Nowadays, Recommender Systems have played a crucial role in several entertainment scenarios by making personalised recommendations and guiding the entire users’ journey from their first interaction. Recent works have addressed it as a Contextual Bandit by providing a sequential decision model to explore items not tried yet (or not tried enough) or exploit the best options learned so far. However, this work noticed these current algorithms are limited to naive non-personalised approaches in the first interactions of a new user, offering random or most popular items. Through experiments in three domains, we identify a negative impact of these first choices. Our study indicates that the bandit performance is directly related to the choices made in the first trials. Then, we propose a new approach to balance exploration and exploitation in the first interactions and handle these drawbacks. This approach is based on the Active Learning theory to catch more information about the new users and improve their long-term experience. Our idea is to explore the potential information gain of items that can also please the user’s taste. This method is named Warm-Starting Contextual Bandits, and it statistically outperforms 10 benchmarks in the literature in the long run.
Current advances in Deep Neural Networks have motivated recent studies to return their interest in Review-Aware Recommender Systems (RSs). In this sense, we employ a systematic mapping approach by selecting 56 papers published on the main vehicles of the area, such as RecSys, VLDB, SIGIR, and others. Reading these works and synthesising their achievements, we provide an updated picture of this field by highlighting relevant outcomes, contributions, and limitations. Especially, we have identified two huge limitations in the area. First, many of these works do not provide their code sources. Second, there is a prevalent tendency to prioritize accuracy above other quality dimensions, despite the consensus within the community to assess the practical effectiveness of RSs through many metrics. Both observations create constraints that restrict reproducibility and hinder straightforward comparisons in the field. Addressing this gap, this work also provides an open-source library that compiles the principal strategies proposed in the literature. Furthermore, we conducted a comprehensive evaluation through several datasets and metrics. Our findings indicate that while no single algorithm demonstrates absolute superiority, RSs based on neural networks have particularly exhibited the most competitive performance.
Interactive recommendation has been recognized as a Multi-Armed Bandit (MAB) problem. Items are arms to be pulled (i.e., recommended) and the user’s satisfaction is the reward to be maximized. Despite the advances, there is still a lack of consensus on the best practices to evaluate such solutions. Recently, two complementary frameworks were proposed to evaluate bandit solutions more accurately: iRec and OBP. The first one has a complete set of offline metrics and bandit models that allows us to perform an comparisons with several evaluation policies. The second one provides a huge set of bandit models to be evaluated through several counterfactual estimators. However, there is a room to be explored when joining these two frameworks. We propose and evaluate an integration between both, demonstrating the potential and richness of such combination.
Location-Based Social Networks (LBSNs) have become important tools for people interested in exploring new places. And, similar to traditional recommendation domains, handling the trade-off between accuracy and diversity is a major challenge to provide useful recommendations. However, this domain adds an equally relevant dimension to this challenge: the geographical distance between users and each point-of-interest (POI). Besides understanding how the characteristics of services offered by each POI fit the user needs, realizing how far users are willing to move to fulfill these needs is of paramount relevance. Moreover, the users present distinct levels of interest in diversification. In this paper, we propose a strategy to provide POI recommendations linearly combining categorical and geographical diversifications in a personalized way. Indeed, our strategy is able to prioritize quality dimensions that better suit the personalized needs of each user, with gains up to 10% when compared with unpersonalized versions.