In this article, we give an overview of ideas for integrating context in recommender systems in general and specifically in various mobile application domains. Our main case study is an approach for vehicular ad-hoc networks (VANETs). The system recommends gas stations based on driver preferences, ratings of other users and context information such as the current location and fuel level of a car. We explain the main design issues behind our recommender. Our approach first filters items based on preferences and context, and then takes ratings of other users and additional information into account, which can be relayed from car to car in a VANET. We also outline other mobile scenarios for contextualized recommender systems: a system for recommending mobile applications based on user context, an approach to find relevant resources in mobile semantic personal information management and a decentralized recommender system for personal digital assistants (PDAs).
Today, a huge variety of sensors is built into vehicles enabling them to obtain a view of their surroundings based on locally observed data. With VANETs (vehicular ad-hoc networks) as an emerging technology, it is possible both to disseminate data to other vehicles and to collect information in order to improve their own picture of the current situation. However, distributed applications based on VANETs still need to agree on a "common understanding" of context for interoperability on a contextual level. In recent years, ontologies have become the "silver bullet" for describing concepts and their relationship in a semantically rich manner. In this paper, we apply this approach to context modeling and propose a collision avoidance application in order to show that applications can benefit from the advantages of ontological models like distributed composition, partial validation, richness and quality of information and a certain level of formality. Also, we want to prove that calculations on the model are still fast enough to fulfill real-time requirements imposed by active safety for vehicles. Two different scenarios with different context models but unchanged applications show the advantages of ontologies with respect to extensibility of models. In addition, it is shown that larger and more complex models do not necessarily increase the execution times.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.