Small sample intention recognition method, system and device and storage medium

Abstract

The invention discloses a small sample intention recognition method, system and device and a storage medium, and the method comprises the steps: obtaining to-be-recognized text information, inputting the to-be-recognized text information to a preset slot extraction model for slot extraction, obtaining target slot text information, and storing the target slot text information in a database; based on a preset knowledge rule and a preset constructed multi-task sub-model, multi-task recognition is carried out on the target slot position text information, recognition results of different tasks are obtained, the multi-task sub-model comprises an intention classification model and an intention matching model, and the intention classification model and the intention matching model are used for matching the intention classification model and the intention matching model; the intention classification model and the intention matching model are both obtained by performing iterative training on a training corpus set constructed based on small sample learning, and based on a preset constructed target fusion model, performing fusion judgment on the recognition results of the different tasks to obtain a target intention recognition result. The technical problem that the accuracy of small sample intention recognition by a model is low is solved.

Description

technical field [0001] The present application relates to the technical field of machine learning, and in particular to a method, system, device and storage medium for small-sample intent recognition. Background technique [0002] The current small-sample intent recognition model is mainly based on a large amount of supervised data, and the text classification model is trained by machine learning or deep learning models to identify and classify. The model training effect is greatly affected by the quantity and quality of supervised data. However, in the initial construction of actual application scenarios, it is difficult to obtain a large amount of high-quality supervised corpus data, usually through manual labeling of the initial data, but the cost of manual labeling is high, the labeling cycle is long, and the semantic space of the labeling corpus is limited, and there is no real application. The scene produces a corpus that is biased, which in turn leads to a low accurac...

AI Technical Summary

Problems solved by technology

[0003] The main purpose of this application is to provide a small-sample intent recognition method, system, device, and storage medium, aiming to solve the technical problem of low accuracy of the models in the prior art for small-sample intent recognition

Images