Designing Algorithmic Editors: How Newspapers Embed and Encode Journalistic Values into News Recommender Systems

Abstract

Recommender systems are increasingly used to prioritise and personalise content on news websites, and many news organisations develop their own systems to align them with internal logics, values, and goals. Based on two case studies of the development and implementation of recommender systems within the Scandinavian daily press, this study contributes with insights into how news organisations are tailoring this technology to journalism by exploring the role of journalistic values in the process. While mainly news values related to relevance and timeliness were encoded into the algorithmic systems, the news organisations embedded other values into practices surrounding the systems by monitoring their output and setting up editorial guardrails in the form of measures for manual control. These efforts illustrate that news organisations are thinking holistically about the role of recommender systems with an attention to normative values of public interest and a recognition of the limitations of the technology, translating into new practices that incorporate editorial control into algorithmic curation. The two studied approaches to algorithmic development further highlight that news organisations wanting to encode journalistic values into technical systems will benefit from proactively identifying and conceptualising the values early in the process for them to inform development and implementation.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 39 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.