CN116452353A - Financial data management method and system - Google Patents

Financial data management method and system Download PDF

Info

Publication number
CN116452353A
CN116452353A CN202310429431.9A CN202310429431A CN116452353A CN 116452353 A CN116452353 A CN 116452353A CN 202310429431 A CN202310429431 A CN 202310429431A CN 116452353 A CN116452353 A CN 116452353A
Authority
CN
China
Prior art keywords
data
financial
initial
text
word
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202310429431.9A
Other languages
Chinese (zh)
Inventor
戴卓君
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guizhou Wujiang Hydropower Development Co Ltd
Original Assignee
Guizhou Wujiang Hydropower Development Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Guizhou Wujiang Hydropower Development Co Ltd filed Critical Guizhou Wujiang Hydropower Development Co Ltd
Priority to CN202310429431.9A priority Critical patent/CN116452353A/en
Publication of CN116452353A publication Critical patent/CN116452353A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/12—Accounting
    • G06Q40/125—Finance or payroll
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/10—File systems; File servers
    • G06F16/17—Details of further file system functions
    • G06F16/174—Redundancy elimination performed by the file system
    • G06F16/1744—Redundancy elimination performed by the file system using compression, e.g. sparse files
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/10—Text processing
    • G06F40/166—Editing, e.g. inserting or deleting
    • G06F40/177—Editing, e.g. inserting or deleting of tables; using ruled lines
    • G06F40/18—Editing, e.g. inserting or deleting of tables; using ruled lines of spreadsheets
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/205—Parsing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/279—Recognition of textual entities
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/279—Recognition of textual entities
    • G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/19—Recognition using electronic means
    • G06V30/191—Design or setup of recognition systems or techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
    • G06V30/19147—Obtaining sets of training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/40—Document-oriented image-based pattern recognition
    • G06V30/41—Analysis of document content
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • General Engineering & Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Computational Linguistics (AREA)
  • Evolutionary Computation (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Business, Economics & Management (AREA)
  • Multimedia (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Software Systems (AREA)
  • Computing Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Databases & Information Systems (AREA)
  • Biomedical Technology (AREA)
  • Mathematical Physics (AREA)
  • Molecular Biology (AREA)
  • Biophysics (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • General Business, Economics & Management (AREA)
  • Medical Informatics (AREA)
  • Development Economics (AREA)
  • Evolutionary Biology (AREA)
  • Technology Law (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention belongs to the technical field of data management, and discloses a financial data management method and system. The method comprises the following steps: acquiring a financial file, and carrying out data analysis on the financial file to obtain a plurality of financial initial data in different data formats; carrying out data extraction on a plurality of financial initial data in different data formats to obtain a plurality of financial key data; carrying out data classification and data conversion on a plurality of financial key data to obtain a plurality of financial data tables of different categories; several tables of financial data of different categories are aggregated and managed. The system comprises a data analysis unit, a data extraction unit, a data classification unit, a data conversion unit and a table collection unit. The invention solves the problems of large labor cost investment, poor data flow, disordered management flow and low management efficiency in the prior art.

