推送算法平台的内容类“热趋同”效应与路径——基于“今日头条”平台的路径分析  

The Effect and Path of Content“Hot Convergence”in Recommendation Algorithm Platform:Path Analysis Based on“Toutiao”Platform

作　　者：徐翔 孙天怡[3] Xu Xiang;Sun Tianyi

机构地区：[1]同济大学艺术与传媒学院 [2]同济大学大数据与计算传播研究中心 [3]南京理工大学紫金学院人文与社会科学学院

出　　处：《媒介批评》2024年第2期144-163,共20页Media Criticism

基　　金：上海市软科学研究项目“从各有所好到共同牢笼:推荐算法平台公共信息茧房形成机理与调控策略”(项目号:23692110600);中央高校基本科研业务费专项资金资助项目“振荡的琴弦:网络舆论主题的振荡周期特征对主题间触发的影响机理与调节策略研究”(项目号:22120240589)的阶段性成果。

摘　　要：推送算法平台隐藏在个性化推荐机制下的是热门内容的同质化和封闭化,这种宏观性的“热趋同”效应将带来平台整体性的内容窄化,影响受众品味和平台中的共识生成演化。本研究结合结构方程模型,明确提出算法平台中各内容主题“热趋同”的生成路径与传导结构。在算法推荐平台中,内容类的热度与其在类相似度矩阵中的中心度之间存在正比关系,并由形成本研究的三个支撑假设及其作用结构。推送算法在打造以平台为边界的内容封闭结构时,有必要采取措施促使算法工具成为受众认知世界的良性渠道,对平台的信息趋同与封闭性演化进行有效调控。What is hidden in the personalized recommendation mechanism of recommendation algorithm platform is the homogenization and closure of popular content.This macroscopic“hot convergence”effect will bring about the overall content narrowing of the platform,and affect the audience s taste and the generation and evolution of consensus in the platform.Based on the structural equation model,the generation path and conduction structure of“hot convergence”in the algorithm platform are proposed.In the algorithm recommendation platform,there is a proportional relationship between the popularity of the content class and its centrality in the class similarity matrix,and the consequences represented by the core hypothesis are formed by three supporting hypotheses and their internal structures.The push algorithm is creating a platform-bounded content closed structure,and it is necessary to take measures to make the algorithm tool become a benign channel for the audience to know the world,and effectively regulate the information convergence and closed evolution of the platform.

关 键 词：推荐算法平台 内容趋同 热趋同 平台内容封闭性 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]