Open AccessWHAT’S ‘UP NEXT’? INVESTIGATING ALGORITHMIC RECOMMENDATIONS ON YOUTUBE ACROSS ISSUES AND OVER TIME
Author(s) -
Ariadna Matamoros-Fernández,
Joanne Gray,
Louisa Bartolo,
Jean Burgess,
Nicolas Suzor
Publication year - 2021
Publication title -
selected papers of internet research
Language(s) - English
Resource type - Journals
ISSN - 2162-3317
DOI - 10.5210/spir.v2021i0.12208
Subject(s) - metadata , computer science , feature (linguistics) , visibility , social media , information retrieval , world wide web , channel (broadcasting) , internet privacy , philosophy , linguistics , physics , optics , computer network
YouTube’s ‘up next’ feature algorithmically suggests videos to watchafter a video that is currently playing. This feature has been criticised for limitingusers’ exposure to diverse media content and information sources; meanwhile, YouTube hasreported that they have implemented technical and policy changes to address these concerns.Yet, there is limited data to support either the existing concerns or YouTube’s claims.Drawing on the concept of platform observability, this paper combines computational andqualitative methods to investigate the types of content YouTube’s ‘up next’ featureamplifies over time, using three search terms associated with sociocultural issues whereconcerns have been raised about YouTube’s role: ‘coronavirus’, ‘feminism’ and ‘beauty’. Oversix weeks, we collected the videos (and their metadata) that were highly ranked in thesearch results for each keyword, as well as the top-ranked recommendations associated witheach video, repeating the exercise for three steps in the recommendation chain. We thenexamined patterns in the recommended videos (and channels) for each query and theirvariation over time. We found evidence of YouTube's stated efforts to boost ‘authoritative’media outlets, but at the same time, misleading and controversial content continues to berecommended. We also found that while algorithmic recommendations offer diversity in videosover time, there are clear ‘winners’ at the channel level that are given a visibility boostin YouTube’s 'up next' feature. These impacts were attenuated differently depending on thenature of the search topic.