Artificial intelligence financial data acquisition system and use method thereof
The AI financial data collection system addresses inaccuracies in stock price predictions by integrating event and news information, improving estimation accuracy by incorporating sentiment scores and event severity into the calculation.
Patent Information
- Application Number
- CN202411171813.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-26
- Publication Date
- 2024-12-03
AI Technical Summary
Existing AI financial data collection systems for stock price prediction rely solely on historical data, failing to account for the impact of events and news information, leading to inaccuracies in stock price estimates.
An AI financial data collection system that integrates news and event information to generate sentiment scores and event severity, calculating an estimated rate of return based on these factors, and incorporating them into the stock price estimation formula to improve accuracy.
Enhances the accuracy of stock price predictions by considering the impact of events and news, reducing errors between estimated and actual stock prices.
Smart Images

Figure CN119067778A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of financial technology, and in particular to an artificial intelligence financial data collection system and a method for using the same. Background Art
[0002] The AI financial data collection system is a system that integrates multiple technologies and tools, and is designed to automatically collect, organize and process financial market data. The AI financial data collection system can efficiently acquire, process and analyze financial market data to provide support for investment decisions, risk management and market forecasting.
[0003] At present, the technical means of stock price prediction by artificial intelligence financial data collection systems mainly rely on quantitative models to make periodic predictions on stock prices. Quantitative models are based on mathematical and statistical models, such as factor models, to analyze market data and make predictions. However, these models only rely on historical data to predict stock prices. They cannot take into account the impact of events and news information related to events on stock price estimates, which leads to a large error between the results of the stock price estimate and the actual stock price of the next day, reducing the accuracy of the estimate and failing to effectively deal with the impact of events on stock prices. Summary of the invention
[0004] 1. Technical issues to be resolved
[0005] In view of the deficiencies in the prior art, the present invention provides an artificial intelligence financial data collection system and a method for using the same, which collects news information and event information, generates a sentiment score Qg and an event severity Sy based on the collected news information and event information, generates a variable Bh based on the sentiment score Qg and the event severity, and calculates an estimated rate of return based on the variable Bh, thereby taking the events that have occurred and the news information related to the events as factors affecting the stock price estimate, and bringing them into the calculation formula for calculating the estimated stock price ZCg of the next day, so as to obtain a more accurate estimated value, avoid the failure to consider the impact of the events that have occurred and the news information related to the events on the stock price estimate, resulting in a large error between the result of the stock price estimate and the actual stock price of the next day, thereby improving the accuracy of the estimate, and can also effectively deal with the impact of the events on the stock price, etc., thereby solving the above-mentioned problems.
[0006] (II) Technical solution
[0007] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: an artificial intelligence financial data collection system, comprising a financial data collection module, a financial data analysis module and a risk assessment module;
[0008] The financial data collection module is connected to the cloud database, and the financial data collection module is composed of a stock price acquisition unit and an event data collection unit;
[0009] The stock price acquisition unit is used to acquire the closing price within five days from the cloud database, the event data acquisition unit is used to collect news information and event information, and the financial data acquisition module sends the collected closing price within five days, collected news information and event information to the financial data analysis module;
[0010] The financial data analysis module forms a five-day stock price data set with the closing prices within five days, calculates the daily rate of return Rs according to the data in the five-day stock price data set, and calculates the five-day stock price historical volatility LB according to the daily rate of return Rs, and after calculating the five-day stock price historical volatility LB, calculates the preliminary estimated next day stock price Cg according to the five-day stock price historical volatility LB, the financial data analysis module forms a sentiment score Qg according to the news information, the financial analysis module forms the event severity Sy according to the event information, the financial data analysis module generates a variable Bh according to the sentiment score Qg and the event severity Sy, the financial data calculates the estimated rate of return Yg according to the variable Bh, and calculates the estimated next day stock price ZCg according to the calculated estimated rate of return Yg and the preliminary estimated next day stock price Cg;
[0011] The financial data analysis module sends the calculated estimated next day stock price ZCg to the risk assessment module;
[0012] The risk assessment module compares the calculated estimated next day stock price ZCg with the stock price threshold GJyz to determine whether the estimated next day stock price ZCg is abnormal. When an abnormality occurs, the risk assessment module issues an early warning.
[0013] Preferably, the five-day stock price data set is expressed as: n-4 , S n-3 ,...,S n}, where the subscripts n-4, n-3, ..., n represent the time point information of the closing prices in the data set.
