CN111583033A - Association analysis method and device based on relation between listed company and stockholder - Google Patents
Association analysis method and device based on relation between listed company and stockholder Download PDFInfo
- Publication number
- CN111583033A CN111583033A CN202010252323.5A CN202010252323A CN111583033A CN 111583033 A CN111583033 A CN 111583033A CN 202010252323 A CN202010252323 A CN 202010252323A CN 111583033 A CN111583033 A CN 111583033A
- Authority
- CN
- China
- Prior art keywords
- stock
- listed
- shareholder
- early warning
- trading
- 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
Links
Images
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
- G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/27—Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
Abstract
The invention discloses a correlation analysis method based on the relation between listed companies and stockholders, which comprises the following steps: acquiring registration information and stockholder information of a listed company, and establishing a related network of the listed company and the stockholder; searching each shareholder holding the target stock code according to the target stock code, further searching other listed companies and stock codes thereof to which each shareholder belongs, and establishing a first sub-association network; and searching each listed company which holds the shareholder according to the target shareholder, further searching other shareholders of each listed company, and establishing a second sub-association network. The invention also provides a correlation analysis device based on the relation between the listed companies and the stockholders. The invention provides for the query of the supervisory personnel, determines suspicious stockholders or listed companies and discovers abnormal stock holding behaviors by constructing the complex association network among the stockholders.
Description
Technical Field
The invention relates to the field of computer software testing. More particularly, the invention relates to a method and a device for analyzing association based on the relation between listed companies and stockholders.
Background
With the continued development of the financial market, more and more funds are being introduced into the securities market. Analysis and monitoring of stock keeping relationships in the stock market has become a challenging issue, with hundreds of shareholders per listed company, and with stocks held by multiple companies. Many internet public data sets provide information about the registration and stock keeping of companies, including information about the top ten stockholders, the ratio of the number of stocks kept to the number of stocks kept, and the relationship of the company. However, these data are scattered data sets in units of listed companies, and it is difficult to find an abnormal shareholder or listed company from a unidirectional shareholding relationship. The supervisors can only rely on experience in the past, and start with suspicious stockholders or listed companies to discover abnormal stock holding behaviors according to their knowledge of the listed companies and historical cases. But for abnormal cases beyond experience and history, it is difficult for the supervisor to effectively monitor. No effective solution has been proposed to the above problems.
Disclosure of Invention
The invention aims to provide a correlation analysis method and a correlation analysis device based on the relation between listed companies and stockholders, which are used for a supervisor to inquire by constructing a complex correlation network between stockholders, determine suspicious stockholders or listed companies and find abnormal stock holding behaviors.
To achieve these objects and other advantages in accordance with the purpose of the invention, there is provided an association analysis method based on a connection between a listed company and a stockholder, comprising:
acquiring registration information and stockholder information of a listed company, and establishing a related network of the listed company and the stockholder;
searching each shareholder holding the target stock code according to the target stock code, further searching other listed companies and stock codes thereof to which each shareholder belongs, and establishing a first sub-association network;
and searching each listed company which holds the shareholder according to the target shareholder, further searching other shareholders of each listed company, and establishing a second sub-association network.
Preferably, the association analysis method based on the contact between the listed company and the shareholder determines the stock codes of which the number of historical early warning records reaches a first threshold value in the first sub-association network and marks the stock codes as early warning stock codes.
Preferably, the association analysis method based on the relationship between the listed companies and the shareholders determines the shareholders with early warning stock codes reaching the second threshold value in the second sub-association network and marks the shareholders as suspected coordinators.
Preferably, the method for analyzing the association between the listed company and the shareholder comprises the following steps: the method comprises the steps of obtaining multi-channel heterogeneous data, classifying and storing the obtained heterogeneous data according to structured data and unstructured data to build an information perception and monitoring database, building a multi-dimensional monitoring index, building an integrated discrimination model with the multi-dimensional monitoring index as a characteristic, carrying out real-time monitoring on data in the information perception and monitoring database by adopting the discrimination model, outputting a primary early warning result, and screening and forming a sub early warning result based on a core index on the basis of the primary early warning result.
Preferably, the method for analyzing the relationship between the listed companies and the shareholders includes at least registered capital, company name, and listing time, and the shareholders includes at least shareholders type and holding stock list.
Preferably, the association analysis method based on the relationship between the listed company and the shareholder stores the registration information and the shareholder information in the stock holding relationship database, and updates the stock holding relationship database at intervals of a set time period.
