Using contextual conditional preferences for recommendation tasks: a case study in the movie domain
Aleksandra Karpus, Tommaso Di Noia, Paolo Tomeo, Krzysztof Goczyła · 2016
Recommendation engines aim to propose users items they are interested in by looking at the user interaction with a system. However, individual interests may be drastically influenced by the context in which decisions are taken. We present an attempt to model user interests via a set of contextual conditional preferences. We show that usage of proposed preferences gives reasonable values of the accuracy and the precision even when the dataset is quite small.