[0014] Preferably, the daily rate of return Rs is calculated as follows:
[0015]
[0016] In the formula, S d Indicates the closing price of the day, S q represents the closing price of the previous day, and S d and S q are two consecutive values in the five-day stock price data set, and ln represents the natural logarithm.
[0017] Preferably, the five-day stock price historical volatility LB calculation expression is as follows:
[0018]
[0019] In the formula, 4 means that the data in the five-day stock price data set can be used to calculate 4 daily yield Rs data. It means to find the average value of the summed results. i=1 means to start the calculation from the first day's yield, and 4 also means to sum up the calculation until the fourth day's yield.
[0020] Preferably, the calculation expression of the preliminary estimated stock price Cg for the next day is as follows:
[0021] Cg=S n *e (LB*Rs*∈)
[0022] In the formula, e represents the base of natural logarithm, ∈ represents a random variable with a value range of -3 to 3, which is used to represent the randomness of the market.
[0023] Preferably, the sentiment score Qg is expressed as: [-1, 0, +1], wherein -1 represents negative news information, 0 represents neutral news information, and +1 represents positive news information.
[0024] Preferably, the event severity Sy is expressed as: [-1 to +1], wherein -1 indicates that the severity of the event is low, i.e., a positive or active event, +1 indicates that the severity of the event is high, i.e., a negative or passive event, and when it is 0, it indicates that the severity of the event is medium, i.e., an event with no positive or negative impact.
[0025] Preferably, the variable Bh is calculated as follows:
[0026] Bh=Qg*Sy
[0027] In the formula, Qg*Sy represents the importance of the event itself and the emotional score of people towards the news information about the event, that is, the total result of the event impact obtained by emotional response.
[0028] Preferably, the estimated rate of return Yg is calculated as follows:
[0029] Yg=α+β*Rs+γ*Bh+ε
[0030] In the formula, α represents the intercept term, β represents the market sensitivity coefficient, γ represents the event impact coefficient, and ε represents the error term. α, β, and γ are obtained by the least squares method in the linear regression model, and ε is the residual in the linear regression model.
[0031] The calculation expression of the estimated next day's stock price ZCg is as follows:
[0032] ZCg=(1+Yg)*Cg
[0033] In the formula, 1 represents the current stock price, 1+Yg represents the current stock price and the estimated rate of return Yg combined to form a growth factor, which means that the value of the stock price will remain at the closing price of the day on the second day, and will increase or decrease based on the estimated rate of return Yg, and multiply it by the preliminary estimated stock price Cg for the next day, which is the estimated stock price for the next day ZCg.
[0034] A method for using an artificial intelligence financial data collection system, based on any one of the artificial intelligence financial data collection systems, comprises the following steps:
[0035] S1. Use the stock price acquisition unit in the financial data acquisition module to obtain the closing price within five days from the cloud database, use the event data acquisition unit in the financial data acquisition module to collect news information and event information, and send the collected information and data to the financial data analysis module;
[0036] S2. The financial data analysis module forms a five-day stock price data set with the closing prices within five days, and calculates the daily rate of return Rs based on the data in the five-day stock price data set. After obtaining the daily rate of return Rs, the five-day stock price historical volatility LB is calculated based on the daily rate of return Rs, and the preliminary estimate of the next day's stock price Cg is calculated based on the five-day stock price historical volatility LB;
[0037] S3, the financial data analysis module forms a sentiment score Qg and an event severity Sy according to the news information and the event information, calculates a variable Bh according to the sentiment score Qg and the event severity Sy, and calculates an estimated rate of return Yg according to the variable Bh;
[0038] S4, the financial data analysis module calculates the estimated next day stock price ZCg based on the estimated rate of return Yg and the preliminary estimated next day stock price Cg;
[0039] S5. The financial data analysis module sends the calculated estimated next day stock price ZCg to the risk assessment module. The risk assessment module compares it with the preset stock price threshold GJyz. When the estimated next day stock price ZCg is less than the stock price threshold GJyz, it is judged to be abnormal and the estimated stock price is lower than expected. The risk assessment module issues a warning to the user.
