Black-box Audit of YouTube's Video Recommendation: Investigation of Misinformation Filter Bubble Dynamics

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F22%3APU146535" target="_blank" >RIV/00216305:26230/22:PU146535 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.ijcai.org/proceedings/2022/749" target="_blank" >https://www.ijcai.org/proceedings/2022/749</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.24963/ijcai.2022/749" target="_blank" >10.24963/ijcai.2022/749</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Black-box Audit of YouTube's Video Recommendation: Investigation of Misinformation Filter Bubble Dynamics

Popis výsledku v původním jazyce

We investigated the creation and bursting dynamics of misinformation filter bubbles on YouTube using a black-box sockpuppeting audit technique. In this study, pre-programmed agents acting as YouTube users stimulated YouTube's recommender systems: they first watched a series of misinformation promoting videos (bubble creation) and then a series of misinformation debunking videos (bubble bursting). Meanwhile, agents recorded videos recommended to them by YouTube. After manually annotating these recommendations, we were able to quantify the portion of misinformative videos among them. The results confirm the creation of filter bubbles (albeit not in all situations) and show that these bubbles can be bursted by watching credible content. Drawing a direct comparison with a previous study, we do not see improvements in overall quantities of misinformation recommended.

Název v anglickém jazyce

Black-box Audit of YouTube's Video Recommendation: Investigation of Misinformation Filter Bubble Dynamics

Popis výsledku anglicky

We investigated the creation and bursting dynamics of misinformation filter bubbles on YouTube using a black-box sockpuppeting audit technique. In this study, pre-programmed agents acting as YouTube users stimulated YouTube's recommender systems: they first watched a series of misinformation promoting videos (bubble creation) and then a series of misinformation debunking videos (bubble bursting). Meanwhile, agents recorded videos recommended to them by YouTube. After manually annotating these recommendations, we were able to quantify the portion of misinformative videos among them. The results confirm the creation of filter bubbles (albeit not in all situations) and show that these bubbles can be bursted by watching credible content. Drawing a direct comparison with a previous study, we do not see improvements in overall quantities of misinformation recommended.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2022

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence Sister Conferences Best Papers

ISBN

9781956792003

ISSN

1045-0823

e-ISSN

—

Počet stran výsledku

5

Strana od-do

5349-5353

Název nakladatele

International Joint Conferences on Artificial Intelligence

Místo vydání

Vienna

Místo konání akce

Vienna

Datum konání akce

23. 7. 2022

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—