"Cyber-Social-Physical Features for Mood Prediction over Online Social Networks", article by Chaima Dhahri, Kazunori Matsumoto, Keiichiro Hoashi, DEIM Forum 2017, 6 pages.
Abstract : Context-Aware Recommendation Systems (CARS) are more effective when adapting their recommendations to a
specific user preference. Since modal context (mood) has a direct impact on user preferences, we aim at having an accurate mood
prediction to improve recommendation performance. Online social networks (OSNs) have grown rapidly over the last decade.
These social platforms provide the opportunity to gather the distributed online activities for each user. Tracking and aggregating
these data could result in useful insights for user modeling and understanding. In this paper, we built a personalized system that
can predict the upcoming user mood even in days without text-type tweets. We, first, studied the correlation of three types of
features (cyber, social and physical) with a user mood. Then, used these features to train a predictive system. The results suggest
a statistically significant correlation between user mood and his cyber, social and physical activities distributed among different
OSNs which leads to a low RMSE in our predictive system.