Description

Financial data management method and system
Technical Field
The invention belongs to the technical field of data management, and particularly relates to a financial data management method and system.
Background
Financial data refers to content reflecting financial status and business outcome of an enterprise. The method mainly comprises the following steps: financial account book data and report data, wherein the financial account data is data which is calculated according to real enterprise operation financial information statistics and then registered, and the report data mainly comprises: asset liability statement data, damage benefit statement data, cash flow statement data, etc., which pertains to the basic financial data of an enterprise. The data of analysis of various indexes of the enterprise are data calculated through a mathematical model or a corresponding formula, such as responsibility assessment data for various departments of the enterprise, financial management data for analyzing various indexes of the enterprise, decision analysis data for investment decision and the like.
When an enterprise analyzes operation business and general assets, statistics and analysis are required on financial data, and thus, the financial data is core data of the enterprise. However, the larger the enterprise scale, the more financial data that needs to be managed, and the existing financial data management mostly relies on manual mode, and the labor cost is high. In addition, the financial management tools adopted by each department, each project and each branch company are different, so that great differences exist in the data format and the recording mode of the financial file, the cross-platform, cross-department and cross-branch company application of the financial file cannot be realized, the data flow is poor, the management flow of the financial data is disordered, and the management efficiency is low.
Disclosure of Invention
The invention aims to solve the problems of large labor cost investment, poor data flow, disordered management flow and low management efficiency in the prior art, and provides a financial data management method and system.
The technical scheme adopted by the invention is as follows:
a financial data management method comprising the steps of:
acquiring a financial file, and carrying out data analysis on the financial file to obtain a plurality of financial initial data in different data formats;
Carrying out data extraction on a plurality of financial initial data in different data formats to obtain a plurality of financial key data;
carrying out data classification and data conversion on a plurality of financial key data to obtain a plurality of financial data tables of different categories;
several tables of financial data of different categories are aggregated and managed.
Further, acquiring a financial document, and performing data analysis on the financial document to obtain a plurality of initial financial data in different data formats, wherein the method comprises the following steps:
acquiring an initial financial document, decompressing the initial financial document, and obtaining a decompressed financial document;
acquiring a suffix name of each piece of financial initial data in the decompressed financial file;
performing data analysis on the decompressed financial file according to the suffix name of the financial initial data to obtain a plurality of financial initial data in different data formats; the data format of the financial initial data includes a text format and an image format.
Further, the data extraction is performed on a plurality of financial initial data in different data formats to obtain a plurality of financial key data, including the following steps:
performing data extraction on the text format financial initial data by using a data extraction model to obtain financial key data corresponding to the text format financial initial data;
Traversing all the text format financial initial data in the financial file to obtain financial key data corresponding to the text format financial initial data; the financial key data comprises financial keywords and corresponding financial values;
text recognition is carried out on the financial initial data in the image format by using a text recognition model to obtain corresponding financial text data, and data extraction is carried out on the financial text data by using a data extraction model to obtain financial key data corresponding to the financial initial data in the image format;
traversing all the financial initial data in the image format in the financial file to obtain financial key data corresponding to the financial initial data in a plurality of image formats.
Further, the data extraction model is constructed based on an IWOA-BILSTM-CRF algorithm, and the construction method of the data extraction model comprises the following steps:
crawling keyword entities of different categories in the financial field in the Internet by using a crawler tool to construct a financial keyword database;
extracting keyword entities in a financial keyword database as data extraction training samples to form a data extraction training sample set;
taking initial network parameters of the BILSTM network as optimization targets;
optimizing an optimization target by using an IWOA optimizing algorithm to obtain optimal initial network parameters of the BILSTM network;
Setting a network structure of the BILSTM network according to the optimal initial network parameters of the BILSTM network, and adding a CRF layer to the BILSTM network after the network structure is set to obtain an initial data extraction model;
and (3) inputting a data extraction training sample set to perform optimization training on the initial data extraction model to obtain an optimal data extraction model.
Further, an IWOA optimizing algorithm is used for optimizing an optimizing target to obtain an initial network parameter of BILSTM network optimization, and the method comprises the following steps:
taking an optimization target of the BILSTM network as the position of a whale individual in an IWOA optimization algorithm;
initializing IWOA optimizing algorithm parameters and IWOA population;
calculating the fitness value of each whale individual in the IWOA population, and reserving the optimal whale individual according to the fitness value of each whale individual;
randomly generating p, if p is less than 0.5 and |A| <1, executing behavior surrounding the prey, updating the position of the IWOA population, if p is less than 0.5 and |A|is more than or equal to 1, executing behavior searching the prey, updating the position of the IWOA population, and if p is more than or equal to 0.5, executing seeker network attack behavior, and updating the position of the IWOA population; wherein, p is an update parameter, A is a step length coefficient which introduces convergence factor optimization;
judging whether the iteration times meet the requirements or whether the optimal fitness value corresponding to the optimal whale individuals of the updated IWOA population meets the requirements, if so, outputting the position of the global optimal solution corresponding to the optimal whale individuals of the updated IWOA population to obtain the optimal initial network parameters of the BILSTM network, otherwise, updating the next generation IWOA population.
Further, the formula for surrounding prey behavior to update the IWOA population location is:
X 1 (t+1)＝X * (t)-AD
wherein X is 1 (t+1) is a whale individual position updated for surrounding hunting behavior; x is X * (t) is the optimal individual whale position; d is the distance between the current whale individual and the optimal whale individual; a=2ar—a, a is a convergence factor decreasing from 2 to 0;
the formula of the convergence factor is:
wherein a is a convergence factor; tan h () is a hyperbolic tangent function; t and tmax are the current iteration number and the maximum iteration number respectively; a, a max 、a min The maximum value and the small value of the convergence factor are respectively; λ is a decreasing rate parameter, k is a decreasing period parameter, λ= -2 pi, k=pi;
the formula for searching for prey behavior to update the IWOA population location is:
X 2 (t+1)＝X rand (t)-AD
wherein X is 2 (t+1) a whale individual location updated for search for hunting behavior; x is X rand (t) is a random whale individual position selected from the IWOA populationPlacing;
