Investigating the User Experience of Recommender Systems in Literature and in Movie Streaming Services
Author
Johansen, Patrick Skov
Term
4. Term
Publication year
2020
Pages
27
Abstract
This thesis investigates the user experience of recommender systems both in research literature and within movie streaming services. It begins with a literature review of more than fifty recent articles (2017–2020) from the ACM Digital Library, synthesized using Knijnenburg & Willemsen’s framework of three objective system aspects: Interaction, Algorithm, and Presentation. Across these, the review identifies themes in preference elicitation and feedback, input modality and interactivity; diversity, fairness, and context; and explanations. A key gap emerging from the literature is the scarcity of empirical studies of users’ perceptions, behaviors, and motives in everyday use outside controlled settings. Addressing this gap, the thesis reports interviews with 14 experienced and frequent movie streaming users to explore how they perceive, use, and interact with recommendations, aiming to derive implications for future system design. Detailed interview findings are not included in this excerpt, but the thesis offers a consolidated mapping of recent UX efforts and an empirical starting point to inform the design and evaluation of recommender systems.
Dette speciale undersøger brugeroplevelsen af anbefalingssystemer både i forskningslitteraturen og i konteksten af filmstreamingtjenester. Først gennemføres et litteraturreview af mere end 50 nyere artikler (2017–2020) fra ACM Digital Library, struktureret efter Knijnenburg & Willemsens ramme med tre systemaspekter: Interaktion, Algoritme og Præsentation. Reviewet identificerer temaer inden for præferenceindsamling og feedback, inputmodalitet og interaktivitet; diversitet, fairness og kontekst; samt forklaringer. På tværs af litteraturen peges der på en mangel på empiriske undersøgelser af brugeres opfattelser, adfærd og motiver i hverdagsbrug uden for kontrollerede miljøer. Med udgangspunkt i dette hul blev 14 erfarne og hyppige brugere af filmstreamingtjenester interviewet for at udforske, hvordan de opfatter, bruger og interagerer med anbefalinger, med henblik på at udlede designimplikationer for fremtidige systemer. Uddraget indeholder ikke detaljerede interviewresultater, men specialet bidrager med en samlet kortlægning af nyere UX-indsatser og et empirisk afsæt, der kan informere design og evaluering af anbefalingssystemer.
[This apstract has been generated with the help of AI directly from the project full text]
Keywords