Print Email Facebook Twitter Context-Aware Recommender Systems in the E-commerce Domain Title Context-Aware Recommender Systems in the E-commerce Domain Author van Kortenhof, B.L. Contributor Hauff, C. (mentor) Faculty Electrical Engineering, Mathematics and Computer Science Department Software Technology / Web Information Systems Programme Computer Science, track Information Architecture Date 2017-01-12 Abstract Recommender systems help users explore a large data set by proposing items in that data set that the system expects to be of interest to that user. The use of context, information that describes in which situation a user interacts with the recommender system, has shown to increase the effectiveness of recommender systems in several domains. In this research, we evaluate the effectiveness of recommender systems that use such contextual information, so-called context-aware recommender systems, in a domain that has not been addressed in literature: the e-commerce domain. In this domain, the goal of the recommender system is to activate users users to buy products in the data set. We explore six very different e-commerce data sets that contain contextual information and we show that the recommender’s effectiveness improves when contextual information is used in the recommendation process. We also find that it is difficult to predict which type of context will increase the effectiveness on which data set, and that not all users always benefit from using a context-aware recommender system. Finally, we also investigate the impact on revenue: when contextual information leads to an increase in effectiveness, the revenue generated by the recommender system is also increased. Subject recommender systemcontextcontext-aware recommender systemscarse-commerceuser experience To reference this document use: http://resolver.tudelft.nl/uuid:80363f35-b986-42fd-9880-e6ef96506daa Part of collection Student theses Document type master thesis Rights (c) 2017 van Kortenhof, B.L. Files PDF Masters thesis - Bas van ... 112830.pdf 4.06 MB Close viewer /islandora/object/uuid:80363f35-b986-42fd-9880-e6ef96506daa/datastream/OBJ/view