Preferably, the method for analyzing the association between the listed company and the shareholder comprises the following steps:
acquiring a trading sample, wherein the trading sample comprises trading behaviors of stock codes at all times and stock market conditions at corresponding times;
dividing a plurality of time intervals, extracting transaction characteristics from transaction behaviors in each time interval, and extracting stock market characteristics from stock market conditions in each time interval;
taking the transaction characteristics and the stock market characteristics in the time interval as input, taking the transaction characteristics in the following time interval as output, and training to obtain a neural network prediction model;
and acquiring the trading behavior, trading characteristics, stock market conditions and stock market characteristics of the target stock code in a time interval, inputting the trading behavior, trading characteristics, stock market conditions and stock market characteristics into a neural network prediction model, and if the difference between the output trading characteristics of the next time interval and the actual trading characteristics of the next time interval is larger than a third threshold, early warning the target stock code.
The invention also provides a correlation analysis device based on the relation between the listed companies and the stockholders, which comprises the following steps:
a processor;
a memory storing executable instructions;
wherein the processor is configured to execute the executable instructions to perform the association analysis method based on the association between the listed company and the shareholder.
The invention at least comprises the following beneficial effects:
the invention collects information of stockholders of listed companies, information of important institution investors and products thereof, and establishes a database for storing the information; collecting registration information of listed companies and other holding information of stockholders, and constructing a knowledge graph; and constructing an incidence relation network of a listed company and an investment stockholder by using the data, realizing intelligent search of specific stocks, mechanism investors and related products, and displaying an incidence relation diagram of the specific stocks, the specific mechanisms or the products. The method has the advantages that the complex association relationship among the listed companies taking the investment stockholders as the links is intelligently displayed for the first time, and a powerful tool is provided for analyzing the behaviors of the listed companies and investors.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention.
Drawings
Fig. 1 is an overall frame diagram of the present invention.
Fig. 2 is a diagram of an association network according to an embodiment of the present invention.
Detailed Description
The present invention is further described in detail below with reference to the attached drawings so that those skilled in the art can implement the invention by referring to the description text.
It will be understood that terms such as "having," "including," and "comprising," as used herein, do not preclude the presence or addition of one or more other elements or groups thereof.
In one technical solution, a method for analyzing association based on a relationship between a listed company and a shareholder, includes:
acquiring registration information and stockholder information of a listed company, and establishing a related network of the listed company and the stockholder;
searching each shareholder holding the target stock code according to the target stock code, further searching other listed companies and stock codes thereof to which each shareholder belongs, and establishing a first sub-association network;
and searching each listed company which holds the shareholder according to the target shareholder, further searching other shareholders of each listed company, and establishing a second sub-association network.
In the above technical solution, the company type shareholder collects company registration information including company nature, affiliated industry, legal person, stock book, etc., and the individual type shareholder collects individual stock holding information, time and quantity of holding other stocks, etc. And extracting the association records, and constructing the listed companies and stockholder association networks of each reporting quarter according to the time section. The overall associated network includes information for all listed companies and stockholders in the quarter. And constructing a query index according to stock codes and shareholder names in the associated network. And constructing a special mode sub-network which can be used for retrieving, outputting and displaying the sub-associated networks according to listed companies and stakeholders, namely the first sub-associated network and the second sub-associated network. It can be seen that the technical scheme is used for the supervision personnel to inquire by constructing the complex association network among the shareholders, so that the suspicious shareholders or listed companies can be conveniently determined, and abnormal stock holding behaviors can be found.
Preferably, the association network visually displays different nodes on the front end in different ways for the user to freely view. The nodes are divided into different colors according to different attributes, the central node is the stock to be inquired and is displayed in orange, and light-colored halo marks are arranged around the central node. The nodes connected with the stocks are stockholders and products thereof, the stockholders are marked with blue, and the products are marked with blue. The shareholder node may further be associated with other early warning stocks, which are shown in red. And the names (and codes) of stocks, stockholders and products corresponding to the nodes are correspondingly displayed on each node. When the mouse cursor moves to the node and the edge between the nodes, the detailed information of the node or the edge, including the name, the time, the number and the like, is displayed in a floating window mode.
For example:
step 1: build stock keeping relation database of all listed companies and keep daily continuous update
The information of the top ten shares of east holdings of all listed companies is collected and stored into an elastic search distributed database by date index. The database field information is as follows:
change ratio of bdbl quantity
rdate report period
Sharehdatio holdup to total strand ratio
sharehdnum holdup number
bz thigh holding variation
Step 2: the registration information of listed companies and the basic information of stockholders are collected and kept updated continuously every day, and the main database fields are as follows:
listed company registration information:
registered _ Fund registered capital
plate to which plate belongs
industry of industry
registered _ time registration time
legal representative of legal _ person
firm _ type Business type
firm _ name company name
listed time to market
ev market value
tp exchange
dashboard _ status large screen display status
Address company address
manipulate early warning record
Shareholder information:
sharehdtype shareholder type
Hold _ stock holding stock list
And step 3: and constructing an association network of stockholders and listed companies by using the established database, and storing the association network in a Neo4j database.
