信息平台的内容封闭性传导效应及机制研究——基于今日头条的样本分析

Content Closure Conduction Effects and Mechanisms on Information Platform:Based on Toutiao

信息平台中的内容封闭性问题日渐受到学界关注,但关于内容封闭性是否存在,其形成原因和运作机理等的相关辩争仍然未有定论。基于今日头条平台,通过引入内容封闭性传导概念,重点探究内容封闭性变化中的传导效应及其机制,回应信息平台中内容是否封闭以及如何封闭等问题。通过大数据抓取和处理,对今日头条平台的数据进行聚类处理和内容基尼系数的计算,并将所有内容按照热度划分为46个层级。在效应检验方面,运用格兰杰因果分析方法得到各个热度层级之间传导封闭性的关系。在机制分析方面,结合二元Logistic回归分析,发现热度越相近的内容之间越有可能产生封闭性的传导作用,即内容封闭性的传导满足热度接近性规律。此外,运用社会网络(QAP)分析方法识别出可能与内容封闭性传导效应相关的25类内容,其主题以分享生活和文娱类型为主。研究结论不仅丰富了信息平台中内容封闭性的表现及机理方面的研究,对网络空间中的信息窄化和舆论调控等议题亦有现实参考价值。

The issue of content closure on information platforms has received increasing attention in academia.However,the debates on whether content closure exists,the formation causes,and its operation mechanisms remain inconclusive.Based on the Toutiao platform,this paper addresses the question of whether and how content closure occurs on information platforms by introducing the concept of content closure conduction,and focusing on the effect and mechanism of this phenomenon.Through big data fetching and processing,more than 18 million data points from the Toutiao platform were clustered and content Gini coefficient was calculated.All content was divided into 46 tiers according to its heat level,in order to analyze the conduction efficiency between different levels of heat in a more fine-grained manner.This paper empirically examines the core hypotheses moving from effect test to mechanism analysis.In the effect test,the granger causality analysis method was applied to determine the conduction between each heat level.Besides,in the mechanism analysis,the binary logistic regression analysis was employed to verify the influence of the heat gap between the tiers on the effect of conduction effect.Subsequently,the QAP regression analysis method was used to identify 25 content categories that might be related to the conduction effect of content closure.The results indicate that the phenomenon of content closure conduction exists between contents with different heat levels,and the more similar the heat levels,the more likely they are to produce closure conduction between contents.In other words,the content closure conduction follows the“heat proximity”law.In terms of content features,only a small number of content categories,which encompass sharing life and entertainment themes,are likely to have an impact on the dissemination effect.This paper reveals the formation and evolution of content closure on the Toutiao platform through the perspective of the conduction effect.Besides,the results confirm the influence of the content heat gap and topic features on the closure conduction effect,enhancing the understanding of mechanisms,characteristics and mechanism of content closure on the information platforms,and offering practical references for addressing issues,such as content homogenization,information narrowing,and public opinion regulation in cyberspace.