Loading...
Paper Type
Complete
Abstract
Using experts in recommender systems can improve the accuracy of recommendations as well as other quality aspects of recommendations. Most studies have tested the impact of expert knowledge in offline tests. However, it is still unclear how user behavior changes when experts are used for recommendation in an online scenario. We therefore deploy a live recommender system based on rules built by employed experts on the video-on-demand platform of a large television network. We find that expert-built rules lead to a similar amount of clip views and platform visits as a standard recommender. However, experts have an influence on the consumed content, focusing users on a few popular categories.
Recommended Citation
Herm-Stapelberg, Nils, "The Impact of Expert Knowledge on User Behavior in Recommender Systems" (2020). AMCIS 2020 Proceedings. 14.
https://aisel.aisnet.org/amcis2020/data_science_analytics_for_decision_support/data_science_analytics_for_decision_support/14
The Impact of Expert Knowledge on User Behavior in Recommender Systems
Using experts in recommender systems can improve the accuracy of recommendations as well as other quality aspects of recommendations. Most studies have tested the impact of expert knowledge in offline tests. However, it is still unclear how user behavior changes when experts are used for recommendation in an online scenario. We therefore deploy a live recommender system based on rules built by employed experts on the video-on-demand platform of a large television network. We find that expert-built rules lead to a similar amount of clip views and platform visits as a standard recommender. However, experts have an influence on the consumed content, focusing users on a few popular categories.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.