Thesis Title
Modelling mobile user behaviors with contextual signals for intelligent assistance

Research Description
Due to the emergence and growth of machine learning and deep learning, the digital assistant has evolved by leaps and bounds in recent years. As a result of this development, next-generation DAs, which are the consequence of this development, can increasingly be found embedded within the devices we use every day. Research by Yonchanok aims to improve DAs by adding rich contextual signals retrieved directly from users’ devices that can better meet their needs. His work has three aspects: modeling mobile user behavior, profiling user attributes, and generating synthetic cross-domain behavior.

Supervisors
Prof Flora Salim
Dr Mohammad Saiedur Rahaman