Algorithm curation and the emergence of filter bubbles: an agent-based modelling approach

Research output: Contribution to conferencePosterAcademic

115 Downloads (Pure)

Abstract

Algorithmic curation is a helpful solution for the massive amount of content on the web. The term is used to describe how platforms automate the recommendation of content to users. News outlets, social networks and search engines widely use recommendation systems. Such automation has led to worries about selective exposure and its side effects. Echo chambers occur when we are over-exposed to the news we like or agree with, distorting our perception of reality (1). Filter bubbles arise where the information we dislike or disagree with is automatically filtered out – narrowing what we know (2). While the idea of Filter Bubbles makes logical sense, the magnitude of the "filter bubble effect", reducing diversity, has been questioned [3]. Most empirical studies indicate that algorithmic recommendations have not locked large audience segments into bubbles [4]. However, little attention has been paid to the interplay between technological, social, and cognitive filters. We proposed an Agent-based Modelling to simulate users' emergent behaviour and track their opinions when getting news from news outlets and social networks. The model aims to understand under which circumstances algorithmic filtering and social network dynamics affect users' innate opinions and which interventions can mitigate the effect. Agent-based models simulate the behaviour of multiple individual agents forming a larger society. The behaviour of the individual agents can be elementary, yet the population's behaviour can be much more than the sum of its parts. We have designed different scenarios to analyse the contributing factors to the emergence of filter bubbles. It includes different recommendation algorithms and social network dynamics. Cognitive filters are based on the Triple Filter Bubble model [5].

References

1.Richard Fletcher, 2020
2.Eli Pariser, 2012
3.Chitra & Musco, 2020
4. Möller et al., 2018
5. Daniel Geschke et al, 2018
Original languageEnglish
Number of pages1
Publication statusPublished - 14 Sept 2022
EventAI, media & democracy lab, opening event - Eye Filmmuseum , Amsterdam
Duration: 14 Sept 202214 Sept 2022

Conference

ConferenceAI, media & democracy lab, opening event
CityAmsterdam
Period14/09/2214/09/22

Fingerprint

Dive into the research topics of 'Algorithm curation and the emergence of filter bubbles: an agent-based modelling approach'. Together they form a unique fingerprint.

Cite this