An Algorithmic Diversity Diet? Questioning Assumptions behind a Diversity Recommendation System for PSM

Jannick Kirk Sørensen, Jan-Hinrik Schmidt

Research output: Working paperResearchpeer-review

Abstract

With the growing influence of personalized algorithmic recommender systems on the exposure of media content to users, the relevance of discussing the diversity of recommendations increases, particularly as far as public service media (PSM) is concerned. An imagined implementation of a diversity diet system however triggers not only the classic discussion of the reach – distinctiveness balance for PSM, but also shows that ‘diversity’ is understood very differently in algorithmic recommender system communities than it is editorially and politically in the context of PSM. The design of a diversity diet system generates questions not just about editorial power, personal freedom and techno-paternalism, but also about the embedded politics of recommender systems as well as the human skills affiliated with PSM editorial work and the nature of PSM content.
Original languageEnglish
Publication statusPublished - 23 Sep 2016

Fingerprint

public service
paternalism
politics
community

Keywords

  • Recommender Systems
  • Exposure Diversity
  • public service media
  • Paternalism
  • media content recommendation
  • reach - distinctiveness
  • algorithmic
  • diversity
  • media exposure

Cite this

@techreport{a5900c5f0dc14cc08727aad0933b0472,
title = "An Algorithmic Diversity Diet?: Questioning Assumptions behind a Diversity Recommendation System for PSM",
abstract = "With the growing influence of personalized algorithmic recommender systems on the exposure of media content to users, the relevance of discussing the diversity of recommendations increases, particularly as far as public service media (PSM) is concerned. An imagined implementation of a diversity diet system however triggers not only the classic discussion of the reach – distinctiveness balance for PSM, but also shows that ‘diversity’ is understood very differently in algorithmic recommender system communities than it is editorially and politically in the context of PSM. The design of a diversity diet system generates questions not just about editorial power, personal freedom and techno-paternalism, but also about the embedded politics of recommender systems as well as the human skills affiliated with PSM editorial work and the nature of PSM content.",
keywords = "Recommender Systems, Exposure Diversity, public service media, Paternalism, media content recommendation, reach - distinctiveness, algorithmic, diversity, media exposure",
author = "S{\o}rensen, {Jannick Kirk} and Jan-Hinrik Schmidt",
year = "2016",
month = "9",
day = "23",
language = "English",
type = "WorkingPaper",

}

An Algorithmic Diversity Diet? Questioning Assumptions behind a Diversity Recommendation System for PSM. / Sørensen, Jannick Kirk; Schmidt, Jan-Hinrik.

2016.

Research output: Working paperResearchpeer-review

TY - UNPB

T1 - An Algorithmic Diversity Diet?

T2 - Questioning Assumptions behind a Diversity Recommendation System for PSM

AU - Sørensen, Jannick Kirk

AU - Schmidt, Jan-Hinrik

PY - 2016/9/23

Y1 - 2016/9/23

N2 - With the growing influence of personalized algorithmic recommender systems on the exposure of media content to users, the relevance of discussing the diversity of recommendations increases, particularly as far as public service media (PSM) is concerned. An imagined implementation of a diversity diet system however triggers not only the classic discussion of the reach – distinctiveness balance for PSM, but also shows that ‘diversity’ is understood very differently in algorithmic recommender system communities than it is editorially and politically in the context of PSM. The design of a diversity diet system generates questions not just about editorial power, personal freedom and techno-paternalism, but also about the embedded politics of recommender systems as well as the human skills affiliated with PSM editorial work and the nature of PSM content.

AB - With the growing influence of personalized algorithmic recommender systems on the exposure of media content to users, the relevance of discussing the diversity of recommendations increases, particularly as far as public service media (PSM) is concerned. An imagined implementation of a diversity diet system however triggers not only the classic discussion of the reach – distinctiveness balance for PSM, but also shows that ‘diversity’ is understood very differently in algorithmic recommender system communities than it is editorially and politically in the context of PSM. The design of a diversity diet system generates questions not just about editorial power, personal freedom and techno-paternalism, but also about the embedded politics of recommender systems as well as the human skills affiliated with PSM editorial work and the nature of PSM content.

KW - Recommender Systems

KW - Exposure Diversity

KW - public service media

KW - Paternalism

KW - media content recommendation

KW - reach - distinctiveness

KW - algorithmic

KW - diversity

KW - media exposure

M3 - Working paper

BT - An Algorithmic Diversity Diet?

ER -