We are now inundated with user data – in the digital world and in the real world – so it makes sense to try to mine that data to look for patterns and rules to guide our recommendation algorithms. We capture data streams from sensors, social media recommendations, mobile location-based information, and the evolving Internet of Things (IoT). The goal is to create a snapshot, or profile, of the user by understanding a person’s behavior when searching for a product, user activities when near a store that has a previously search-for product, and how social recommendations may influence a decision. The data tells the much of the user’s story, but we need tools and techniques to look for patterns, and turn those patterns into knowledge that can guide our algorithms in making smarter recommendations.
Data is being collected constantly on user behavior on the Web, by location-based services using mobile phones, tele-monitoring and home support systems, and on our mobile fitness apps, and by sensors, cameras, and the IoT. Our goal is to transform that data into knowledge in ways that support and enhance the user experience. We want to make recommender systems smarter and more responsive to user needs, so we need to understand our users better. One important requirement is for users to be able to provide feedback regarding the recommendations provided by the system. Another important factor is the role of social media in the way users are influenced in their decision-making.