A label knowledge graph powered multi-task framework for crowdsourcing and mobile crowd sensing tasks
Abstract: Highlights•A novel label knowledge graph powered multi-task framework (LKG-MF) for enhancing classification accuracy.•A label knowledge graph integrating heterogeneous semantics and relations to enhance label representation.•A task-adaptive feature extraction and loss function to mine the task-specific features.•When integrated into MCS platforms, LKG-MF leads to a 38.6% improvement in task assignment efficiency.
Loading