SORT: SOcial HelpeR SelecTion Scheme for Mobile Question Answering Systems

Abstract

Recent research has suggested that answers to the particular questions may be found when the questions are forwarded via suitable friends (or helpers) in a multi-hop manner. In mobile social network-based question answering systems, users are associated with limited resources; therefore, a light-weight scheme is crucial. However, improving the performance of the system with a light-weight scheme is challenging. To this end, we propose a distributed SOcial helpeR selecTion (SORT) scheme where all users share information with their 1-hop friends. SORT records and leverages the information to select and forward the questions to capable and cooperative helpers among the friends to improve the performance of the system. Our extensive trace-driven simulations reveal that, on average, SORT achieves higher response rate, higher best-answer rate, and lower response time by more than 14%, 13%, and 14%, respectively, in comparison to state-of-the-art helper selection schemes. Further, we observe that SORT performs consistently better than the state-of-the-art systems under a diverse number of system parameters.