the formula for updating the bubble network attack behavior of the IWOA population position is as follows:
X 3 (t+1)＝D'c bl cos(2πl)+X * (t)
wherein X is 3 (t+1) is a whale individual location updated for bubble network aggression; d' is the current distance between the whale individual and the prey; b is a constant defining a spiral equation, b=1; l is [ - 1,1]Random numbers in between.
Further, the data extraction model comprises an input layer, a vector characterization layer, a BILSTM layer, a feature fusion layer, a CRF layer and an output layer which are connected in sequence;
Data extraction of text formatted financial initiation data/financial text data using a data extraction model comprising the steps of
Preprocessing the initial financial data/text financial data in a text format to obtain preprocessed data;
converting the preprocessed data into word sequences by using an input layer of a data extraction model;
converting a plurality of word segments in the word sequence into word vectors by using a vector characterization layer of the data extraction model to obtain a word sequence comprising a plurality of word vectors;
converting each word vector in the word sequence comprising a plurality of word vectors into a word vector by using a vector characterization layer of the data extraction model to obtain a word sequence comprising a plurality of word vectors;
extracting word semantic features of each word vector and word semantic features of each word vector by using a BILSTM layer of the data extraction model;
using a feature fusion layer of the data extraction model to perform feature fusion on word meaning features of all word vectors and word meaning features of all word vectors to obtain a fusion feature sequence;
using a CRF layer of the data extraction model, carrying out dependency processing on each word vector in the word sequence according to the fusion feature sequence, and adding a keyword entity tag to obtain a keyword entity tag word sequence;
Extracting corresponding financial keywords according to the keyword entity labels in the keyword entity label word sequence by using an output layer of the data extraction model;
using an output layer of the data extraction model, and extracting a corresponding financial value of the next position of the keyword entity tag in the keyword entity tag word sequence according to the position information of the keyword entity tag in the keyword entity tag word sequence;
and obtaining financial key data according to the financial key words and the corresponding financial values.
Further, the text recognition model is constructed based on a CTPN-CRNN algorithm, and the construction method of the text recognition model comprises the following steps:
acquiring a plurality of image data containing text areas, and forming a text positioning training sample set according to the plurality of image data containing the text areas;
acquiring a plurality of image data containing deformed characters, and forming a text recognition training sample set according to the plurality of image data containing the deformed characters;
acquiring a plurality of image data containing deformed text areas, and forming a text recognition training sample set according to the plurality of image data containing the deformed text areas;
according to the text positioning training sample set, performing optimization training by using a CTPN algorithm to obtain a text positioning sub-model;
According to the text recognition training sample set, performing optimization training by using a CRNN algorithm to obtain a text recognition sub-model;
and combining the text positioning sub-model and the text recognition sub-model, and inputting a text recognition training sample set for optimization training to obtain the text recognition model.
Further, data classification and data conversion are performed on the plurality of financial key data to obtain a plurality of financial data tables of different categories, including the following steps:
performing similarity matching on all financial keywords in the initial financial keyword data and keyword entities of different categories in the financial field in a financial keyword database to obtain a category corresponding to each financial keyword in the initial financial keyword data;
if the number of the financial keywords belonging to the same category in the initial financial key data exceeds a threshold value, the category is taken as the category of the initial financial key data; the initial category of financial critical data includes financial base data, liability assessment data financial management data, decision analysis data, stock financial data, and professional financial data;
null value processing, format normalization processing, data splitting processing, data correctness verification and data replacement processing are carried out on paired financial keywords and financial values in the initial financial key data, so that financial key data after data conversion are obtained;
According to the relation between the financial keywords in the financial key data after data conversion and the corresponding financial values, inputting the financial key data after data conversion into a blank financial data table to obtain a corresponding financial data table, and taking the initial category of the financial key data as the category of the corresponding financial data table;
traversing all the financial key data to obtain a plurality of corresponding financial data tables of different categories.
A financial data management system is used for realizing a financial data management method, and comprises a data analysis unit, a data extraction unit, a data classification unit, a data conversion unit and a form collection unit, wherein the data analysis unit, the data extraction unit, the data classification unit, the data conversion unit and the form collection unit are sequentially connected, and the form collection unit is connected with an external service management system;
the data analysis unit is used for receiving the financial file uploaded by the user, carrying out data analysis on the financial file to obtain a plurality of financial initial data with different data formats, and sending the plurality of financial initial data with different data formats to the data extraction unit;
the data extraction unit is used for carrying out data extraction on the plurality of financial initial data in different data formats sent by the data analysis unit to obtain a plurality of financial key data, and sending the plurality of financial key data to the data classification unit;
The data classification unit is used for classifying the data of the plurality of financial key data sent by the data extraction unit to obtain categories corresponding to the plurality of financial key data, and sending the plurality of financial key data of different categories to the data conversion unit;
the data conversion unit is used for carrying out data conversion on the plurality of financial key data of different categories sent by the data classification unit to obtain a plurality of financial data tables of different categories, and sending the plurality of financial data tables of different categories to the table collecting unit;
and the form collecting unit is used for collecting a plurality of financial data forms of different types sent by the data conversion unit according to the types of the financial data forms to obtain financial data form files of different types, compressing, packaging and naming the financial data form files of the same type according to the types of the financial data forms to obtain financial data form compression packages of different types, and sending the financial data form compression packages of different types to the service management system.
The beneficial effects of the invention are as follows:
