Method and device for extracting short message text abstract based on named entity recognition
A named entity recognition and short message technology, applied in the field of text information recognition, can solve the problems of low efficiency and low accuracy, and achieve the effect of simple labeling data and improving accuracy and efficiency
Inactive Publication Date: 2020-08-28
上海创蓝云智信息科技股份有限公司
View PDF9 Cites 1 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0007] In view of the deficiencies in the above-mentioned prior art, the purpose of the present invention is to provide a method and device for extracting short message text summaries based on named entity recognition, so as to solve the problems of low accuracy and low efficiency in the prior art
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
[0032] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings, but the present invention can be implemented in many different ways defined and covered by the claims.
[0033] In order to solve the above problems, the present invention discloses a method and device for extracting short message text summaries based on named entity recognition, such as figure 1 described, including the following steps:
[0034] S1. Prepare the short message text collection to be extracted;
[0035] S2. Mark the SMS text collection, each text message needs to be marked with two parts, the organization entity word and the product entity word;
[0036] S3. Collect the labeled data of text messages for AI model training. The training model is the language model BERT and the conditional random field model CRF, expressed as y=f(x), where x is the text of the text message, and y is the summary of the text message, that is, in S2 Organize ...
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 relates to the technical field of text information recognition, in particular to a method and a device for extracting a short message text abstract based on named entity recognition. Themethod comprises the following steps of: S1, preparing a short message text set of an abstract to be extracted; s2, marking a short message text set, marking two parts of each short message text, andorganizing entity words and product entity words; s3, collecting annotation data of the short message text to perform AI model training; s4, after the AI model training in the S3 is completed, predicting and recognizing the data; and S5, taking the organization entity words and the product entity words predicted by the AI model as abstracts of the short message text. According to the invention, the short message abstract can be automatically extracted; data labeling is simple, and only two groups of words need to be labeled; the short message text abstract extracted by the trained AI model ishigh in accuracy and concise in content, and the accuracy and efficiency of short message text auditing are greatly improved.
Description
technical field [0001] The invention relates to the technical field of text information recognition, in particular to a method and device for extracting short message text summaries based on named entity recognition. Background technique [0002] As an SMS sending platform provider, a lot of resources are spent on storing and reviewing SMS. Therefore, it is very important to extract the text summary of the short message, and the summary can compress and simplify the original short message text to a great extent. Abstract extraction is based on semantic rules. This method relies heavily on expert knowledge and has many rules. It is also prone to errors in the face of informal grammar texts. Currently, deep learning methods are relatively rare and have not been applied in the field of Chinese text messages. To sum up, the prior art extracting short message text summaries based on semantic rules has low accuracy and low efficiency. Therefore, the present invention proposes a m...
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 Applications(China)
IPC IPC(8): G06F40/295G06F16/34
CPCG06F16/345G06F40/295
Inventor 元方唐小波宋争光郭乐郭盛楠
Owner 上海创蓝云智信息科技股份有限公司