In the network diagram, the entity design is as follows:
1. stockholders: including companies and individuals, wherein the stakeholders of the company are divided into funds, investment companies, trusts, insurance, social security, security dealer, collective financing plan, finance company, enterprise annuity, finance, colleges, basic pension funds, foreign accounts, countries, special accounts, etc
2. Stock: the attributes include stock code, stock name, and early warning record
3. The product is as follows: attributes include company of interest, number of holdings, etc
After the entity is constructed, reading the database according to the following association relationship, and importing the database into the overall relational network:
the network is divided into pieces according to time when being imported, the time granularity is day, and the network held on the day can be inquired according to the day.
And after the construction is completed, constructing a stock code index in the database so as to be convenient for query.
And 4, step 4: and designing a subgraph construction mode to support intelligent subgraph searching.
Designing a subgraph construction mode as follows:
1. sub-association network with city company as core: given a suspected stock code of a listed company, the shareholder holding the listed company's stock is searched and further other listed companies held by the shareholder are sought, forming a multi-level associative subnetwork.
2. Sub-associative networks with shareholders as cores: given a suspicious shareholder, all listed companies holding stocks are searched, and shareholders of the listed companies are further searched to form a multi-level association sub-network.
In another technical scheme, the association analysis method based on the contact between the listed company and the shareholder determines the stock codes of which the number of historical early warning records reaches a first threshold value in the first sub-association network and marks the stock codes as early warning stock codes. According to the technical scheme, the suspicious stock codes in the first sub-associated network are marked through marking the early warning records, so that the monitoring personnel can conveniently check and supervise the suspicious stock codes. The pre-warning may be achieved by prior art or experience. The first threshold is preferably obtained by a domain expert after multi-azimuth experiments and investigations.
In another technical scheme, the association analysis method based on the connection between the listed companies and the shareholders determines the shareholders with early warning stock codes reaching the second threshold value in the second sub-association network, and marks the shareholders as suspected coordinators. According to the technical scheme, suspicious shareholders in the second sub-associated network are marked by marking suspected coordinators, so that the supervision and inspection are facilitated for a supervisor. The early warning can be realized by the prior art or experience as in the above technical scheme. The second threshold is preferably obtained by a domain expert after multi-directional experiments and investigations.
In another technical solution, the method for performing early warning on a stock code based on a correlation analysis method of a connection between a listed company and a shareholder includes: the method comprises the steps of obtaining multi-channel heterogeneous data, classifying and storing the obtained heterogeneous data according to structured data and unstructured data to build an information perception and monitoring database, building a multi-dimensional monitoring index, building an integrated discrimination model with the multi-dimensional monitoring index as a characteristic, carrying out real-time monitoring on data in the information perception and monitoring database by adopting the discrimination model, outputting a primary early warning result, and screening and forming a sub early warning result based on a core index on the basis of the primary early warning result.