1) According to the financial data management method provided by the invention, the automatic data analysis is carried out on the financial file, the data format of the financial initial data can be accurately identified, the practicability of the method is improved, the data extraction is carried out on the financial initial data with different data formats, only the financial key data is reserved, the cross-platform, cross-department and cross-branch company application of the financial file is realized, the data flow is improved, the financial key data avoids the interference and influence of other irrelevant information or data on the financial analysis, the subsequent financial data analysis is well laid, meanwhile, the process and the storage space of the financial data storage and management are simplified, the financial data is uniformly managed by establishing the financial data table, the management efficiency is improved, the management personnel can conveniently check and carry out statistical analysis, the manual data management is avoided, and the labor cost investment is reduced.
2) The financial data management system provided by the invention can automatically and intelligently process the financial files uploaded by the user, simplifies the processing flow of the financial data, avoids the manual mode of carrying out uniform format and statistical analysis on the financial files, reduces the labor cost investment, and improves the efficiency of financial data management by sending the financial data form compression packages of different types to the service management system.
Other advantageous effects of the present invention will be further described in the detailed description.
Drawings
FIG. 1 is a block flow diagram of a method of financial data management in accordance with the present invention.
FIG. 2 is a block diagram of the financial data management system of the present invention.
Detailed Description
The invention is further illustrated by the following description of specific embodiments in conjunction with the accompanying drawings.
Example 1:
as shown in fig. 1, the present embodiment provides a financial data management method, including the following steps:
acquiring a financial file, carrying out data analysis on the financial file to obtain a plurality of financial initial data with different data formats, and comprising the following steps:
acquiring an initial financial document, decompressing the initial financial document, and obtaining a decompressed financial document;
Acquiring a suffix name of each piece of financial initial data in the decompressed financial file;
performing data analysis on the decompressed financial file according to the suffix name of the financial initial data to obtain a plurality of financial initial data in different data formats;
the financial management tools adopted by each department, each project and each branch company are different, so that great differences exist in data formats and record modes of financial files, for example, some departments adopt EXCEL tools to generate financial initial data, the suffix name of the EXCEL tools is equal to xls, some departments adopt word tools to generate financial initial data, the suffix name of the EXCEL tools is equal to doc, some departments adopt PDF tools to generate financial initial data, the suffix name of the PDF tools is equal to PNG, and some departments adopt image tools to generate financial initial data, the suffix name of the PNG, so that the data formats of the financial initial data comprise text formats (& xls, & doc) and image formats (& PNG);
data extraction is carried out on a plurality of financial initial data in different data formats to obtain a plurality of financial key data, and the method comprises the following steps:
constructing a data extraction model;
the data extraction model is constructed based on an IWOA-BILSTM-CRF algorithm, and the construction method of the data extraction model comprises the following steps:
Optimizing an initial crawler tool by using a big-data SVM (support vector machine) -grid-search algorithm and a keyword similarity algorithm to obtain an optimized crawler tool; because the meaning of the financial data is very extensive and relates to massive keywords, the crawler tool is adopted to crawl different types of keyword entities in the financial field in the Internet, the traditional crawler tool is low in efficiency and accuracy, and the crawler tool with the big-data SVM grid-search algorithm and keyword similarity algorithm optimized can crawl the keyword entities more effectively, accurately and rapidly and selectively, so that the method is more suitable for the financial field;
crawling keyword entities of different categories in the financial field in the Internet by using the optimized crawler tool, and constructing a financial keyword database; the keyword entities of different categories stored in the financial keyword database are not only used for training samples of the data extraction model, but also used for subsequent data classification;
extracting keyword entities in a financial keyword database as data extraction training samples to form a data extraction training sample set;
taking initial network parameters of a Bi-directional Long Short-Term Memory (BILSTM) network as optimization targets, wherein the initial network parameters comprise the number of hidden layer neurons, initial weights and initial thresholds of the hidden layer neurons and initial learning rate of the BILSTM network;
Optimizing the optimization target by using an improved whale (Improved Whale Optimization Algorithm, IWOA) optimizing algorithm to obtain optimal initial network parameters of the BILSTM network, comprising the following steps:
taking an optimization target of the BILSTM network as the position of a whale individual in an IWOA optimization algorithm;
initializing IWOA optimizing algorithm parameters and IWOA population;
calculating the fitness value of each whale individual in the IWOA population, and reserving the optimal whale individual according to the fitness value of each whale individual;
randomly generating p, if p is less than 0.5 and |A| <1, executing behavior surrounding the prey, updating the position of the IWOA population, if p is less than 0.5 and |A|is more than or equal to 1, executing behavior searching the prey, updating the position of the IWOA population, and if p is more than or equal to 0.5, executing seeker network attack behavior, and updating the position of the IWOA population; wherein, p is an update parameter, A is a step length coefficient which introduces convergence factor optimization;
the formula for surrounding prey behavior to update IWOA population locations is:
X 1 (t+1)＝X * (t)-AD
wherein X is 1 (t+1) is a whale individual position updated for surrounding hunting behavior; x is X * (t) is the optimal individual whale position; d is the distance between the current whale individual and the optimal whale individual; a=2ar—a, a is a convergence factor decreasing from 2 to 0;
the formula of the convergence factor is:
Wherein a is a convergence factor; tan h () is a hyperbolic tangent function; t, t max The current iteration number and the maximum iteration number are respectively; a, a max 、a min The maximum value and the small value of the convergence factor are respectively; λ is a decreasing rate parameter, k is a decreasing period parameter, λ= -2 pi, k=pi;
the formula for searching for prey behavior to update the IWOA population location is:
X 2 (t+1)＝X rand (t)-AD
wherein X is 2 (t+1) search for prey linesFor updated whale individual positions; x is X rand (t) is a random whale individual location selected from the IWOA population;
the formula for updating the bubble network attack behavior of the IWOA population position is as follows:
X 3 (t+1)＝D'c bl cos(2πl)+X * (t)
wherein X is 3 (t+1) is a whale individual location updated for bubble network aggression; d' is the current distance between the whale individual and the prey; b is a constant defining a spiral equation, b=1; l is [ - 1,1]Random numbers in between;
the value of a is larger in the early iteration stage, the updated value of a is larger, the value of A is larger than or equal to 1, the IWA algorithm is enabled to be in the behavior of searching for the prey for a long time in the early iteration stage, the global searching capability of the algorithm is enhanced, the value of a is smaller in the later iteration stage, the updated value of A is smaller, the IWA algorithm is enabled to be in the behavior of surrounding the prey for a long time in the early iteration stage, the local surrounding capability of the algorithm is enhanced, the local hunting capability is improved, the convergence performance and the convergence speed of the BILSTM network are improved by the IGWO optimizing algorithm, the premature and sinking into the local optimal value of the BILSTM network are avoided, and the accuracy and the efficiency of data extraction are improved;
