算法驯化:个性化推荐平台的自媒体内容生产网络及其运作

Algorithmic Domestication: the Content Production Network of Entrepreneurial Media in Personalised Recommendation Platforms

本文整合行动者网络理论的"转译"概念和"驯化"理论,研究面向算法推荐平台的自媒体生产实践。笔者观察了4个推荐平台,访谈了6位平台管理者和6位自媒体创作者,基于实证数据阐释发生在自媒体生产过程中的两次算法驯化。两次驯化构建了两个行动者网络--在平台运营实践中形成的"职责共识网络"、在自媒体创作群体中形成的"规律共识网络",二者再共同构成自媒体内容生产网络。算法规则在两个局部网络中都是"强制通过点",同时也是局部网络之间的连接点,由此成为整个网络的权力中心。

Based on integrating the concept of translation from the framework of Actornetwork Theory and the Domestication Theory, this study investigated the content production practice of entrepreneurial media that distribute content outputs in algorithmic recommendation platform. Authors had observed 4 different platforms, interviewed 6 platform managers and 6 content producers. Drawing on the empirical data, the authors interpreted two-time of algorithmic domestication occurring inside and outside of the platform organizations. Twice processes of domestication construct two actor networks. One is the actor network of responsibility consensus, which is inside of the platform practices, is led by content managers and followed by technological and commercial managers. Another is the actor network of pattern consensus, which is inside of the community of entrepreneurial media, is led by head producers and followed by middle of tail producers. These two networks constitute the content production network of entrepreneurial media. Algorithmic rules play the role of obligatory passage point in each actor network and the role of connection between them, so then the power center of the content production network.