In this paper, we address two research topics in Recommender Systems (RSs) which have been developed in parallel without a deeper integration: Cross-Domain RS (CDRS) and Context-Aware RS (CARS). CDRS have emerged to enhance the quality of recommendations in a target domain by leveraging sources of information in different domains. CDRS are especially useful to address cold-start, sparsity and diversity problems in target domains with scarce information. CARS, on its turn, have been proposed to consider contextual information for recommendations. Such systems are suitable when the users’ interests change according to factors like time, location, among others. By combining these two approaches, better RSs can be developed, considering both the availability of useful data from multiple domains and the use of contextual information. In this paper, we formalize the combination of CDRS and CARS, which represents a more systematic integration of these approaches compared to previous work. Based on this formulation, we developed novel RSs techniques, named CD-CARS. To evaluate the developed CD-CARS techniques, we performed extensive experimentation through real datasets taking into account several scenarios. The recommendations were evaluated in terms of predictive and ranking performance, respectively achieving up to 62.6% and 45%, depending on the scenario, in comparison to traditional cross-domain collaborative filtering techniques. Therefore, the experimental results have shown that the integration of techniques developed in isolation can be useful in a variety of situations, in which recommendations can be improved by information gathered from different sources and can be refined by considering specific contextual information.

Introduction

A large number of Web sites and applications, such as Amazon,1 Netflix,2 Youtube,3 Last.fm,4 among many others, have adopted recommender systems (RS) (Adomavicius & Tuzhilin, 2005; Park, Kim, Choi, & Kim, 2012; Ricci, Rokach, Shapira, & Kantor, 2015) to provide their users with more relevant items. In the RS area, collaborative filtering (CF) is the most popular and widely implemented approach, since its implementation is relatively easy in different domains, and its quality is generally higher than other approaches, such as content-based filtering (CBF) (Nilashi, Ibrahim, & Bagherifard, 2018; Ricci et al., 2015). However, a common criticism of CF recommenders is that they tend to be biased toward popularity, constraining the degree of diversity (FernándezTobías, Cantador, Kaminskas, & Ricci, 2012). Furthermore, CFs are not able to recommend new items for which no ratings are available (a.k.a. cold-start problem) resulting in a low user satisfaction (Cantador, Fernández-Tobías, Berkovsky, & Cremonesi, 2015). In order to minimize these problems, Cross-Domain Recommender Systems (CDRS) (Cantador et al., 2015; Cremonesi, Tripodi, & Turrin, 2011; Fernández-Tobías et al., 2012; Gao et al., 2013; Taneja & Arora, 2018; Winoto & Tang, 2008) have been developed to use knowledge or user preferences acquired in a source domain to improve recommendation in a target domain where data is scarce (e.g. using a consolidated database of book preferences to recommend in a new movie recommendation application). Instead of handling each domain independently, CDRS recommend items of a target domain by exploring similarities between users considering ratings from source and target domains (Cremonesi et al., 2011).