Exploiting Re-ranking Information in a Case-Based Personal

Abstract

Intelligent software assistants are becoming more common in the e-commerce domain. We are working on a personal travel assistant. The goal of this application is to use case based reasoning to assist the user in arranging flights. It offers personalised service to its users and automatically learns their travel preferences. It stores these preferences in a user model that is directly related to the CBR process. It learns the user preferences by exploiting user feedback on sets of presented travel offers. When the user selects a preferred offer, the PTA establishes a preference ordering among the whole set. This ordering is calculated by measuring the similarity between the selected offer and each of the other offers. This ordering is used to rate these offers and store them in the user profile as cases. This ordering is also used to refine the user’s overall travel preferences by altering their personal similarity measure.