Judging whether the iteration times meet the requirements or whether an optimal fitness value corresponding to an optimal whale individual of the updated IWOA population meets the requirements, if so, outputting the position of a global optimal solution corresponding to the optimal whale individual of the updated IWOA population to obtain an optimal initial network parameter of the BILSTM network, otherwise, updating the next generation IWOA population;
according to the optimal initial network parameters of the BILSTM network, setting the network structure of the BILSTM network, solving the problem that the BILSTM is sensitive to the initial values of the network parameters, accelerating the training speed and accuracy of the model, and adding a linear chain member random field (Conditional Random Field, CRF) layer to the BILSTM network after the network structure is set to obtain an initial data extraction model;
inputting a data extraction training sample set to perform optimization training on the initial data extraction model to obtain an optimal data extraction model; the data extraction model comprises an input layer, a vector characterization layer, a BILSTM layer, a feature fusion layer, a CRF layer and an output layer which are connected in sequence;
data extraction is carried out on the financial initial data in a text format by using a data extraction model to obtain financial key data corresponding to the financial initial data in the text format, and the method comprises the following steps of
Preprocessing the initial financial data in a text format to obtain first preprocessed data;
converting the first preprocessed data into a first word sequence using an input layer of a data extraction model;
converting a plurality of word segments in the first word sequence into first word vectors by using a vector characterization layer of the data extraction model to obtain a first word sequence comprising a plurality of first word vectors;
converting each first word vector in the first word sequence comprising a plurality of first word vectors into a first word vector by using a vector characterization layer of the data extraction model to obtain a first word sequence comprising a plurality of first word vectors;
extracting word semantic features of each first word vector and word semantic features of each first word vector by using a BILSTM layer of the data extraction model;
feature fusion is carried out on the word semantic features of all the first word vectors and the word semantic features of all the first word vectors by using a feature fusion layer of the data extraction model, so that a first fusion feature sequence is obtained;
using a CRF layer of a data extraction model, performing dependency processing on each first word vector in the first word sequence according to the first fusion feature sequence, and adding a keyword entity tag to obtain a keyword entity tag first word sequence;
Extracting financial keywords corresponding to the financial initial data in a text format according to the keyword entity labels in the keyword entity label first word sequence by using an output layer of the data extraction model;
using an output layer of the data extraction model, extracting a corresponding financial value of a next position of the keyword entity tag at the keyword entity tag first word sequence according to the position information of the keyword entity tag in the keyword entity tag first word sequence;
according to the financial keywords and the corresponding financial values of the financial initial data in the text format, obtaining the financial key data corresponding to the financial initial data in the text format;
traversing all the text format financial initial data in the financial file to obtain financial key data corresponding to the text format financial initial data; the financial key data comprises financial keywords and corresponding financial values;
constructing a text recognition model; the text recognition model is constructed based on a CTPN-CRNN algorithm, and the construction method of the text recognition model comprises the following steps:
acquiring a plurality of image data containing text areas, and forming a text positioning training sample set according to the plurality of image data containing the text areas;
Acquiring a plurality of image data containing deformed characters, and forming a text recognition training sample set according to the plurality of image data containing the deformed characters;
acquiring a plurality of image data containing deformed text areas, and forming a text recognition training sample set according to the plurality of image data containing the deformed text areas;
performing optimization training by using a text detection (Detecting Text in Natural Image with Connectionist Text Proposal Network, CTPN) algorithm based on a connection pre-selection frame network according to the text positioning training sample set to obtain a text positioning sub-model;
according to the text recognition training sample set, performing optimization training by using a convolutional recurrent neural network (Convolutional Recurrent Neural Network, CRNN) algorithm to obtain a text recognition sub-model;
combining the text positioning sub-model and the text recognition sub-model, and inputting a text recognition training sample set for optimization training to obtain a text recognition model;
the CTPN algorithm can improve the positioning accuracy and efficiency of the text recognition model on the text region, and the CRNN algorithm can improve the text recognition accuracy and efficiency of the text recognition model on the text region, so that the text recognition model can accurately and rapidly recognize the text data of the financial initial data in the image format;
Text recognition is carried out on the financial initial data in the image format by using a text recognition model to obtain corresponding financial text data, and the method comprises the following steps:
inputting the financial initial data in the image format into a text recognition model;
using a text positioning sub-model of the text recognition model to position a text region of the financial initial data in the image format, and obtaining the text region of the financial initial data in the image format;
image segmentation is carried out on the financial initial data in the image format according to the text region, so that all text region images corresponding to the financial initial data in the image format are obtained;
using a text recognition sub-model of the text recognition model to perform text recognition on all text area images corresponding to the financial initial data in the image format to obtain a plurality of corresponding image text data;
combining a plurality of image text data according to the positions and the sequence of all the text area images in the financial initial data in the image format to obtain financial text data corresponding to the financial initial data in the image format;
and carrying out data extraction on the financial text data by using a data extraction model to obtain financial key data corresponding to the financial initial data in an image format, wherein the method comprises the following steps of:
Preprocessing the financial text data to obtain second preprocessed data;
converting the second preprocessed data into a second word sequence using an input layer of the data extraction model;
converting a plurality of word segments in the second word sequence into second word vectors by using a vector characterization layer of the data extraction model to obtain a second word sequence comprising a plurality of second word vectors;
converting each second word vector in the second word sequence comprising a plurality of second word vectors into a second word vector by using a vector characterization layer of the data extraction model to obtain a second word sequence comprising a plurality of second word vectors;
extracting word semantic features of each second word vector and word semantic features of each second word vector by using a BILSTM layer of the data extraction model;
using a feature fusion layer of the data extraction model to perform feature fusion on word semantic features of all second word vectors and word semantic features of all second word vectors to obtain a second fusion feature sequence;