[0040] Compared with the prior art, the present invention provides an artificial intelligence financial data collection system and a method for using the same, which has the following beneficial effects:
[0041] The present invention collects news information and event information, generates a sentiment score Qg and an event severity Sy according to the collected news information and event information, generates a variable Bh according to the sentiment score Qg and the event severity, and calculates an estimated rate of return according to the variable Bh, thereby taking the events that have occurred and the news information related to the events as factors affecting the stock price estimate, and bringing them into the calculation formula for calculating the estimated stock price ZCg of the next day, so as to obtain a more accurate estimated value, avoid the failure to consider the impact of the events that have occurred and the news information related to the events on the stock price estimate, resulting in a large error between the result of the stock price estimate and the actual stock price of the next day, thereby improving the accuracy of the estimate, and can also effectively deal with the impact of the events on the stock price. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] Figure 1 It is a schematic diagram of the system flow of the present invention;
[0043] Figure 2 It is a schematic diagram of the steps of the method of the present invention. DETAILED DESCRIPTION
[0044] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0045] See also Figure 1 , an artificial intelligence financial data collection system, including a financial data collection module, a financial data analysis module and a risk assessment module;
[0046] The financial data collection module is connected to the cloud database, and can obtain the timely real-time data required in the evaluation process from the cloud database.
[0047] The financial data collection module consists of a stock price acquisition unit and an event data collection unit. The stock price acquisition unit is used to obtain the closing price within five days from the cloud database, and the event data collection unit is used to collect news information and event information.
[0048] When these data are collected, the financial data collection module sends the collected data and information to the financial data analysis module connected to it;
[0049] The financial data analysis module converts the closing prices within five days into a five-day stock price data set, which is expressed as: n-4 , S n-3 ,...,S n}, where the subscripts n-4, n-3, ..., n represent the time point information of the closing price collection in the data set. The financial data analysis module calculates the daily rate of return RS based on the data in the five-day stock price data set. The calculation expression is as follows:
[0050]
[0051] In the formula, S d Indicates the closing price of the day, S q represents the closing price of the previous day, and S d and S q are two consecutive values in the five-day stock price data set, and ln represents the natural logarithm.
[0052] After the daily rate of return Rs is calculated, the financial data analysis module calculates the five-day stock price historical volatility LB based on the daily rate of return Rs. The calculation expression is as follows:
[0053]
[0054] In the formula, 4 means that the data in the five-day stock price data set can be used to calculate 4 daily yield Rs data. It means to find the average value of the summed results. i=1 means to start the calculation from the first day's yield, and 4 also means to sum up the calculation until the fourth day's yield.
[0055] After the five-day stock price historical volatility LB is calculated, the financial data analysis module calculates the preliminary estimate of the next day's stock price Cg based on the five-day stock price historical volatility LB. The calculation expression is as follows:
[0056] Cg=S n *e (LB*Rs*∈)
[0057] In the formula, e represents the base of natural logarithm, ∈ represents a random variable with a value range of -3 to 3, which is used to represent the randomness of the market.
[0058] The financial data analysis module constructs a sentiment score Qg based on the news information, where the sentiment score Qg is expressed as: [-1, 0, +1], where -1 represents negative news information, 0 represents neutral news information, and +1 represents positive news information.
[0059] The financial analysis module constructs the event severity Sy based on the event information, where the event severity Sy is expressed as: [-1~+1], where -1 indicates that the severity of the event is low, that is, a positive or active event, +1 indicates that the severity of the event is high, that is, a negative or passive event, and when it is 0, it means that the severity of the event is medium, that is, an event with no positive or negative impact.
[0060] The financial data analysis module generates the variable Bh based on the sentiment score Qg and the event severity Sy. Its calculation expression is as follows:
[0061] Bh=Qg*Sy
[0062] In the formula, Qg*Sy represents the importance of the event itself and the emotional score of people towards the news information about the event, that is, the sum of the impact of the event obtained through emotional response.
[0063] After the variable Bh is calculated, the financial data calculates the estimated rate of return Yg based on the variable Bh. The calculation expression is as follows:
[0064] Yg=α+β*Rs+γ*Bh+ε
[0065] In the formula, α represents the intercept term, β represents the market sensitivity coefficient, γ represents the event impact coefficient, and ε represents the error term. α, β, and γ are obtained through the least squares method in the linear regression model, and ε is the residual in the linear regression model. The data used in the linear regression model is the corresponding historical data in the cloud data, and the quantity is several.
[0066] After calculating the estimated rate of return Yg, the financial data analysis module calculates the estimated next day stock price ZCg based on the estimated rate of return Yg and the preliminary estimated next day stock price Cg. The specific expression is as follows:
[0067] ZCg=(1+Yg)*Cg
[0068] In the formula, 1 represents the current stock price, 1+Yg represents the current stock price and the estimated rate of return Yg combined to form a growth factor, which means that the value of the stock price will remain at the closing price of the day on the second day, and will increase or decrease based on the estimated rate of return Yg, and multiply it by the preliminary estimated stock price Cg for the next day, which is the estimated stock price for the next day ZCg.
