January  2018 43
專輯論文Special Issue Articles
雙強寡頭平台新聞推薦算法機制研究
Critical Algorithm Study on News Recommendation of Duopoly Platform
  (852)
作者 方師師
Author Shishi FANG
關鍵詞 平台媒體、算法機制、批判算法研究
Keywords platisher, algorithm mechanism, critical algorithm study
摘要 算法是平台媒體新聞推薦的核心。本文採用批判算法研究的視角對「雙強寡頭」平台——臉書和谷歌的「算法機制」進行了研究。「算法機制」包含核心排序算法、新聞要素權重和算法價值觀三個維度,關注的是決定算法背後的關鍵技術、社會規範和價值觀念。文章發現,兩大平台的「算法機制」呈現兩分,谷歌注重客觀要素並試圖建立技術性的網絡秩序,而臉書則更多強調以社交為基礎的人與人之間的關係構成。而蘊藏在不同「算法機制」內部的系統性與結構性偏向,最終指向了對於「算法機制」可信任框架的探尋。
Abstract Algorithm occupies the core of the platisher’s news recommendation. This paper uses the perspective of critical algorithm study (CAS) to explore the algorithmic mechanism bedding in the news recommendations of the Duopoly platform, Facebook and Google. It finds out that the algorithmic mechanism of Facebook and Google in the news recommendation business presents their own key technologies, social norms and values. The systematic and structural bias in the algorithm mechanism finally points to the trustworthy framework of the algorithm mechanism.



本文引用格式:

方師師(2018)。〈雙強寡頭平台新聞推薦算法機制研究〉。《傳播與社會學刊》,第43期,頁103–122。



Citation of this article:

Fang, S. (2018). Critical algorithm study on news recommendation of Duopoly platform. Communication & Society, 43, 103–122.
No.45  2018 July
No.44  2018 April
No.43  2018 January
No.42  2017 October
No.41  2017 July
© 2017 Communication & Society Journal. All Rights Reserved.