Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Company announcement processing method for multi-task learning and server

A technology of multi-task learning and processing method, which is applied in the company announcement processing method and server field of multi-task learning, which can solve the problems of not learning related relations, complicated technical process, and long time-consuming, etc., and achieves the convenience of project deployment and maintenance , Improve learning efficiency and adaptability

Active Publication Date: 2019-09-10
深圳司南数据服务有限公司
View PDF5 Cites 22 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] 1. The existing technology needs to manually establish a rule lexicon, and the cost of manual maintenance is relatively high
[0005] 2. The existing technical feature extraction is based on the LSTM model. There is a length limit when extracting long-distance features, and distributed training cannot be performed. It requires a large amount of computing power and takes a long time
[0006] 3. The existing technical process is complicated, and it is necessary to classify the announcement before extracting the abstract
Did not learn a direct correlation with the task

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Company announcement processing method for multi-task learning and server
  • Company announcement processing method for multi-task learning and server
  • Company announcement processing method for multi-task learning and server

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0099] Please refer to Figure 1 to Figure 7 , Embodiment 1 of the present invention is:

[0100] Investors, researchers, and shareholders need to pay close attention to the announcements of listed companies. There are about 2,000 A-share announcements every day, and the peak period can reach more than 10,000. They need to spend a lot of time reading analysis reports. The following method is used to process tasks for each announcement, and investors will process the results according to the tasks to provide a basis for investment.

[0101] A company announcement processing method for multi-task learning, comprising steps:

[0102] S1. Input the historical announcement data into the shared layer of the multi-task learning model, and pre-train the historical announcement data through Bert;

[0103] S2. Input the data set corresponding to the processing task into the task layer of the multi-task learning model to train the multi-task learning model;

[0104] S3. Obtain the cur...

Embodiment 2

[0105] Please refer to Figure 1 to Figure 7 , the second embodiment of the present invention is:

[0106] A multi-task learning method for processing company announcements. On the basis of the first embodiment above, the processing tasks in this embodiment include sentiment classification, announcement classification, and abstract generation, that is, each announcement has its classification, emotion (favorable , negative, neutral) and core information (summary), investors will judge sentiment based on their classification, summary and sentiment, thus providing a basis for investment.

[0107] In this embodiment, before performing Bert pre-training, it also includes figure 2 and image 3 The data preprocessing steps shown, that is, before step S1, also include:

[0108] S0.1. Crawl web page information from financial websites that release announcement data to obtain public historical announcement data;

[0109] S0.2. Perform denoising processing on historical announcemen...

Embodiment 3

[0135] Please refer to Figure 8 , the third embodiment of the present invention is:

[0136] Investors, researchers, and shareholders need to pay close attention to the announcements of listed companies. There are about 2,000 A-share announcements every day, and the peak period can reach more than 10,000. They need to spend a lot of time reading analysis reports. The following server is used to process tasks for each announcement, and investors will provide a basis for investment based on the results of task processing.

[0137] A company announcement processing server 1 for multi-task learning, including a memory 3, a processor 2, and a computer program stored on the memory 3 and operable on the processor 3, when the processor 2 executes the computer program, the implementation is as in Embodiment 1 the method described.

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

No PUM Login to View More

Abstract

The invention discloses a company announcement processing method for multi-task learning and a server, and the method comprises the steps: inputting historical announcement data into a sharing layer of a multi-task learning model, and carrying out the pre-training of the historical announcement data through Bert; inputting the data set corresponding to the processing task into a task layer of themulti-task learning model to train the multi-task learning model; obtaining current announcement data, and inputting the current announcement data into the trained multi-task learning model to obtaina task processing result. The multi-task learning model is constructed through using in a transfer learning and multi-task learning mode, and the method has the advantages of being high in learning efficiency, high in generalization, low in manual maintenance cost, high in accuracy of multiple tasks, high in recall rate and convenient for engineering deployment and maintenance.

Description

technical field [0001] The present invention relates to the technical field of natural language processing, in particular to a multi-task learning company announcement processing method and a server. Background technique [0002] In the securities market, listed companies will regularly publish some prospectuses, listing announcements, semi-annual reports, quarterly reports, and interim reports to the public. Investors, researchers, and shareholders need to pay attention to the announcements of listed companies, and A-shares disclose announcements every day. There are about 2,000 articles, and the peak period can reach more than 10,000 articles. It takes a lot of time to read the analysis report, resulting in low efficiency in analyzing and judging specific items. [0003] The patent application number is CN201710255476.3, which discloses the following technical scheme based on the deep learning-based classification and abstract generation of listed company announcements: th...

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
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F16/35G06F16/33G06F17/27
CPCG06F16/35G06F16/3344G06F40/30Y02D10/00
Inventor 李勇辉魏志森杨双
Owner 深圳司南数据服务有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com