Method and system for automatically extracting Chinese financial events based on graph attention network

An automatic extraction and attention technology, applied in neural learning methods, biological neural network models, computer components, etc., can solve the problems of time-consuming and labor-intensive construction of event patterns, poor portability, and sparse data, so as to solve the problem of scarcity of extracted corpus. , the effect of good classification effect and high degree of feature interpretability

Pending Publication Date: 2022-03-18
BEIJING TECHNOLOGY AND BUSINESS UNIVERSITY
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Existing methods based on pattern matching are poor in portability, and the construction of event patterns is time-consuming and laborious, requiring the guidance of domain experts; methods based on traditional machine learning require large-scale annotation corpus, but the amount of data in the financial field is huge, and the existing annotation The corpus is sparse, which will lead to data sparse problems

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
  • Method and system for automatically extracting Chinese financial events based on graph attention network
  • Method and system for automatically extracting Chinese financial events based on graph attention network
  • Method and system for automatically extracting Chinese financial events based on graph attention network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0027] figure 1 It is a flowchart of a method for automatically extracting Chinese financial events based on a graph attention network, and a method for automatically extracting Chinese financial events based on a graph attention network includes the following steps:

[0028] Step S1: Preprocessing the data obtained from financial related websites, the processing flow is as follows figure 2 shown, including:

[0029] Step S11: Obtain relevant data of financial events and financial domain dictionaries from related domain websites. An example of a financial domain dictionary is as follows: {"Wall Street": 0, "Rodgers": 1, "Beijing": 2, "Tiger Securities": 3, "China": 4, "Appeared": 5, "Announced": 6 , "investment": 7, "Tencent": 8, "Tesla": 9, "Tencent Holdings": 10, "acquisition": 11, "representation": 12, "rise": 13}.

[0030] Step S12: This method uses HanLP and the financial field dictionary to process the data to obtain word segmentation results. Examples of word segmen...

Embodiment 2

[0069] Such as Figure 8 As shown, the embodiment of the present invention provides a system for automatically extracting Chinese financial events based on graph attention network, including the following modules:

[0070] Event library module, the event library is used to store marked events and representation vectors of prototype events, which will be continuously expanded and updated during the classification process;

[0071] The data preprocessing module preprocesses the acquired data: extract text, remove stop words, etc.; then use HanLP and the financial field dictionary to segment the data, and input the result of word segmentation into the Chinese financial event extraction model;

[0072] The Chinese financial event extraction model module uses the Chinese financial event extraction model to realize the classification of its event categories, and obtains the following results: the event category is "stock holdings"; by calculating the word vectors in the sen...

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 relates to a graph attention network-based Chinese financial event automatic extraction method and system, and the method comprises the steps: S1, obtaining the related data of a Chinese financial event and a financial domain dictionary from a related domain website, and carrying out the preprocessing of the obtained data; s2, constructing prototype events, manually marking financial event data of typical categories as seed data, and obtaining prototype events corresponding to each category; and S3, constructing a Chinese financial event automatic extraction model, classifying Chinese financial events to be classified, and extracting event elements. According to the method, small sample data can be utilized to expand Chinese financial event corpora, the problem of scarcity of Chinese event extracted corpora can be solved, and powerful data support is provided for related research; and meanwhile, the event extraction effect can be improved by utilizing the graph attention network and the bidirectional long-short-term memory network, and the prototype event and similarity calculation are combined, so that the demand of the model on the labeled sample can be greatly reduced, and the generalization ability of the model is improved.

Description

technical field [0001] The invention relates to the field of extracting Chinese financial events, in particular to a method and system for automatically extracting Chinese financial events based on a graph attention network. Background technique [0002] The rapid development of the Internet has brought massive data, how to manage and utilize these data has become a key and urgent issue in the field of information extraction. At the same time, the connection between the financial industry and the Internet is getting closer and closer. Major media platforms publish a large number of news related to the financial field every day. The financial events behind these news will have a significant impact on the fluctuations of the stock market and the investment decisions of enterprises. influences. For users, it is of great research significance to obtain hot financial events from massive texts. Event extraction is an important direction in the field of information extraction. It...

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): G06F16/33G06F16/35G06F40/194G06F40/242G06K9/62G06N3/04G06N3/08
CPCG06F16/3335G06F16/35G06F40/242G06F40/194G06N3/084G06N3/047G06N3/048G06N3/044G06F18/22G06F18/24
Inventor 段大高刘文文张秋丽韩忠明
Owner BEIJING TECHNOLOGY AND BUSINESS UNIVERSITY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap