Method, device, device and storage medium for generating text summarization
A text and abstract technology, applied in the field of text abstract generation methods, devices, equipment and storage media, can solve the problems of semantic loss, abstract sentences are not smooth, low accuracy, etc., achieve high accuracy and readability, and improve performance. Ability, effect of good summary generation function
Active Publication Date: 2022-06-24
PINGAN INT SMART CITY TECH CO LTD
View PDF3 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0006] In order to solve the technical problems in the prior art that the abstract sentences extracted or generated are not smooth, semantic loss or low accuracy
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
[0039] In order to make the purposes, technical solutions and advantages of the embodiments of the present application more clear, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments It is a part of the embodiments of this application, but not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of the present application. It should be understood that the specific embodiments described herein are only used to explain the present application, but not to limit the present application.
[0040] The text abstract generation method provided in this application can be applied in such as figure 1 , in which the terminal device can be but not limited to vari...
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
This application proposes a method, device, device, and storage medium for generating text abstracts. This application uses the sample text marked by the sample abstract to perform model training on the pre-trained Bert model, and conducts model training on the detection model during the training process through random inspection. Detection, to roll back the training node from the bad training node to the better historical training node in time, and then correct the training direction through the rollback operation until the training is completed. While this application effectively improves Bert's performance on the summary model, it also tries to avoid catastrophic forgetting and other loss of basic functions of the Bert model during the training and fine-tuning process, ensuring that the final trained summary generation model retains both In addition to the original powerful functions of the Bert model, it also has a good summary generation function, thereby ensuring that the generated summary has high accuracy and readability.
Description
technical field [0001] The present application relates to the technical field of natural language processing, and in particular, to a text abstract generating method, apparatus, device and storage medium. Background technique [0002] In the prior art, methods for obtaining abstracts of articles include extractive abstracts, RNN and seq-to-seq models, or generative abstracts such as Transformer. [0003] The traditional extractive abstract is mainly generated by the extractive method. The extractive method is limited to searching in the original text, and the obtained result is also the paragraph of the original text. The extraction method mainly consists of three main steps: sentence importance evaluation, sentence de-duplication operation, and sentence length constraints. The importance of a sentence can be determined by using tf-idf and textrank or heuristic rules. The general method is to use textrank to use the similarity of sentences, use sentences as nodes to constru...
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 Patent Type & Authority Patents(China)
IPC IPC(8): G06F16/34G06F40/289G06F40/30
CPCG06F16/345G06F40/30G06F40/289
Inventor 王改吴少康
Owner PINGAN INT SMART CITY TECH CO LTD