The technical scheme provides a mode of early warning, and other patents of the applicant can be referred. Specifically, the method comprises the following steps:
step one, collecting multi-channel heterogeneous data;
determining a perception source for identifying information type manipulation according to an implementation means of the information type manipulation, and acquiring massive unstructured text data and other structured data by a customized strategy;
the specific multichannel heterogeneous data acquisition adopts an expandable crawler cluster technology based on Scapy and a dynamic webpage grabbing technology based on a Selenium tool to acquire data of social media, financial media and regulatory agency channels;
specifically, the data of the open internet is regularly crawled every day, and comprises social media represented by microblogs, WeChat public numbers, post bars and known letters, financial media represented by stock bars, east wealth networks and New and useless finance, and a supervision organization official network represented by certificate guild, exchange and deep exchange, and third-party financial software data represented by Wander and national Taian is regularly acquired every day on the basis of an API (application program interface) data interface technology, wherein the third-party financial software data comprises transaction data, financial data and the like;
the data comprises transaction data, financial data and the like, the transaction data refers to closing prices, volume of trades and the like of all stocks on the market every day, and the financial data is data of three financial statements of listed companies;
step two, constructing an information perception and monitoring database;
classifying and storing the obtained heterogeneous data according to structured data and unstructured data, and simultaneously corresponding to each listed company name to form a listed company information perception and monitoring database index and an information perception and monitoring database; the constructed information sensing and monitoring database is obtained by storing structured data and unstructured data based on an Elasticissearch engine database of a highly-expanded distributed full-text search engine so as to realize near-real-time data storage and search;
the structured data comprises transaction data and financial data, wherein the transaction data comprises opening price, closing price, daily maximum price, daily minimum price, transaction amount, hand-changing rate, bulk transaction date, bulk transaction bargaining price, bulk transaction bargaining amount, bulk transaction trading overflow rate, bulk transaction buyer business department, bulk transaction seller business department and the like of stock of listed companies, and the financial data comprises profit statement data represented by net profits of the listed companies, flow cash statement data represented by cash flow net amount generated by operation activities and asset liability statement data represented by accounts receivable;
the unstructured data comprises stock right structure information, announcement information, official information disclosure data and Internet public opinion data;
in order to realize the correspondence between the unstructured text data and the listed companies, the listed company index of the monitoring database is completed according to the names, short names, stock names and short names of the listed companies and the hit conditions of stock codes in the text;
step three, constructing and calculating a multi-dimensional detection index;
establishing a multi-dimensional monitoring index, establishing an integrated discrimination model taking the multi-dimensional monitoring index as a characteristic, adopting the discrimination model to perform real-time monitoring on data in an information perception and monitoring database, and outputting a first-stage early warning result;
the method specifically comprises the following steps: the information type operation case for checking and analyzing the certificate and prison penalty judges summarizes the behavior characteristics of the listed company, the related institutions and the related personnel for implementing the information type operation, and the summarizing standard is the behavior characteristics which have appeared in the information type operation penalty case, wherein the characteristics refer to that the listed company performs market value management, the listed company manages the market to perform deduction and cash register at the high position of stock price, the risk of the pledge stock touching the flat warehouse line is relieved, the current situation of poor operation of the listed company is improved, and the listed company, the related institutions and the related personnel determine the operation;
the control means mainly utilizes the information advantages, namely the content and the rhythm of the favorable message release are controlled, and some private raising institutions, cattle raising companies, investment companies, trust authorities and the like enter the market to cooperate, and the capital advantages are utilized to carry out frequent buying and selling and the like;
the achievement of the information type manipulation is embodied on transaction data, such as stock price rise, sudden rise and large rise of transaction amount, frequent occurrence of bulk transactions and the like;
the multi-dimensional monitoring index construction starts from an operation cause, an operation behavior and an operation effect, wherein the multi-dimensional monitoring index construction specifically refers to the following information in 8 aspects:
market information, which refers to percentage of change of stock price and percentage of change of transaction amount;
the block transaction means the number of the block transaction, the average transaction amount of the block transaction, and the ratio of the transaction number of the business department with the maximum block transaction number to the block transaction number of all the business departments so as to reflect the concentration degree of the transaction of the business departments;
the shareholder reduction refers to the share reduction rate (%), the share reduction rate (%) and the number of times of shareholder reduction;
the capital investment of the stakeholder means the percentage (%) of the capital investment, the percentage (%) of each capital investment, the number of times of the capital investment, the percentage of change in the price of each non-decompensated capital from the date of the capital investment (%)
Market information, the number of news items of interest information of each listed company, and the ratio of the number of positive comments of the netizens under news related to the interest information of each listed company to the number of all comments;
the investment abnormal movement means that the number of the stockholders (private recruitment, cattle release, investment companies and trust institutions) of special types of newly entered stocks is actually controlled by the control of the stock ratio of people and the control of the stock ratio of institution investors;
financial reports are abnormal, which refers to important financial indicators, such as net profit, net assets, cash flow generated by business activities, size of accounts receivable and number of consecutive negative quarters;