[0069] After the estimated next day's stock price ZCg is calculated, the financial data analysis module sends it to the risk assessment module connected to it.
[0070] The risk assessment module is preset with a stock price threshold GJyz. The stock price threshold is the unit price per share acceptable to the investment owner set by the monitoring personnel. The risk assessment module compares the estimated next day's stock price ZCg calculated with the stock price threshold GJyz. When the estimated next day's stock price ZCg is greater than the stock price threshold GJyz, the estimated next day's stock price ZCg is determined to be an abnormal value and an abnormality is determined. When an abnormality occurs, the risk assessment module issues an early warning to remind the monitoring personnel that they should focus on the price of the stock when the market opens the next day to avoid economic losses.
[0071] By collecting news information and event information, and generating a sentiment score Qg and event severity Sy based on the collected news information and event information, and generating a variable Bh based on the sentiment score Qg and event severity, and calculating the estimated rate of return based on the variable Bh, the events that occurred and the news information related to the events are used as factors affecting the stock price estimate, and are brought into the calculation formula for calculating the estimated stock price ZCg of the next day, so as to obtain a more accurate estimate value, avoiding the failure to consider the impact of the events that occurred and the news information related to the events on the stock price estimate, resulting in a large error between the result of the stock price estimate and the actual stock price of the next day, thereby improving the accuracy of the estimate, and can also effectively deal with the impact of the event on the stock price.
[0072] See also Figure 2 , the practical method steps of the above system are as follows:
[0073] S1. Use the stock price acquisition unit in the financial data acquisition module to obtain the closing price within five days from the cloud database, use the event data acquisition unit in the financial data acquisition module to collect news information and event information, and send the collected information and data to the financial data analysis module;
[0074] S2. The financial data analysis module forms a five-day stock price data set with the closing prices within five days, and calculates the daily rate of return Rs based on the data in the five-day stock price data set. After obtaining the daily rate of return Rs, the five-day stock price historical volatility LB is calculated based on the daily rate of return Rs, and the preliminary estimate of the next day's stock price Cg is calculated based on the five-day stock price historical volatility LB;
[0075] S3, the financial data analysis module forms a sentiment score Qg and an event severity Sy according to the news information and the event information, calculates a variable Bh according to the sentiment score Qg and the event severity Sy, and calculates an estimated rate of return Yg according to the variable Bh;
[0076] S4, the financial data analysis module calculates the estimated next day stock price ZCg based on the estimated rate of return Yg and the preliminary estimated next day stock price Cg;
[0077] S5. The financial data analysis module sends the calculated estimated next day stock price ZCg to the risk assessment module. The risk assessment module compares it with the preset stock price threshold GJyz. When the estimated next day stock price ZCg is less than the stock price threshold GJyz, it is judged to be abnormal and the estimated stock price is lower than expected. The risk assessment module issues a warning to the user.
[0078] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. An artificial intelligence financial data collection system, characterized by: Including financial data collection module, financial data analysis module and risk assessment module; The financial data collection module is connected to the cloud database, and the financial data collection module is composed of a stock price acquisition unit and an event data collection unit; The stock price acquisition unit is used to acquire the closing price within five days from the cloud database, the event data acquisition unit is used to collect news information and event information, and the financial data acquisition module sends the collected closing price within five days, collected news information and event information to the financial data analysis module; The financial data analysis module forms a five-day stock price data set with the closing prices within five days, calculates the daily rate of return Rs according to the data in the five-day stock price data set, and calculates the five-day stock price historical volatility LB according to the daily rate of return Rs, and after calculating the five-day stock price historical volatility LB, calculates the preliminary estimated next day stock price Cg according to the five-day stock price historical volatility LB, the financial data analysis module forms a sentiment score Qg according to the news information, the financial analysis module forms the event severity Sy according to the event information, the financial data analysis module generates a variable Bh according to the sentiment score Qg and the event severity Sy, the financial data calculates the estimated rate of return Yg according to the variable Bh, and calculates the estimated next day stock price ZCg according to the calculated estimated rate of return Yg and the preliminary estimated next day stock price Cg; The financial data analysis module sends the calculated estimated next day stock price ZCg to the risk assessment module; The risk assessment module compares the calculated estimated next day stock price ZCg with the stock price threshold GJyz to determine whether the estimated next day stock price ZCg is abnormal. When an abnormality occurs, the risk assessment module issues an early warning.
