Method and system for improving anti-noise capability of reply generation model
A technology for generating models and capabilities, applied in instruments, electrical digital data processing, computing, etc., can solve problems such as the expansion of input sentence distribution and training data distribution deviation, unsatisfactory performance, etc., to improve coding ability and anti-noise ability Effect
Pending Publication Date: 2020-10-16
AISPEECH CO LTD
View PDF0 Cites 1 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
A large amount of noise leads to an expansion of the deviation between the input sentence distribution and the training data distribution, making the mainstream reply generation system seq2seq-attention and Transformer perform unsatisfactorily in the face of noisy sentences
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0025] In order to make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be described clearly and completely in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of the embodiments of the present invention, not all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of the present invention.
[0026] It should be noted that the embodiments in the application and the features in the embodiments can be combined with each other if there is no conflict.
[0027] The invention can be described in the general context of computer-executable instructions executed by a computer, such as a program module. Generally, program modules in...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a method for improving the anti-noise capability of a reply generation model, and the reply generation model employs a coding-decoding architecture. The method comprises the steps: adding simulation noise into a training set so as to obtain a mixed training sample; inputting a mixed training sample into the reply generation model; performing binary classification on each character at a coding end output layer to identify noise characters; and adopting a multi-task training mode for the whole model, and training noise character recognition and reply generation at the same time. According to the method, the noise for simulating the real situation is added into the training sample in the training process, and the semantic features of the noise are learned through training, so that the coding capability of the reply generation model to the noise features is improved, and the anti-noise capability of the reply generation model is improved.
Description
Technical field [0001] The invention relates to the field of artificial intelligence technology, and in particular to a method and system for improving the anti-noise ability of a response generation model. Background technique [0002] In recent years, voice dialogue systems have been widely used in people's daily life, which greatly facilitates and enriches people's lives, such as smart navigation, smart speakers, smart homes, and so on. From the perspective of application, voice dialogue systems are mainly divided into two categories. One is task-oriented closed-domain dialogue systems. This type of dialogue system only receives task sentences in a limited field and completes user-specific tasks, such as navigation, playing music, and querying. Weather, etc.; the other is the open-domain dialogue system of small chat type, this type of dialogue system does not limit the user's conversation topic, receives any small chat sentences of the user, and allows the user to talk freely...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More IPC IPC(8): G06F40/232
CPCG06F40/232Y02T90/00
Inventor 朱钦佩缪庆亮俞凯
Owner AISPEECH CO LTD