favorable bulletins, namely the number of favorable bulletins, the average interval time (day) for releasing favorable bulletins, the time interval (day) between the releasing time of favorable bulletins and the actual occurrence time of events in the bulletins, whether to release and purchase a recombination class plan and terminate due to subjective reasons of companies (binary variable is 1, whether to be 0), whether to release a high forwarding plan under the condition that the unallocated profit is negative (binary variable is 1, whether to be 0), and whether to release bulletins related to hot concepts (binary variable is 1, whether to be 0);
the establishment of the primary early warning model comprises the following steps: collecting information-type controlled stocks with definite penalties judged by the guild and stocks without penalties judged by the guild in the same industry to form a training data set;
the market information dimension comprises a stock price change percentage, a transaction quantity change percentage, a stock sale prohibition and forestock price change rate, and a stock pledge explosion and announcement forestock price change rate;
step four, a first-level early warning model based on the integrated discrimination model
Collecting information type manipulated stocks with definite penalties and stocks without penalties judged by the certificated guild to form a training data set, regarding each trading day of each stock as a sample point, regarding samples in the stock manipulation period with definite penalties as positive samples, and regarding other samples as negative samples;
comparing the time dimension and the space dimension of the multi-dimensional monitoring index to form a discrimination dimension, and integrating a discrimination model and a training model based on Xgboost;
combining a grid search method and a five-fold cross inspection method to determine key parameters in the Xgboost model, wherein the key parameters comprise: the method comprises the steps of learning rate (the parameter can improve the robustness of a model by reducing the weight of each step), random sampling parameters (the parameter is used for controlling the proportion of random sampling, when the parameter value is reduced, the algorithm is more conservative so as to avoid overfitting), regularization parameters (the parameter is a coefficient of a regularization item and has the function of avoiding overfitting), unbalance parameters (when various types of samples are quite unbalanced, the parameter is set to be a positive value so that the algorithm can be converged more quickly), selecting the model with the optimal precision performance as an accurate key parameter, applying the trained model to stocks in a whole market sample period to obtain the probability of implementing information type operation on each trading day of each stock, and if the calculated probability is more than 0.5, judging that the information type operation is implemented and outputting a primary early warning result;
the grid search method is an exhaustion method, namely various parameter combinations are set, an F index of comprehensive accuracy and recall rate is calculated by using a five-fold cross-validation method, the parameter combination with the highest F value is selected, and a model is trained for subsequent calculation;
through the above steps, the time series of whether each trading day of each stock is engaged in information-based manipulation can be obtained.
If 3 or more trading days appear in the past 5 trading days as the information type operation, the trading day is judged as the starting date of the information type operation, and if the trading days in the future 20 trading days do not appear as the information type operation any more, the trading day is judged as the ending date of the information type operation. The result of the first-level warning, namely the stock suspected to be engaged in the information type manipulation, the manipulation start date and the manipulation end date is obtained.
Step five, sub-early warning model based on core index abnormity
The stock pledge rate in the manipulation period is at the level of 10% before the whole market, the guarantee rate is at the level of 10% before the whole market, the bulk trade is frequent, namely the bulk trade times during the manipulation start date and the manipulation end date are at the level of 10% before the same industry, the concentrated interest bulletins are issued, namely the number of the interest bulletins during the manipulation start date and the manipulation end date is at the level of 10% before the same industry, the actual stock holding proportion of a controller is over 50%, and the stock price sharply rises after the suggestive bulletin which the pledge touches the flat warehouse line and the stock price sharply rises after the suggestive bulletin in which the stock limit is forbidden, so as to form a sub-early warning result.
The sub-early warning results are of the same level, the sub-early warning results comprise stocks suspected of being engaged in information type manipulation, manipulation starting date and manipulation ending date, and the sub-early warning results are higher in precision compared with the first level early warning results.
In another embodiment, the method for analyzing the association based on the relationship between the listed companies and the shareholders includes at least registered capital, company name, and listing time, and the shareholders includes at least shareholders type and holding stock list. The technical scheme provides the preferred registration information and stockholder information.
In another technical scheme, the association analysis method based on the connection between the listed company and the shareholder stores the registration information and the shareholder information into the stock holding relation database, and updates the stock holding relation database at intervals of set time periods. The technical scheme updates data in real time and facilitates real-time supervision.
In another technical solution, the method for performing early warning on a stock code based on a correlation analysis method of a connection between a listed company and a shareholder includes:
acquiring a trading sample, wherein the trading sample comprises trading behaviors of stock codes at all times and stock market conditions at corresponding times;
dividing a plurality of time intervals, extracting transaction characteristics from transaction behaviors in each time interval, and extracting stock market characteristics from stock market conditions in each time interval;
taking the transaction characteristics and the stock market characteristics in the time interval as input, taking the transaction characteristics in the following time interval as output, and training to obtain a neural network prediction model;
and acquiring the trading behavior, trading characteristics, stock market conditions and stock market characteristics of the target stock code in a time interval, inputting the trading behavior, trading characteristics, stock market conditions and stock market characteristics into a neural network prediction model, and if the difference between the output trading characteristics of the next time interval and the actual trading characteristics of the next time interval is larger than a third threshold, early warning the target stock code.
