No Cover Image

Journal article 947 views 127 downloads

Recommender systems and the amplification of extremist content

Joe Whittaker Orcid Logo, Sean Looney, Alastair Reed Orcid Logo, Fabio Votta

Internet Policy Review, Volume: 10, Issue: 2

Swansea University Authors: Joe Whittaker Orcid Logo, Sean Looney, Alastair Reed Orcid Logo

  • policyreview-2021-2-1565.pdf

    PDF | Version of Record

    Released under the terms of a Creative Commons Attribution 3.0 Germany

    Download (483.51KB)

Check full text

DOI (Published version): 10.14763/2021.2.1565

Abstract

Policymakers have recently expressed concerns over the role of recommendation algorithms and their role in forming “filter bubbles.” This is a particularly prescient concern in the context of extremist content online; these algorithms may promote extremist content at the expense of more moderate voi...

Full description

Published in: Internet Policy Review
ISSN: 2197-6775
Published: Internet Policy Review, Alexander von Humboldt Institute for Internet and Society 2021
Online Access: Check full text

URI: https://cronfa.swan.ac.uk/Record/cronfa57054
Tags: Add Tag
No Tags, Be the first to tag this record!
first_indexed 2021-07-05T09:03:08Z
last_indexed 2023-01-11T14:36:42Z
id cronfa57054
recordtype SURis
fullrecord <?xml version="1.0"?><rfc1807><datestamp>2022-11-02T13:38:14.7793384</datestamp><bib-version>v2</bib-version><id>57054</id><entry>2021-06-08</entry><title>Recommender systems and the amplification of extremist content</title><swanseaauthors><author><sid>112ed59957393e783f913443ec80faab</sid><ORCID>0000-0001-7342-6369</ORCID><firstname>Joe</firstname><surname>Whittaker</surname><name>Joe Whittaker</name><active>true</active><ethesisStudent>false</ethesisStudent></author><author><sid>98bae6ad8004bb8fa78d382f2630dbe3</sid><firstname>Sean</firstname><surname>Looney</surname><name>Sean Looney</name><active>true</active><ethesisStudent>false</ethesisStudent></author><author><sid>115297b63e005e2b75991efe269cd4a2</sid><ORCID>0000-0002-9060-5518</ORCID><firstname>Alastair</firstname><surname>Reed</surname><name>Alastair Reed</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2021-06-08</date><deptcode>CSSP</deptcode><abstract>Policymakers have recently expressed concerns over the role of recommendation algorithms and their role in forming &#x201C;filter bubbles.&#x201D; This is a particularly prescient concern in the context of extremist content online; these algorithms may promote extremist content at the expense of more moderate voices. In this article, we make two contributions to this debate. Firstly, we provide a novel empirical analysis of three platforms&#x2019; recommendation systems when interacting with far-right content. We find that one platform &#x2013; YouTube &#x2013; does amplify extreme and fringe content, while two &#x2013; Reddit and Gab &#x2013; do not. Secondly, we contextualise these findings into the regulatory debate. There are currently few policy instruments for dealing with algorithmic amplification, and those that do exist largely focus on transparency. We argue that policymakers have yet to fully understand the problems inherent in &#x201C;de-amplifying&#x201D; legal, borderline content and argue that a co-regulatory approach may offer a route towards tackling many of these challenges.</abstract><type>Journal Article</type><journal>Internet Policy Review</journal><volume>10</volume><journalNumber>2</journalNumber><paginationStart/><paginationEnd/><publisher>Internet Policy Review, Alexander von Humboldt Institute for Internet and Society</publisher><placeOfPublication/><isbnPrint/><isbnElectronic/><issnPrint>2197-6775</issnPrint><issnElectronic/><keywords>Filter bubble, Online radicalisation, Algorithms, Extremism, Regulation</keywords><publishedDay>30</publishedDay><publishedMonth>6</publishedMonth><publishedYear>2021</publishedYear><publishedDate>2021-06-30</publishedDate><doi>10.14763/2021.2.1565</doi><url/><notes/><college>COLLEGE NANME</college><department>Criminology, Sociology and Social Policy</department><CollegeCode>COLLEGE CODE</CollegeCode><DepartmentCode>CSSP</DepartmentCode><institution>Swansea University</institution><apcterm/><funders>Global Research Network on Terrorism &amp; Technology</funders><projectreference/><lastEdited>2022-11-02T13:38:14.7793384</lastEdited><Created>2021-06-08T10:01:54.8473674</Created><path><level id="1">Faculty of Humanities and Social Sciences</level><level id="2">School of Social Sciences - Criminology, Sociology and Social Policy</level></path><authors><author><firstname>Joe</firstname><surname>Whittaker</surname><orcid>0000-0001-7342-6369</orcid><order>1</order></author><author><firstname>Sean</firstname><surname>Looney</surname><order>2</order></author><author><firstname>Alastair</firstname><surname>Reed</surname><orcid>0000-0002-9060-5518</orcid><order>3</order></author><author><firstname>Fabio</firstname><surname>Votta</surname><order>4</order></author></authors><documents><document><filename>57054__20330__2741bca41e484a13adfb2a5cec60f23f.pdf</filename><originalFilename>policyreview-2021-2-1565.pdf</originalFilename><uploaded>2021-07-05T09:48:24.3804485</uploaded><type>Output</type><contentLength>495114</contentLength><contentType>application/pdf</contentType><version>Version of Record</version><cronfaStatus>true</cronfaStatus><documentNotes>Released under the terms of a Creative Commons Attribution 3.0 Germany</documentNotes><copyrightCorrect>true</copyrightCorrect><language>eng</language><licence>https://creativecommons.org/licenses/by/3.0/de/deed.en</licence></document></documents><OutputDurs/></rfc1807>
