Trip@dvice methodology

Trip@dvice methodology is fully described in:

Destination Recommendation Systems: Behavioural Foundations and Applications

by D R Fesenmaier, H Werthner, K W Wober.


Trip@dvice Methodology Book
Recommendation Technology
Case Base


To develop recommendations, Trip@dvice uses Case Based Reasoning (CBR) Technology. CBR is a problem solving methodology that faces a new problem by first retrieving a past, already solved similar case, and then uses that for solving the current one. In Trip@dvice completed travel plans are stored in the Case Base. The system memorizes them as good examples. During a user's travel recommendation session the system retrieves cases similar to the one under construction. The similarity function uses all the information coming in, including current and historic user and travel characteristics. This leads to highly personalized results.
 

Interaction

The Interaction technology helps to refine a user's query that has not returned a manageable set of items. A query fails when it returns either too many or no items. If a user's request returns no results, the system tells which conditions can not be satisfied at the same time and gives suggestions how to refine it in order to obtain some results. Vice versa, if a user's request returns too many results, the system suggests ways of refining it in order to obtain a reasonable and manageable number of results. Thus, on one hand, Trip@dvice avoids that the user leaves the Site due to unsatisfactory search results and on the other ensures that the user doesn't get overloaded with too many search results.

Ranking

The Ranking technology sorts suitable items from a catalogue and presents candidate trips based on user input and the satisfaction of other users on similar trips. Ranking can be used for complete travels as well as for single items of a travel, e.g. destination, accommodation, etc.

Mediator

The Mediator is a data integration system based on XML which provides a uniform view over different data sources. The electronic catalogues can be distributed across several databases, even if they have different structures. The Mediator collects heterogeneous data and makes it available in a uniform way.