The technical scheme provides another early warning technical scheme, stock trading has continuity, and two adjacent trading time intervals have correlation. And training by taking the transaction characteristics and the stock market characteristics in the previous time period as input and taking the transaction characteristics in the next time period as output to establish a neural network prediction model. In the actual monitoring stage, the trading behavior of the target stock code in a time interval is collected, the trading characteristics are extracted and input into the neural network prediction model to obtain the trading characteristics of the next time interval, and whether the target stock code needs to be warned or not is judged through the difference by comparing the trading characteristics with the actual trading characteristics of the next time interval. The transaction characteristics comprise transaction times, transaction amount, transaction objects and the like, and the stock market characteristics comprise stock price trend, transaction amount, closing price, opening price and the like. The third threshold may be, for example, 10%, and when the difference reaches more than 10%, the target stock code is judged to be suspicious, i.e., an early warning is performed.
The invention also provides a correlation analysis device based on the relation between the listed companies and the stockholders, which comprises the following steps:
a processor;
a memory storing executable instructions;
wherein the processor is configured to execute the executable instructions to perform the association analysis method based on the association between the listed company and the shareholder.
The technical scheme is obtained based on the same inventive concept as the association analysis method based on the connection between the listed companies and the stockholders, and reference can be made to the description of the method part. The device of the technical scheme is not limited to a PC, a terminal and a server. The device can be arranged in a server, and data acquisition and processing are carried out at set time intervals.
The number of apparatuses and the scale of the process described herein are intended to simplify the description of the present invention. Applications, modifications and variations of the present invention based on the association analysis method and apparatus of the relationship between the listed companies and the stakeholders will be apparent to those skilled in the art.
While embodiments of the invention have been described above, it is not limited to the applications set forth in the description and the embodiments, which are fully applicable in various fields of endeavor to which the invention pertains, and further modifications may readily be made by those skilled in the art, it being understood that the invention is not limited to the details shown and described herein without departing from the general concept defined by the appended claims and their equivalents.
Claims (8)
1. An association analysis method based on a connection between a listed company and a stockholder, comprising:
acquiring registration information and stockholder information of a listed company, and establishing a related network of the listed company and the stockholder;
searching each shareholder holding the target stock code according to the target stock code, further searching other listed companies and stock codes thereof to which each shareholder belongs, and establishing a first sub-association network;
and searching each listed company which holds the shareholder according to the target shareholder, further searching other shareholders of each listed company, and establishing a second sub-association network.
2. The association analysis method based on the relation between the listed company and the shareholder as claimed in claim 1, wherein the stock codes whose number of the historical early warning records reaches the first threshold value are determined in the first sub-association network and marked as early warning stock codes.
3. The method of claim 1, wherein shareholders with early warning stock codes reaching a second threshold are identified in a second sub-association network and marked as suspected collaborators.
4. The method for analyzing association based on the relation between listed companies and stockholders as claimed in claim 2, wherein the method for early warning the stock code comprises: the method comprises the steps of obtaining multi-channel heterogeneous data, classifying and storing the obtained heterogeneous data according to structured data and unstructured data to build an information perception and monitoring database, building a multi-dimensional monitoring index, building an integrated discrimination model with the multi-dimensional monitoring index as a characteristic, carrying out real-time monitoring on data in the information perception and monitoring database by adopting the discrimination model, outputting a primary early warning result, and screening and forming a sub early warning result based on a core index on the basis of the primary early warning result.
5. The method of claim 1, wherein the registration information includes at least registered capital, company name, time to market, and the shareholder information includes at least shareholder type, holding stock list.
6. The association analysis method based on the relation between the listed company and the stockholder as claimed in claim 1, wherein the registration information and the stockholder information are stored in the stock holding relationship database, and the stock holding relationship database is updated at set time intervals.
7. The method for analyzing association based on the relation between listed companies and stockholders as claimed in claim 2, wherein the method for early warning the stock code comprises:
acquiring a trading sample, wherein the trading sample comprises trading behaviors of stock codes at all times and stock market conditions at corresponding times;
dividing a plurality of time intervals, extracting transaction characteristics from transaction behaviors in each time interval, and extracting stock market characteristics from stock market conditions in each time interval;
taking the transaction characteristics and the stock market characteristics in the time interval as input, taking the transaction characteristics in the following time interval as output, and training to obtain a neural network prediction model;
and acquiring the trading behavior, trading characteristics, stock market conditions and stock market characteristics of the target stock code in a time interval, inputting the trading behavior, trading characteristics, stock market conditions and stock market characteristics into a neural network prediction model, and if the difference between the output trading characteristics of the next time interval and the actual trading characteristics of the next time interval is larger than a third threshold, early warning the target stock code.
