摘要
[目的/意义]对数智时代下跨平台用户信息过滤气泡强度进行比较研究,探讨同类内容智能分发平台的差异性和互补性,有助于平台和用户采取有效措施破除过滤气泡,提升信息多样性。[方法/过程]结合LDA主题模型和信息熵理论,以30746条用户文本评论为样本,测度今日头条和腾讯新闻平台用户信息过滤气泡强度,比较分析跨平台之间的异同之处。[结果/结论]研究发现两大平台不会导致用户陷入过滤气泡,而是提高了用户信息多样性,且今日头条的提升幅度大于腾讯新闻。此外,两大平台的信息存在互补性,用户复合使用有助于提升信息多样性。
[Purpose/significance]A comparative study on the strength of cross-platform user information filter bub?bles in the era of digital intelligence,to explore the differences and complementarities of similar content intelligent dis?tribution platforms,will help platforms and users to take effective measures to break the filter bubbles and improve in?formation diversity.[Method/process]Combined with LDA topic model and information entropy theory,taking 30746 user text comments as a sample,measuring the filter bubble strength of user information on Toutiao and Tencent News,and comparing and analyzing the similarities and differences between cross-platforms.[Result/conclusion]The study found that the two platforms did not cause users to fall into filter bubbles,but increased the diversity of user informa?tion,and the improvement of Toutiao was greater than that of Tencent News.In addition,the information of the two plat?forms is complementary,and the combined use of two platforms helps to improve the diversity of user information.
作者
王益成
张梅
李会
Wang Yicheng;Zhang Mei;Li Hui(School of Management Science and Engineering at Anhui University of Finance and Economics,Bengbu,233030;China Institute of Science and Technology Information,Beijing,100038)
出处
《情报资料工作》
北大核心
2023年第3期88-97,共10页
Information and Documentation Services
关键词
过滤气泡
跨平台比较
推荐算法
LDA模型
信息熵
filter bubbles
cross-platform comparison
recommendation algorithm
LDA mode
information entropy