Multi-stage and multi decision travel recommenders

The interactions between the user and the recommender system in a basic recommendation process (moving from needs to products with explanations), is not always linear and straightforward.
 
While some systems e.g. used by Amazon produce non-interactive recommendations, others such as www.activebuyersguide.com involves a user searching for a vacation in a multistage interaction. First, the site asks about the vacation’s general characteristics (type of vacation, activities, accommodation, and so forth). Second, it asks for details related to these characteristics, then for trade-offs between characteristics. Finally, it recommends destinations. It is argued that an interactive approach, where questions are fine-tuned as the human–machine interaction unfolds, has more potential [1].
 
Researchers have therefore argued that recommender systems should support multiple decision styles. The DieToRecs recommender [2] supports these decision styles by letting the user enter the system through: iterative single-item selection, complete travel selection, and inspiration-driven selection. Iterative single-item selection allows experienced users to efficiently navigate in the potentially overwhelming information space. The user can select whatever products he or she likes and in the preferred order, using the selections done up to a certain point (and in the past) to personalize the next stage. For example, if the user selects a particular destination, that destination is used to recommend a particular accommodation.
 
Complete travel selection gives to the user the option to select a personalized travel plan that bundles items by reusing the structure of travels built by other users in similar sessions. Finally, inspiration-driven selection lets the user choose a complete trip by means of a simpler user interface and a short interaction. This approach integrates case-based reasoning with interactive query refinement. Interactive query refinement allows a more flexible dialogue management that handles failures due to over- or under-specified user needs, suggesting precise repair actions. Case based reasoning therefore provides the framework to cast a recommendation session into a case and similarity-based ordering of both complete trips and single services [].

As Ricci et al argued that an effective travel recommender system should not only notice the user’s main needs or constraints in a top-down way but also allow the exploration of the option space and support the active construction of user preferences (in a bottom-up way). Recent research has emphasized this change of perspective, defining it as navigation by proposing. In this approach, the system shows the user examples of products, selected from those that the initial query retrieved. The user can choose a product as the current best choice, which updates the initial query and lets the recommender identify a new set of suggestions. These approaches use the concept of relevance feedback, used in information retrieval in a conversational, multistage interaction, in a dialog that interleaves needs elicitation with products. In conclusion, recommender systems for travel must carefully manage the human–machine dialogue to achieve usability and acceptance by the users.

[1] View at www.grouplens.org 
[2] F. Ricci, Daniel R. Fesenmaier, Nader Mirzadeh, Hildegard Rumetshofer, Erwin Schaumlechner, Adriano Venturini, Karl W. Wöber and Andreas H. Zins. DieToRecs: A Case-based Travel Advisory system. Travel Advisory System. In CAB International 2006. Destination Recommendation Systems: Behavioural Foundations and Applications (eds D.R. Fesenmaier, H. Werthner and K.W. Wöber)

<