Training method, device, computer equipment and storage medium of intention recognition model

Abstract

This application relates to the field of artificial intelligence technology, and provides a training method, device, computer equipment, and storage medium for an intention recognition model. The method includes: obtaining sensitive text and normal text; performing text expansion processing on sensitive text based on random mosaic processing; Use the specified sensitive text and normal text as training samples, and obtain the contextual text data corresponding to the training samples; use the training samples and contextual text data as the input of the initial intent recognition model, and use the intent label as the output of the initial intent recognition model for model training , to obtain the first intent recognition model; verify the first intent recognition model based on the test sample data; if the verification is passed, use the first intent recognition model as the target intent recognition model. This application can alleviate the problem of sample category imbalance and increase the generalization ability of the training model. The present application can also be applied to the blockchain field, and data such as the above-mentioned target intention recognition model can be stored on the blockchain.

Description

technical field [0001] The present application relates to the technical field of artificial intelligence, in particular to a training method, device, computer equipment and storage medium for an intent recognition model. Background technique [0002] At present, the intelligent customer service system has been applied in various fields, including the financial field, e-commerce field, communication field and so on. During the dialogue between the customer service and the customer, the customer may inadvertently disclose content containing sensitive information. How to determine whether the voice of the customer during the dialogue has sensitive content has become an urgent problem to be solved. With the popularity of deep learning, the method based on the neural network model has become the mainstream sensitive content detection method, that is, the pre-trained recognition model is used to identify the intention of the speech input by the customer in the dialogue, and then j...

AI Technical Summary

Problems solved by technology

[0003] The main purpose of this application is to provide a training method, device, computer equipment, and storage medium for an intention recognition model, aiming at solving the problem of class imbalance in the existing positive and negative samples used to train and generate a recognition model, so that the training The generalization ability of the generated recognition model is poor, and the technical problem of the generated model’s poor intent prediction effect

Images