Implicit Feedback Based Recommendation and Collaboration

Recommendation, collaboration and other tasks play important role on the adaptive web. For such tasks, user feedback is needed. Explicit feedback interrupts user and obtaining quality explicit feedback is problematic. On the other hand, users provide implicit feedback uninterrupted without knowing that they are rating. Traditional implicit feedback on the web – tracking of mouse and keyboard interaction, displayed parts of document, etc. – is problematic, when user passively reads the document and does not provide any inputs. We do not want to force him to provide inputs; therefore we have to track him physically. In our work we proposed a method for identification of important fragments based on implicit interest indicators with included commodity gaze tracking in common settings of the user’s home. We use collected information in recommendation of fragments, adaptive explicit feedback collection and we proposed additional scenarios.