2. The artificial intelligence financial data collection system according to claim 1, characterized in that: The five-day stock price data set expression is: n-4 , S n-3 ,...,S n }, where the subscripts n-4, n-3, ..., n represent the time point information of the closing prices in the data set.
3. The artificial intelligence financial data collection system according to claim 2, characterized in that: The daily rate of return Rs is calculated as follows: In the formula, S d Indicates the closing price of the day, S q represents the closing price of the previous day, and S d and S q are two consecutive values in the five-day stock price data set, and ln represents the natural logarithm.
4. The artificial intelligence financial data collection system according to claim 3, characterized in that: The five-day stock price historical volatility LB calculation expression is as follows: In the formula, 4 means that the data in the five-day stock price data set can be used to calculate 4 daily yield Rs data. It means to find the average value of the summed results. i=1 means to start the calculation from the first day's yield, and 4 also means to sum up the calculation until the fourth day's yield.
5. The artificial intelligence financial data collection system according to claim 4, characterized in that: The calculation expression of the preliminary estimated stock price Cg for the next day is as follows: Cg=S n *e (LB*Rs*∈) In the formula, e represents the base of natural logarithm, ∈ represents a random variable with a value range of -3 to 3, which is used to represent the randomness of the market.
6. The artificial intelligence financial data collection system according to claim 5, characterized in that: The sentiment score Qg is expressed as: [-1, 0, +1], where -1 represents negative news information, 0 represents neutral news information, and +1 represents positive news information.
7. An artificial intelligence financial data collection system according to claim 6, characterized in that: The event severity Sy is expressed as: [-1~+1], where -1 indicates that the severity of the event is low, that is, a positive or active event, +1 indicates that the severity of the event is high, that is, a negative or passive event, and when it is 0, it means that the severity of the event is medium, that is, an event with no positive or negative impact.
8. An artificial intelligence financial data collection system according to claim 7, characterized in that: The variable Bh is calculated as follows: Bh=Qg*Sy In the formula, Qg*Sy represents the importance of the event itself and the emotional score of people towards the news information about the event, that is, the sum of the impact of the event obtained through emotional response.
9. An artificial intelligence financial data collection system according to claim 8, characterized in that: The estimated rate of return Yg is calculated as follows: Yg=α+β*Rs+γ*Bh+ε In the formula, α represents the intercept term, β represents the market sensitivity coefficient, γ represents the event impact coefficient, and ε represents the error term. α, β, and γ are obtained by the least squares method in the linear regression model, and ε is the residual in the linear regression model. The calculation expression of the estimated next day's stock price ZCg is as follows: ZCg=(1+Yg)*Cg In the formula, 1 represents the current stock price, 1+Yg represents the current stock price and the estimated rate of return Yg combined to form a growth factor, which means that the value of the stock price will remain at the closing price of the day on the second day, and will increase or decrease based on the estimated rate of return Yg, and multiply it by the preliminary estimated stock price Cg for the next day, which is the estimated stock price for the next day ZCg.
10. A method for using an artificial intelligence financial data collection system, based on an artificial intelligence financial data collection system according to any one of claims 1 to 9, characterized in that: The following steps are involved: S1. Use the stock price acquisition unit in the financial data acquisition module to obtain the closing price within five days from the cloud database, use the event data acquisition unit in the financial data acquisition module to collect news information and event information, and send the collected information and data to the financial data analysis module; S2. The financial data analysis module forms a five-day stock price data set with the closing prices within five days, and calculates the daily rate of return Rs based on the data in the five-day stock price data set. After obtaining the daily rate of return Rs, the five-day stock price historical volatility LB is calculated based on the daily rate of return Rs, and the preliminary estimate of the next day's stock price Cg is calculated based on the five-day stock price historical volatility LB; S3, the financial data analysis module forms a sentiment score Qg and an event severity Sy according to the news information and the event information, calculates a variable Bh according to the sentiment score Qg and the event severity Sy, and calculates an estimated rate of return Yg according to the variable Bh; S4, the financial data analysis module calculates the estimated next day stock price ZCg based on the estimated rate of return Yg and the preliminary estimated next day stock price Cg; S5. The financial data analysis module sends the calculated estimated next day stock price ZCg to the risk assessment module. The risk assessment module compares it with the preset stock price threshold GJyz. When the estimated next day stock price ZCg is less than the stock price threshold GJyz, it is judged to be abnormal and the estimated stock price is lower than expected. The risk assessment module issues a warning to the user.
