Publication at Faculty of Mathematics and Physics |
2014
Abstract
The following paper presents our work in progress in the area of learning user preference. User preference is often learned from explicit feedback (user rating), but on some domains, namely e-commerce, this kind of feedback is hard to come by, so collecting various implicit user actions is needed instead.
The paper categorizes usable user actions, other factors affecting user preference and describes IPIget - a component for observing implicit user preference indicators. The component is easily deployable on majority of e-commerce sites and available for download.
Several initial experiments are also described.