Automatic preference learning on semantic attributes

Identifier:  TFM:155

Handle: https://hdl.handle.net/20.500.11797/TFM155

Authors:  Naha, Kallol

Abstract:
Summary Recommender systems need to know the preferences of users to provide accurate suggestions. This information may be learnt by analysing the interaction of the user with the system. Previous works showed how to learn the preferences on numeric and categorical attributes. In this work we propose a framework to learn the preferences on uni-valued and multi-valued semantic attributes. The possibility to tune the learning parameters to obtain a good performance is shown in a case study of tourist destinations.