Publication | Open Access
Know your algorithm: what media organizations need to explain to their users about news personalization
40
Citations
0
References
2019
Year
EngineeringMedia StandardsSocial Media CompaniesCommunicationJournalismText MiningInteractive JournalismComputational Social ScienceSocial MediaMedia OrganizationsTransparency InstrumentsNews RecommendationNews AnalyticsSocial Medium NewsNews SemanticsContent AnalysisDisinformation DetectionComputational JournalismMedia InstitutionsMedia DistributionData PrivacyPersonalized SearchPrivacy ConcernPublic Perception StudiesMedia PoliciesSocial ComputingNews PersonalizationArts
If the right to an explanation is expected to effectively safeguard users’ rights, it must be interpreted in a manner that takes the contextual risks algorithms pose to those rights into account. This article provides a framework of transparency instruments in the context of the news personalization algorithms employed by both traditional media organizations and social media companies. Explaining the impact on a user’s news diet and the role of editorial values in the algorithm is especially important in this context. Conversely, explanations of individual decisions and counterfactual explanations face specific practical and normative barriers that limit their utility.