Abstract

Info

Title:: Embedding Algorithmic Values: A Study on Value Sensitive Design of Algorithmic Systems for News Production

Keywords:: empirical turn; critical algorithm studies; value sensitive design; algorithm audit

Abstract:: The “algorithmic turn” has confronted journalism with a real problem: how to embed journalistic values into algorithms and match the values of algorithmic systems with those of the media. In this paper, we attempt to explore how algorithms have values from the “empirical turn” in the philosophy of technology and critical algorithm research through combining value-sensitive design with the algorithm design process consisting of “context analysis”, “value identification”, “technical translation”, and “prototype design”. On this basis, we propose a specific mechanism for embedding news values into algorithms, reflects on value-sensitive design, and points out that value sensitive design embeds limited values. We argue that the key to successful value sensitive design lies in the collaborative participation of stakeholders, and the value sensitive design should follow the ethical principles of transparency, explainability and auditability.