spelling 2022-11-02T13:38:14.7793384 v2 57054 2021-06-08 Recommender systems and the amplification of extremist content 112ed59957393e783f913443ec80faab 0000-0001-7342-6369 Joe Whittaker Joe Whittaker true false 98bae6ad8004bb8fa78d382f2630dbe3 Sean Looney Sean Looney true false 115297b63e005e2b75991efe269cd4a2 0000-0002-9060-5518 Alastair Reed Alastair Reed true false 2021-06-08 CSSP Policymakers have recently expressed concerns over the role of recommendation algorithms and their role in forming “filter bubbles.” This is a particularly prescient concern in the context of extremist content online; these algorithms may promote extremist content at the expense of more moderate voices. In this article, we make two contributions to this debate. Firstly, we provide a novel empirical analysis of three platforms’ recommendation systems when interacting with far-right content. We find that one platform – YouTube – does amplify extreme and fringe content, while two – Reddit and Gab – do not. Secondly, we contextualise these findings into the regulatory debate. There are currently few policy instruments for dealing with algorithmic amplification, and those that do exist largely focus on transparency. We argue that policymakers have yet to fully understand the problems inherent in “de-amplifying” legal, borderline content and argue that a co-regulatory approach may offer a route towards tackling many of these challenges. Journal Article Internet Policy Review 10 2 Internet Policy Review, Alexander von Humboldt Institute for Internet and Society 2197-6775 Filter bubble, Online radicalisation, Algorithms, Extremism, Regulation 30 6 2021 2021-06-30 10.14763/2021.2.1565 COLLEGE NANME Criminology, Sociology and Social Policy COLLEGE CODE CSSP Swansea University Global Research Network on Terrorism & Technology 2022-11-02T13:38:14.7793384 2021-06-08T10:01:54.8473674 Faculty of Humanities and Social Sciences School of Social Sciences - Criminology, Sociology and Social Policy Joe Whittaker 0000-0001-7342-6369 1 Sean Looney 2 Alastair Reed 0000-0002-9060-5518 3 Fabio Votta 4 57054__20330__2741bca41e484a13adfb2a5cec60f23f.pdf policyreview-2021-2-1565.pdf 2021-07-05T09:48:24.3804485 Output 495114 application/pdf Version of Record true Released under the terms of a Creative Commons Attribution 3.0 Germany true eng https://creativecommons.org/licenses/by/3.0/de/deed.en
title Recommender systems and the amplification of extremist content
spellingShingle Recommender systems and the amplification of extremist content
Joe Whittaker
Sean Looney
Alastair Reed
title_short Recommender systems and the amplification of extremist content
title_full Recommender systems and the amplification of extremist content
title_fullStr Recommender systems and the amplification of extremist content
title_full_unstemmed Recommender systems and the amplification of extremist content
title_sort Recommender systems and the amplification of extremist content
author_id_str_mv 112ed59957393e783f913443ec80faab
98bae6ad8004bb8fa78d382f2630dbe3
115297b63e005e2b75991efe269cd4a2
author_id_fullname_str_mv 112ed59957393e783f913443ec80faab_***_Joe Whittaker
98bae6ad8004bb8fa78d382f2630dbe3_***_Sean Looney
115297b63e005e2b75991efe269cd4a2_***_Alastair Reed
author Joe Whittaker
Sean Looney
Alastair Reed
author2 Joe Whittaker
Sean Looney
Alastair Reed
Fabio Votta
format Journal article
container_title Internet Policy Review
container_volume 10
container_issue 2
publishDate 2021
institution Swansea University
issn 2197-6775
doi_str_mv 10.14763/2021.2.1565
publisher Internet Policy Review, Alexander von Humboldt Institute for Internet and Society
college_str Faculty of Humanities and Social Sciences
hierarchytype
hierarchy_top_id facultyofhumanitiesandsocialsciences
hierarchy_top_title Faculty of Humanities and Social Sciences
hierarchy_parent_id facultyofhumanitiesandsocialsciences
hierarchy_parent_title Faculty of Humanities and Social Sciences
department_str School of Social Sciences - Criminology, Sociology and Social Policy{{{_:::_}}}Faculty of Humanities and Social Sciences{{{_:::_}}}School of Social Sciences - Criminology, Sociology and Social Policy
document_store_str 1
active_str 0
description Policymakers have recently expressed concerns over the role of recommendation algorithms and their role in forming “filter bubbles.” This is a particularly prescient concern in the context of extremist content online; these algorithms may promote extremist content at the expense of more moderate voices. In this article, we make two contributions to this debate. Firstly, we provide a novel empirical analysis of three platforms’ recommendation systems when interacting with far-right content. We find that one platform – YouTube – does amplify extreme and fringe content, while two – Reddit and Gab – do not. Secondly, we contextualise these findings into the regulatory debate. There are currently few policy instruments for dealing with algorithmic amplification, and those that do exist largely focus on transparency. We argue that policymakers have yet to fully understand the problems inherent in “de-amplifying” legal, borderline content and argue that a co-regulatory approach may offer a route towards tackling many of these challenges.
published_date 2021-06-30T04:12:30Z
_version_ 1763753853003497472
score 11.013082

Similar Items