using a CRF layer of the data extraction model, performing dependency processing on each second word vector in the second word sequence according to the second fusion feature sequence, and adding a keyword entity tag to obtain a keyword entity tag second word sequence;
Extracting financial keywords corresponding to the financial text data according to the keyword entity labels in the keyword entity label second word sequence by using an output layer of the data extraction model;
using an output layer of the data extraction model, extracting a corresponding financial value of a next position of the keyword entity tag at the keyword entity tag second word sequence according to the position information of the keyword entity tag in the keyword entity tag second word sequence;
according to the financial keywords of the financial text data and the corresponding financial values, obtaining financial key data corresponding to the financial initial data in an image format;
traversing all the financial initial data in the image format in the financial file to obtain financial key data corresponding to the financial initial data in a plurality of image formats;
data classification and data conversion are carried out on a plurality of financial key data to obtain a plurality of financial data tables of different categories, and the method comprises the following steps:
performing similarity matching on all financial keywords in the initial financial keyword data and keyword entities of different categories in the financial field in a financial keyword database to obtain a category corresponding to each financial keyword in the initial financial keyword data;
If the number of the financial keywords belonging to the same category in the initial financial key data exceeds a threshold value, the category is taken as the category of the initial financial key data; the categories of initial financial key data include financial base data, liability assessment data, financial management data, decision analysis data, stock financial data, and professional financial data;
null value processing, format normalization processing, data splitting processing, data correctness verification and data replacement processing are carried out on paired financial keywords and financial values in the initial financial key data, so that financial key data after data conversion are obtained; if the keyword entity tag word sequence does not have a financial value at the next position of the keyword entity tag, a null value exists in the initial financial key data, and the unpaired financial key words and financial values in the initial financial key data are deleted; format normalization processing, data splitting processing, data correctness verification and data replacement processing are processing of formats, lengths and correctness of paired financial keywords and financial values, and validity and standardability of the financial key data are guaranteed;
According to the relation between the financial keywords in the financial key data after data conversion and the corresponding financial values, inputting the financial key data after data conversion into a blank financial data table to obtain a corresponding financial data table, and taking the initial category of the financial key data as the category of the corresponding financial data table;
traversing all financial key data to obtain a plurality of corresponding financial data tables of different categories;
the method for collecting and managing the financial data tables of different categories comprises the following steps:
collecting a plurality of financial data tables belonging to the same category to obtain financial data table files of different categories;
according to the types of the financial data tables, compressing, packaging and naming the financial data table files of the same type to obtain financial data table compression packages of different types;
and sending the financial data table compression packages of different categories to a business management system for management.
According to the financial data management method provided by the invention, the automatic data analysis is carried out on the financial file, the data format of the financial initial data can be accurately identified, the practicability of the method is improved, the data extraction is carried out on the financial initial data with different data formats, only the financial key data is reserved, the cross-platform, cross-department and cross-branch company application of the financial file is realized, the data flow is improved, the financial key data avoids the interference and influence of other irrelevant information or data on the financial analysis, the subsequent financial data analysis is well laid, meanwhile, the process and the storage space of the financial data storage and management are simplified, the financial data is uniformly managed by establishing the financial data table, the management efficiency is improved, the management personnel can conveniently check and carry out statistical analysis, the manual data management is avoided, and the labor cost investment is reduced.
Example 2:
as shown in fig. 2, the present embodiment provides a financial data management system, for implementing a financial data management method, where the system includes a data analysis unit, a data extraction unit, a data classification unit, a data conversion unit, and a table collection unit, where the data analysis unit, the data extraction unit, the data classification unit, the data conversion unit, and the table collection unit are sequentially connected, and the table collection unit is connected with an external service management system;
the data analysis unit is used for receiving the financial file uploaded by the user, carrying out data analysis on the financial file to obtain a plurality of financial initial data with different data formats, and sending the plurality of financial initial data with different data formats to the data extraction unit;
the data extraction unit is used for carrying out data extraction on the plurality of financial initial data in different data formats sent by the data analysis unit to obtain a plurality of financial key data, and sending the plurality of financial key data to the data classification unit;
the data classification unit is used for classifying the data of the plurality of financial key data sent by the data extraction unit to obtain categories corresponding to the plurality of financial key data, and sending the plurality of financial key data of different categories to the data conversion unit;
The data conversion unit is used for carrying out data conversion on the plurality of financial key data of different categories sent by the data classification unit to obtain a plurality of financial data tables of different categories, and sending the plurality of financial data tables of different categories to the table collecting unit;
and the form collecting unit is used for collecting a plurality of financial data forms of different types sent by the data conversion unit according to the types of the financial data forms to obtain financial data form files of different types, compressing, packaging and naming the financial data form files of the same type according to the types of the financial data forms to obtain financial data form compression packages of different types, and sending the financial data form compression packages of different types to the service management system.
The financial data management system provided by the invention can automatically and intelligently process the financial files uploaded by the user, simplifies the processing flow of the financial data, avoids the manual mode of carrying out uniform format and statistical analysis on the financial files, reduces the labor cost investment, and improves the efficiency of financial data management by sending the financial data form compression packages of different types to the service management system.
The invention is not limited to the alternative embodiments described above, but any person may derive other various forms of products in the light of the present invention. The above detailed description should not be construed as limiting the scope of the invention, which is defined in the claims and the description may be used to interpret the claims.