8. An association analysis apparatus based on a relation between a listed company and a stockholder, comprising:
a processor;
a memory storing executable instructions;
wherein the processor is configured to execute the executable instructions to perform the method for association analysis based on a connection between a listed company and a shareholder as claimed in any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010252323.5A CN111583033A (en) | 2020-04-01 | 2020-04-01 | Association analysis method and device based on relation between listed company and stockholder |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010252323.5A CN111583033A (en) | 2020-04-01 | 2020-04-01 | Association analysis method and device based on relation between listed company and stockholder |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111583033A true CN111583033A (en) | 2020-08-25 |
Family
ID=72117081
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010252323.5A Pending CN111583033A (en) | 2020-04-01 | 2020-04-01 | Association analysis method and device based on relation between listed company and stockholder |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111583033A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112102087A (en) * | 2020-09-21 | 2020-12-18 | 中国工商银行股份有限公司 | Transaction abnormity detection method and device |
| CN112598513A (en) * | 2020-12-07 | 2021-04-02 | 深圳价值在线信息科技股份有限公司 | Method and device for identifying shareholder risk transaction behavior |
| CN112612937A (en) * | 2020-12-07 | 2021-04-06 | 深圳价值在线信息科技股份有限公司 | Associated information acquisition method and equipment |
| CN114461867A (en) * | 2022-04-12 | 2022-05-10 | 深圳格隆汇信息科技有限公司 | Financial big data analysis processing method and system |
| CN116542774A (en) * | 2023-06-27 | 2023-08-04 | 深圳市迪博企业风险管理技术有限公司 | Probability diffusion model-based method for detecting compliance of company-associated transactions on sale |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106780024A (en) * | 2016-11-30 | 2017-05-31 | 东软集团股份有限公司 | Transaction method for monitoring abnormality and device |
| CN106991506A (en) * | 2017-05-16 | 2017-07-28 | 深圳先进技术研究院 | Intelligent terminal and its stock trend forecasting method based on LSTM |
| CN107358527A (en) * | 2017-06-19 | 2017-11-17 | 重庆誉存大数据科技有限公司 | A kind of investment relation network visualization analysis method and system |
| CN108596439A (en) * | 2018-03-29 | 2018-09-28 | 北京中兴通网络科技股份有限公司 | A kind of the business risk prediction technique and system of knowledge based collection of illustrative plates |
| CN109189867A (en) * | 2018-10-23 | 2019-01-11 | 中山大学 | Relationship discovery method, apparatus and storage medium based on Corporate Intellectual map |
| CN109360097A (en) * | 2018-09-28 | 2019-02-19 | 中山大学 | Prediction of Stock Index method, apparatus, equipment and storage medium based on deep learning |
| CN109767322A (en) * | 2018-12-20 | 2019-05-17 | 平安科技（深圳）有限公司 | Suspicious transaction analysis method, apparatus and computer equipment based on big data |
| CN110659996A (en) * | 2019-10-25 | 2020-01-07 | 重庆第二师范学院 | Stock investment risk early warning system and method based on machine learning |
| CN110738570A (en) * | 2019-10-14 | 2020-01-31 | 北京智信度科技有限公司 | Information type manipulation automatic identification method based on multi-channel heterogeneous data |
-
2020
- 2020-04-01 CN CN202010252323.5A patent/CN111583033A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106780024A (en) * | 2016-11-30 | 2017-05-31 | 东软集团股份有限公司 | Transaction method for monitoring abnormality and device |
| CN106991506A (en) * | 2017-05-16 | 2017-07-28 | 深圳先进技术研究院 | Intelligent terminal and its stock trend forecasting method based on LSTM |
| CN107358527A (en) * | 2017-06-19 | 2017-11-17 | 重庆誉存大数据科技有限公司 | A kind of investment relation network visualization analysis method and system |
| CN108596439A (en) * | 2018-03-29 | 2018-09-28 | 北京中兴通网络科技股份有限公司 | A kind of the business risk prediction technique and system of knowledge based collection of illustrative plates |
| CN109360097A (en) * | 2018-09-28 | 2019-02-19 | 中山大学 | Prediction of Stock Index method, apparatus, equipment and storage medium based on deep learning |
| CN109189867A (en) * | 2018-10-23 | 2019-01-11 | 中山大学 | Relationship discovery method, apparatus and storage medium based on Corporate Intellectual map |
| CN109767322A (en) * | 2018-12-20 | 2019-05-17 | 平安科技（深圳）有限公司 | Suspicious transaction analysis method, apparatus and computer equipment based on big data |
| CN110738570A (en) * | 2019-10-14 | 2020-01-31 | 北京智信度科技有限公司 | Information type manipulation automatic identification method based on multi-channel heterogeneous data |
