络和主题情感分析方法对用户文本评论进行跨平台比较分析。[结果/结论] 研究发现两大平台用户在评论主题、主题社会网络和主题情感上的异同之处,从微观用户评论角度解释了两大平台在产品和服务上的替代性和互补性。本文结果为平台管理者进行住宿产品和服务的开发和改进提供重要的实践借鉴。

 

关键词: 跨平台比较, 文本主题挖掘, 社会网络分析, 情感分析

Abstract: [Purpose/significance] There may be substitutability and complementarity between shared accommodation and hotel reservation platforms. However, there is still a lack of discussion on which products and services this kind of substitutability and complementarity are embodied in. Therefore, a cross-platform comparative analysis is needed.[Method/process] In this paper, 86 635 reviews of relevant rental rooms in Beijing were collected from Ctrip.com and Xiaozhu.com. The cross-platform comparative analysis of online reviews is further carried out by integrated latent dirichlet allocation (LDA), social network analysis (SNA) and sentiment analysis.[Result/conclusion] The study found the similarities and differences of the two platforms' users in the comment theme, social network and the emotion of each topic, and further explained the substitutability and complementarity of the two platforms in products and services from the perspective of users' online reviews. The results of this paper also provide important practical reference for platform managers to develop and improve accommodation products and services.

Key words: cross-platform comparison, text topic mining, social network analysis, sentiment analysis

中图分类号: