INTENT IDENTIFICATION METHOD BASED ON DEEP LEARNING NETWORK

摘要

The present invention relates to the field of intelligent identification. Disclosed is an intent identification method based on a deep learning network, wherein the technical problem of low intent identification accuracy is solved. The key point of the technical scheme of the present invention is to migrate a feature of a first deep learning network to a second deep learning network, and mainly lies in: converting a data set of all fields into a word sequence WS and a corresponding Pinyin sequence PS, and performing manual labeling on a data set of a certain field and converting the data set into a word sequence WD, a Pinyin sequence PD and a label; inputting the word sequence WS and the Pinyin sequence PS into the first deep learning network for training and obtaining a language model; initializing and updating a coding layer parameter matrix of the language model; and then, after the word sequence WD and the Pinyin sequence PD are input into the second deep learning network for coding, weighting the word sequence WD and the Pinyin sequence PD and inputting same into the second deep learning network for training an intent identification model. The intent identification model has high intent identification accuracy.