Claims (10)

1. A financial data management method, characterized by: the method comprises the following steps:
acquiring a financial file, and carrying out data analysis on the financial file to obtain a plurality of financial initial data in different data formats;
carrying out data extraction on a plurality of financial initial data in different data formats to obtain a plurality of financial key data;
carrying out data classification and data conversion on a plurality of financial key data to obtain a plurality of financial data tables of different categories;
several tables of financial data of different categories are aggregated and managed.
2. A method of financial data management as claimed in claim 1, wherein: acquiring a financial file, carrying out data analysis on the financial file to obtain a plurality of financial initial data with different data formats, and comprising the following steps:
acquiring an initial financial document, decompressing the initial financial document, and obtaining a decompressed financial document;
Acquiring a suffix name of each piece of financial initial data in the decompressed financial file;
performing data analysis on the decompressed financial file according to the suffix name of the financial initial data to obtain a plurality of financial initial data in different data formats; the data format of the financial initial data comprises a text format and an image format.
3. A method of financial data management as claimed in claim 2, wherein: data extraction is carried out on a plurality of financial initial data in different data formats to obtain a plurality of financial key data, and the method comprises the following steps:
performing data extraction on the text format financial initial data by using a data extraction model to obtain financial key data corresponding to the text format financial initial data;
traversing all the text format financial initial data in the financial file to obtain financial key data corresponding to the text format financial initial data; the financial key data comprises financial key words and corresponding financial values;
text recognition is carried out on the financial initial data in the image format by using a text recognition model to obtain corresponding financial text data, and data extraction is carried out on the financial text data by using a data extraction model to obtain financial key data corresponding to the financial initial data in the image format;
Traversing all the financial initial data in the image format in the financial file to obtain financial key data corresponding to the financial initial data in a plurality of image formats.
4. A method of financial data management according to claim 3, wherein: the data extraction model is constructed based on an IWOA-BILSTM-CRF algorithm, and the construction method of the data extraction model comprises the following steps:
crawling keyword entities of different categories in the financial field in the Internet by using a crawler tool to construct a financial keyword database;
extracting keyword entities in a financial keyword database as data extraction training samples to form a data extraction training sample set;
taking initial network parameters of the BILSTM network as optimization targets;
optimizing an optimization target by using an IWOA optimizing algorithm to obtain optimal initial network parameters of the BILSTM network;
setting a network structure of the BILSTM network according to the optimal initial network parameters of the BILSTM network, and adding a CRF layer to the BILSTM network after the network structure is set to obtain an initial data extraction model;
and (3) inputting a data extraction training sample set to perform optimization training on the initial data extraction model to obtain an optimal data extraction model.
5. A method of financial data management as claimed in claim 4, wherein: optimizing an optimization target by using an IWOA optimizing algorithm to obtain optimal initial network parameters of a BILSTM network, comprising the following steps:
taking an optimization target of the BILSTM network as the position of a whale individual in an IWOA optimization algorithm;
initializing IWOA optimizing algorithm parameters and IWOA population;
calculating the fitness value of each whale individual in the IWOA population, and reserving the optimal whale individual according to the fitness value of each whale individual;
randomly generating p, if p is less than 0.5 and |A| <1, executing behavior surrounding the prey, updating the position of the IWOA population, if p is less than 0.5 and |A|is more than or equal to 1, executing behavior searching the prey, updating the position of the IWOA population, and if p is more than or equal to 0.5, executing seeker network attack behavior, and updating the position of the IWOA population; wherein, p is an update parameter, A is a step length coefficient which introduces convergence factor optimization;
judging whether the iteration times meet the requirements or whether the optimal fitness value corresponding to the optimal whale individuals of the updated IWOA population meets the requirements, if so, outputting the position of the global optimal solution corresponding to the optimal whale individuals of the updated IWOA population to obtain the optimal initial network parameters of the BILSTM network, otherwise, updating the next generation IWOA population.
6. A method of financial data management as claimed in claim 5, wherein: the formula for surrounding prey behavior to update IWOA population locations is:
X 1 (t+1)＝X * (t)-AD
wherein X is 1 (t+1) is a whale individual position updated for surrounding hunting behavior; x is X * (t) is the optimal individual whale position; d is the distance between the current whale individual and the optimal whale individual; a=2ar—a, a is a convergence factor decreasing from 2 to 0;
the formula of the convergence factor is:
wherein a is a convergence factor; tan h () is a hyperbolic tangent function; t, t max The current iteration number and the maximum iteration number are respectively; a, a max 、a min The maximum value and the small value of the convergence factor are respectively; λ is a decreasing rate parameter, k is a decreasing period parameter, λ= -2 pi, k=pi;
the formula for searching for prey behavior to update the IWOA population location is:
X 2 (t+1)＝X rand (t)-AD
wherein X is 2 (t+1) is more behavior for searching for preyNew whale individual location; x is X rand (t) is a random whale individual location selected from the IWOA population;
the formula for updating the bubble network attack behavior of the IWOA population position is as follows:
X 3 (t+1)＝D'c bl cos(2πl)+X * (t)
wherein X is 3 (t+1) is a whale individual location updated for bubble network aggression; d' is the current distance between the whale individual and the prey; b is a constant defining a spiral equation, b=1; l is [ - 1,1]Random numbers in between.
7. A method of financial data management as claimed in claim 4, wherein: the data extraction model comprises an input layer, a vector characterization layer, a BILSTM layer, a feature fusion layer, a CRF layer and an output layer which are connected in sequence;
data extraction of text formatted financial initiation data/financial text data using a data extraction model comprising the steps of
Preprocessing the initial financial data/text financial data in a text format to obtain preprocessed data;
converting the preprocessed data into word sequences by using an input layer of a data extraction model;
converting a plurality of word segments in the word sequence into word vectors by using a vector characterization layer of the data extraction model to obtain a word sequence comprising a plurality of word vectors;
converting each word vector in the word sequence comprising a plurality of word vectors into a word vector by using a vector characterization layer of the data extraction model to obtain a word sequence comprising a plurality of word vectors;
extracting word semantic features of each word vector and word semantic features of each word vector by using a BILSTM layer of the data extraction model;
using a feature fusion layer of the data extraction model to perform feature fusion on word meaning features of all word vectors and word meaning features of all word vectors to obtain a fusion feature sequence;
Using a CRF layer of the data extraction model, carrying out dependency processing on each word vector in the word sequence according to the fusion feature sequence, and adding a keyword entity tag to obtain a keyword entity tag word sequence;
extracting corresponding financial keywords according to the keyword entity labels in the keyword entity label word sequence by using an output layer of the data extraction model;
using an output layer of the data extraction model, and extracting a corresponding financial value of the next position of the keyword entity tag in the keyword entity tag word sequence according to the position information of the keyword entity tag in the keyword entity tag word sequence;
and obtaining financial key data according to the financial key words and the corresponding financial values.
8. A method of financial data management according to claim 3, wherein: the text recognition model is constructed based on a CTPN-CRNN algorithm, and the construction method of the text recognition model comprises the following steps:
acquiring a plurality of image data containing text areas, and forming a text positioning training sample set according to the plurality of image data containing the text areas;
acquiring a plurality of image data containing deformed characters, and forming a text recognition training sample set according to the plurality of image data containing the deformed characters;
Acquiring a plurality of image data containing deformed text areas, and forming a text recognition training sample set according to the plurality of image data containing the deformed text areas;
according to the text positioning training sample set, performing optimization training by using a CTPN algorithm to obtain a text positioning sub-model;
according to the text recognition training sample set, performing optimization training by using a CRNN algorithm to obtain a text recognition sub-model;
and combining the text positioning sub-model and the text recognition sub-model, and inputting a text recognition training sample set for optimization training to obtain the text recognition model.
9. A method of financial data management as claimed in claim 4, wherein: data classification and data conversion are carried out on a plurality of financial key data to obtain a plurality of financial data tables of different categories, and the method comprises the following steps:
performing similarity matching on all financial keywords in the initial financial keyword data and keyword entities of different categories in the financial field in a financial keyword database to obtain a category corresponding to each financial keyword in the initial financial keyword data;
if the number of the financial keywords belonging to the same category in the initial financial key data exceeds a threshold value, the category is taken as the category of the initial financial key data; the categories of initial financial key data include financial base data, liability assessment data, financial management data, decision analysis data, stock financial data, and professional financial data;
Null value processing, format normalization processing, data splitting processing, data correctness verification and data replacement processing are carried out on paired financial keywords and financial values in the initial financial key data, so that financial key data after data conversion are obtained;
according to the relation between the financial keywords in the financial key data after data conversion and the corresponding financial values, inputting the financial key data after data conversion into a blank financial data table to obtain a corresponding financial data table, and taking the initial category of the financial key data as the category of the corresponding financial data table;
traversing all the financial key data to obtain a plurality of corresponding financial data tables of different categories.
10. A financial data management system for implementing a financial data management method as claimed in any one of claims 1 to 9, wherein: the system comprises a data analysis unit, a data extraction unit, a data classification unit, a data conversion unit and a form collection unit, wherein the data analysis unit, the data extraction unit, the data classification unit, the data conversion unit and the form collection unit are sequentially connected, and the form collection unit is connected with an external service management system;
The data analysis unit is used for receiving the financial file uploaded by the user, carrying out data analysis on the financial file to obtain a plurality of financial initial data with different data formats, and sending the plurality of financial initial data with different data formats to the data extraction unit;
the data extraction unit is used for carrying out data extraction on the plurality of financial initial data in different data formats sent by the data analysis unit to obtain a plurality of financial key data, and sending the plurality of financial key data to the data classification unit;
the data classification unit is used for classifying the data of the plurality of financial key data sent by the data extraction unit to obtain categories corresponding to the plurality of financial key data, and sending the plurality of financial key data of different categories to the data conversion unit;
the data conversion unit is used for carrying out data conversion on the plurality of financial key data of different categories sent by the data classification unit to obtain a plurality of financial data tables of different categories, and sending the plurality of financial data tables of different categories to the table collecting unit;
and the form collecting unit is used for collecting a plurality of financial data forms of different types sent by the data conversion unit according to the types of the financial data forms to obtain financial data form files of different types, compressing, packaging and naming the financial data form files of the same type according to the types of the financial data forms to obtain financial data form compression packages of different types, and sending the financial data form compression packages of different types to the service management system.
CN202310429431.9A 2023-04-20 2023-04-20 Financial data management method and system Pending CN116452353A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310429431.9A CN116452353A (en) 2023-04-20 2023-04-20 Financial data management method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310429431.9A CN116452353A (en) 2023-04-20 2023-04-20 Financial data management method and system