| CN110659996A (en) * | 2019-10-25 | 2020-01-07 | 重庆第二师范学院 | Stock investment risk early warning system and method based on machine learning |
Non-Patent Citations (1)
| Title |
|---|
| 伍楠林著: "《中国金融市场风险预警研究》", 31 December 2012, 中国经济出版社 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112102087A (en) * | 2020-09-21 | 2020-12-18 | 中国工商银行股份有限公司 | Transaction abnormity detection method and device |
| CN112598513A (en) * | 2020-12-07 | 2021-04-02 | 深圳价值在线信息科技股份有限公司 | Method and device for identifying shareholder risk transaction behavior |
| CN112612937A (en) * | 2020-12-07 | 2021-04-06 | 深圳价值在线信息科技股份有限公司 | Associated information acquisition method and equipment |
| CN112598513B (en) * | 2020-12-07 | 2024-04-05 | 深圳价值在线信息科技股份有限公司 | Method and device for identifying stockholder risk transaction behaviors |
| CN114461867A (en) * | 2022-04-12 | 2022-05-10 | 深圳格隆汇信息科技有限公司 | Financial big data analysis processing method and system |
| CN116542774A (en) * | 2023-06-27 | 2023-08-04 | 深圳市迪博企业风险管理技术有限公司 | Probability diffusion model-based method for detecting compliance of company-associated transactions on sale |
| CN116542774B (en) * | 2023-06-27 | 2023-12-22 | 深圳市迪博企业风险管理技术有限公司 | Probability diffusion model-based method for detecting compliance of company-associated transactions on sale |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Dahooie et al. | A novel dynamic credit risk evaluation method using data envelopment analysis with common weights and combination of multi-attribute decision-making methods | |
| CN111583033A (en) | Association analysis method and device based on relation between listed company and stockholder | |
| Teti et al. | The relationship between twitter and stock prices. Evidence from the US technology industry | |
| US20140081652A1 (en) | Automated Healthcare Risk Management System Utilizing Real-time Predictive Models, Risk Adjusted Provider Cost Index, Edit Analytics, Strategy Management, Managed Learning Environment, Contact Management, Forensic GUI, Case Management And Reporting System For Preventing And Detecting Healthcare Fraud, Abuse, Waste And Errors | |
| Campbell et al. | Tone at the bottom: Measuring corporate misconduct risk from the text of employee reviews | |
| CN111402064A (en) | Risk value evaluation method and device | |
| CN112102076B (en) | Comprehensive risk early warning system of platform | |
| Gupta et al. | Audit selection strategy for improving tax compliance: application of data mining techniques | |
| CN112116464B (en) | Abnormal transaction behavior analysis method and system based on event sequence frequent item set | |
| Omidi et al. | The efficacy of predictive methods in financial statement fraud | |
| Plaksiy et al. | Applying big data technologies to detect cases of money laundering and counter financing of terrorism | |
| Gupta et al. | A data mining framework for prevention and detection of financial statement fraud | |
| Chen | An effective going concern prediction model for the sustainability of enterprises and capital market development | |
| CN113609193A (en) | Method and device for training prediction model for predicting customer transaction behavior | |
| CN111882420A (en) | Generation method of response rate, marketing method, model training method and device | |
| Chimonaki et al. | Identification of financial statement fraud in Greece by using computational intelligence techniques | |
| Perko | Behaviour-based short-term invoice probability of default evaluation | |
| Marais et al. | Predicting financial statement manipulation in South Africa: A comparison of the Beneish and Dechow models | |
| Gupta et al. | Feature selection for dimension reduction of financial data for detection of financial statement frauds in context to Indian companies | |
| Zaki et al. | “Stock‐touting” through spam e‐mails: a data mining case study | |
| KR102499182B1 (en) | Loan regular auditing system using artificia intellicence | |
| Murugan | Creation of a recommendation system to recommend cryptocurrency portfolio using Association rule mining | |
| CN110738570A (en) | Information type manipulation automatic identification method based on multi-channel heterogeneous data | |
| Koffi et al. | Trade impacts of the New Silk Road in Africa: Insight from Neural Networks Analysis | |
| Groşanu et al. | INSTRUMENTAL CONTEXT OF A FORENSIC ACCOUNTING INVESTIGATION: A SYSTEMATIC REVIEW OF THE CURRENT LITERATURE. |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200825 |
|
| RJ01 | Rejection of invention patent application after publication |