摘要
[目的/意义]本研究以具有代表性的在校大学生为实验对象,试图探索用户在不同逻辑的个性化信息推荐系统中信息偶遇的差异。[方法/过程]本研究将信息偶遇和心理学领域的眼动实验相结合,以及定量定性分析相结合,验证了所提出的研究假设。[结论/发现]得到了以下个性化认识:类“知网节”用户维度的推荐系统在网站中信息偶遇效果更明显;用户的信息偶遇行为是一个复杂性系统的黑箱过程,可以通过调试网站的信息构建来提供更好的偶遇机会;揭示出规范算法推荐,避免过度与无序推荐而造成的“信息焦虑与信息茧房”双向叠加“过滤泡”效应的必要性。
[Objective/Meaning]The representative college students are taken as testees to explore the significant differences of information encounter among the users in different logics of personalized information recommendation systems.[Method/Process]Based on interdisciplinary research of encountering information and eye movement experiment,combined with quantitative and qualitative analysis,hypothesis was verified.[Conclusion/Discovery]The recommendation system derived from user behaviors like“CNKI Nodes”gets more web IE results;Although IE behaviour of users is a black box process of complex system,reasonable information architecture can increase the effectiveness of it;Is is essential to standard the algorithm recommendation to avert“the filter bubble”effect of superimposed“information anxiety and information cocoons”result in excessive disorder recommendations.
作者
刘春茂
张学佳
周悦
Liu Chunmao;Zhang Xuejia;Zhou Yue(School of Management,Tianjin Normal University,Tianjin 300387,China)
出处
《现代情报》
CSSCI
2022年第5期26-37,共12页
Journal of Modern Information
基金
2018年国家社会科学基金项目“面向语义信息关联的网站信息偶遇机理的实证研究”(项目编号:18BTQ068)。
关键词
信息偶遇
信息推荐
眼动实验
信息构建
信息茧房
information encountering
information recommendation
eye movement experiment
information architecture
information cocoons