Trip@dvise mobile extension of a case-based travel recommender system

In the majority of e-commerce applications for travel and tourism, users usually experience problems related to an overload of information and options to consider. Often the user receives a poor support in finding, filtering, comparing, and combining products. For the on-the-move traveller, finding interesting travel products is often highly expensive (time, and cognitive effort). We propose an interactive, case-based, and judgment-based approach to the problem of mapping travellers’ needs to the available travel products and services. The approach is case-based since it uses the knowledge of past recommendations when making a new one. The approach is based on an interactive adaptation of the dialogue since it involves the user in the process of searching products. The approach is judgment-based since it allows the user to provide feedbacks to the system’s recommendations, and incorporates such feedbacks in the successive recommendations. The user is not required to define a completely specified set of needs and wants at the beginning, but is guided through a recommendation process to interleave needs elicitation with recommendations. This approach has been implemented in a prototype which is here described