id	author	title	date	pages	extension	mime	words	sentence	flesch	summary	cache	txt
fcis-13124	Li, Qixuan	Joint Modelling of Slot Filling and Intent Detection in Constrained Resource Scenarios	2023	5	.pdf	application/pdf	3571	174	51	For the study of joint models for NLU, the earliest work used a three-layer CRF, where the three layers are token features, slot labels, and intent labels, whereas this architecture is better than performing two sub-tasks in a pipeline, and the first neural model for solving the joint task is (different from recurrent neural networks) Conclusion This study aims to explore the construction of joint models for intent recognition and slot filling in a constrained resource environment.	cache/fcis-13124.pdf	txt/fcis-13124.txt