Attention-Based Recurrent Neural Network Models for Joint Intent Detection and Slot Filling

摘要

Attention-based encoder-decoder neural network models have recently shownpromising results in machine translation and speech recognition. In this work,we propose an attention-based neural network model for joint intent detectionand slot filling, both of which are critical steps for many speechunderstanding and dialog systems. Unlike in machine translation and speechrecognition, alignment is explicit in slot filling. We explore differentstrategies in incorporating this alignment information to the encoder-decoderframework. Learning from the attention mechanism in encoder-decoder model, wefurther propose introducing attention to the alignment-based RNN models. Suchattentions provide additional information to the intent classification and slotlabel prediction. Our independent task models achieve state-of-the-art intentdetection error rate and slot filling F1 score on the benchmark ATIS task. Ourjoint training model further obtains 0.56% absolute (23.8% relative) errorreduction on intent detection and 0.23% absolute gain on slot filling over theindependent task models.