Case-based travel recommendations.

Abstract

This chapter provides the methodological foundations and the rationale for approaching travel destination recommendation as a problem-solving activity. Case-base reasoning (CBR) is presented both as a cognitive plausible approach and as an integrated paradigm to build advisory systems. The chapter then describes an integrated solution called Trip@dvice that employs CBR. Finally, the chapter discusses the most important findings of various validation activities and future research.

Cooperative Query Rewriting for Decision Making Support and Recommender Systems

“…Considering the benefits and the limitations of both content-and collaborative-based filtering technologies, we have designed a novel hybrid collaborative=content based recommendation methodology, described in Ricci et al (2003) and Ricci et al (2006a). In this methodology, a unique human-machine interaction session is modeled as a case in a case-based reasoning framework (Aamodt and Plaza 1994;Bridge et al 2006).…”

“…Many recommendation techniques have been proposed to automate this mapping (Adomavicius and Tuzhilin 2005;Anand and Mobasher 2005;Burke 2002b) and there are many deployed recommender systems for CD, movies, music, and travel recommendation (MovieLens 2007;Linden et al 2003;Ricci et al 2006a). The most popular approach is called collaborative filtering, which recommends products to a user based on his commonalities (e.g., product rates) with other users (Resnick et al 1994;Shardanand and Maes 1995;Breese et al 1998;Sarwar et al 2001).…”

“…IQM is a conversational system that manages both the ''no result'' and ''too many results'' problems. When finally the user accepts a query, which is suggested by IQM, returning an acceptable set of items, these are ranked using a collaborative-by-content approach (Pazzani 1999), which exploits a double similarity computation Ricci et al 2006a). The architecture, algorithms, and evaluation of this IQM module is the core contribution of this article, and will be described in the next sections.…”

“…What emerges from this dialogue is that the clerk not only replies to the customer (decision maker), but also suggests modifications to the user's requirements (query) to keep the conversation alive and to better converge to a positive outcome, i.e., a user selection. In order to support such types of dialogues, we have developed a technology called interactive query management (IQM) that cooperatively and interactively helps the user to autonomously rewrite a query issued in the context of a decision-making problem supported by a recommender system (Ricci et al 2006a). The next section describes the logical architecture of IQM.…”

Location-Adapted Music Recommendation Using Tags

“…The problem of matching POI with a music track is more closely related to that found in cross-selling, e.g., recommending a type of boots that suit a kind of ski. This is a rather unpopular recommendation problem, that have only be considered by applications that recommend a good bundling of items, e.g., a travel planning [13] or music compilation [1].…”