Publications (1)

Publication Number Publication Date
CN116452353A true CN116452353A (en) 2023-07-18

Family

ID=87125252

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310429431.9A Pending CN116452353A (en) 2023-04-20 2023-04-20 Financial data management method and system

Country Status (1)

Country Link
CN (1) CN116452353A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116757334A (en) * 2023-08-16 2023-09-15 江西科技学院 Financial data processing method, system, readable storage medium and computer

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116757334A (en) * 2023-08-16 2023-09-15 江西科技学院 Financial data processing method, system, readable storage medium and computer
CN116757334B (en) * 2023-08-16 2023-11-24 江西科技学院 Financial data processing method, system, readable storage medium and computer

Similar Documents

Publication Publication Date Title
CN108415953B (en) Method for managing bad asset management knowledge based on natural language processing technology
CN110275936B (en) Similar legal case retrieval method based on self-coding neural network
CN113312501A (en) Construction method and device of safety knowledge self-service query system based on knowledge graph
CN112732871B (en) Multi-label classification method for acquiring client intention labels through robot induction
CN111428028A (en) Information classification method based on deep learning and related equipment
CN113505200A (en) Sentence-level Chinese event detection method combining document key information
CN113742733B (en) Method and device for extracting trigger words of reading and understanding vulnerability event and identifying vulnerability type
CN115952292B (en) Multi-label classification method, apparatus and computer readable medium
CN113836896A (en) Patent text abstract generation method and device based on deep learning
CN110222192A (en) Corpus method for building up and device
CN113064990A (en) Hot event identification method and system based on multi-level clustering
CN115577080A (en) Question reply matching method, system, server and storage medium
CN114611491A (en) Intelligent government affair public opinion analysis research method based on text mining technology
CN116150361A (en) Event extraction method, system and storage medium for financial statement notes
CN116452353A (en) Financial data management method and system
CN116150651A (en) AI-based depth synthesis detection method and system
CN110781297A (en) Classification method of multi-label scientific research papers based on hierarchical discriminant trees
CN114328800A (en) Text processing method and device, electronic equipment and computer readable storage medium
CN113342949A (en) Matching method and system of intellectual library experts and topic to be researched
CN116629258B (en) Structured analysis method and system for judicial document based on complex information item data
CN113535928A (en) Service discovery method and system of long-term and short-term memory network based on attention mechanism
CN111061939B (en) Scientific research academic news keyword matching recommendation method based on deep learning
CN112749530B (en) Text encoding method, apparatus, device and computer readable storage medium
CN115186138A (en) Comparison method and terminal for power distribution network data
AU2021428503A1 (en) Systems and methods for generating document numerical representations

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination