图书情报工作 ›› 2023, Vol. 67 ›› Issue (2): 64-75.DOI: 10.13266/j.issn.0252-3116.2023.02.007

• 情报研究 • 上一篇    下一篇

过滤气泡情境下用户行为及其形成机制研究

张庭玮1, 朱沁雨1, 宋德凤1, 陈烨2   

  1. 1 华中师范大学信息管理学院 武汉 430079;
    2 南京大学信息管理学院 南京 210023
  • 收稿日期:2022-08-14 修回日期:2022-10-12 出版日期:2023-02-09 发布日期:2023-02-09
  • 通讯作者: 陈烨,长聘助理教授,博士,通信作者,E-mail:chenye@nju.edu.cn。
  • 作者简介:张庭玮,本科生;朱沁雨,硕士研究生;宋德凤,本科生。
  • 基金资助:
    本文系国家自然科学基金青年项目“基于多视角学习的社会化问答平台用户画像研究”(项目编号:71904057)研究成果之一。

Research on User Behaviors and their Forming Mechanism under Filter Bubbles

Zhang Tingwei1, Zhu Qinyu1, Song Defeng1, Chen Ye2   

  1. 1 School of Information Management, Central China Normal University, Wuhan 430079;
    2 School of Information Management, Nanjing University, Nanjing 210023
  • Received:2022-08-14 Revised:2022-10-12 Online:2023-02-09 Published:2023-02-09

摘要: [目的/意义] 推荐算法技术快速发展所产生的“过滤气泡”现象给用户信息行为带来深刻的影响。从用户角度出发,对其面对过滤气泡时的行为类型以及行为产生的机制进行探索性研究,帮助用户建立对过滤气泡现象的理性认知,为信息服务平台明确推荐算法的建设方向、改进服务水平提供一定的参考。[方法/过程] 采用扎根理论,选取对互联网信息服务平台有一定使用经验的30位用户进行半结构化访谈,并进行编码分析,构建用户面对过滤气泡时的行为与形成机制模型。[结果/结论] 用户面对过滤气泡时的行为类型主要包括忽略行为、缓解行为、加强行为、突破行为以及脱离行为。感知控制、态度和信息需求直接影响用户面对过滤气泡时的行为;推荐算法通过态度以及感知控制的中介作用对行为产生影响;此外,个人特质对行为产生的全过程起到调节作用。对特定情境下用户信息行为的研究以及进一步探究过滤气泡现象提供了一种研究视角和研究基础。

关键词: 过滤气泡, 用户行为, 扎根理论, 形成机制

Abstract: [Purpose/Significance] The "filter bubbles" phenomenon brought about by the rapid development of recommendation algorithm technology may have a profound impact on users' information behaviors. Based on the perspective of users, this study conducts an exploratory study on the types of behaviors and their forming mechanism under filter bubbles, helping users establish a rational cognition of the filter bubbles, providing a certain reference for the information service platform to clarify the construction direction of the recommendation algorithm and improve the service level.[Method/Process] The study selected 30 users who have some experience in using the Internet information service platform to conduct semi-structured interviews, employing grounded theory to proceed code analysis to construct a model of user behaviors and its forming mechanism in face of filter bubbles.[Result/Conclusion] The behavior types of users in face of filter bubbles mainly include ignoring behavior, mitigation behavior, reinforcement behavior, breakthrough behavior and disengagement behavior. Perceived control, attitude and information demand directly affect users' behavior when faced with filter bubbles; recommendation algorithm affect behaviors through the mediating effect of attitude and perceived control; in addition, personal character act as a regulator and affect the overall formation of users' behaviors. This study enriches research content of users' information behaviors under specific contexts and provide a research perspective and a basis for further research on filter bubbles.

Key words: filter bubbles, user behavior, grounded theory, forming mechanism

中图分类号: