Incorporating Transition Matrix With Prior Knowledge and Dynamic Graph Attention Within Graph CRF for Joint Slot Filling and Intent Detection

Dialogue systems are widely used in human-machine interaction applications, such as voice assistants and service robots. Among the core tasks in these systems, intent detection and slot filling are particularly important. The strong semantic interdependence between these two tasks has driven the dev...

Descripción completa

Detalles Bibliográficos

Autores principales:	Reza Zadkamali, Hossein Zeinali, Saeedeh Momtazi
Formato:	Artigo
Lenguaje:	Inglês
Publicado:	IEEE 2026-01-01
Colección:	IEEE Access
Materias:	Natural language understanding intent detection slot filling graph-based CRF multi-task
Acceso en línea:	https://ieeexplore.ieee.org/document/11534220/
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!