Transaction signal generation method and device, program product and electronic equipment

Abstract

The invention discloses a transaction signal generation method and device, a program product and electronic equipment, and relates to the field of financial science and technology, and the method comprises the steps: collecting first information and second information corresponding to a target transaction, determining a transaction risk value corresponding to the target transaction based on the first information, the second information, L first features and M second features, wherein the L first features are transaction features of historical transactions of which the income proportion is greater than or equal to a preset proportion under different dimensions, the M second features are event features of historical events of which the income proportion is related to the historical transactions under different dimensions, and when the transaction risk value is less than or equal to a preset risk value, the M second features are the event features of the historical events of which the income proportion is greater than or equal to the preset risk value. A target signal is generated. The technical problem of low transaction signal generation accuracy caused by lack of understanding of financial product information and / or transaction object information involved in user transactions in an existing financial transaction system is solved.

Description

Technical Field

[0001] The present application relates to the field of financial technology, and specifically, to a method, device, program product, and electronic device for generating a trading signal. Background Art

[0002] Existing financial trading signal generation technology is usually limited to the analysis of historical price data, ignoring the specific information of financial products and trading objects, as well as the impact of external events on the market. This one-sided analysis method leads to severe limitations on the accuracy of trading signals. Especially in rapidly changing market environments, it is often unable to reflect the latest situation in a timely manner, thereby affecting the quality of decision-making. As a result, in existing financial trading systems, due to the lack of understanding of the financial product information and / or trading object information involved in user transactions, the technical problem of low accuracy in the generation of trading signals is caused.

[0003] To address the above-mentioned problems, no effective solutions have been proposed so far. Summary of the Invention

[0004] The present application provides a method, apparatus, program product, and electronic device for generating a transaction signal to at least address the technical problem of low accuracy in generating transaction signals in existing financial transaction systems due to a lack of understanding of the financial product information and / or transaction object information involved in user transactions.

[0005] According to one aspect of the present application, a method for generating a transaction signal is provided, comprising: collecting first information and second information corresponding to a target transaction, wherein the target transaction is a financial transaction to be initiated by a target user, the first information at least including the transaction time, transaction type, financial product type and transaction object, and the second information is information of a target event in an open source platform that is relevant to the financial products and / or transaction objects involved in the target transaction; determining a transaction risk value corresponding to the target transaction based on the first information, the second information, L first features and M second features, wherein the L first features are transaction features in different dimensions of historical transactions with a return ratio greater than or equal to a preset ratio, and the M second features are event features in different dimensions of historical events that are relevant to the return ratio of historical transactions, and both L and M are positive integers; generating a target signal when the transaction risk value is less than or equal to the preset risk value, wherein the target signal is used to prompt the user to complete the target transaction.

[0006] Optionally, a target model is used to determine a transaction risk corresponding to a target transaction based on the first information, the second information, L first features, and M second features. The target model training step includes: determining a profit label corresponding to each of the P historical transactions based on transaction information of the P historical transactions, where P is a positive integer and the profit label is used to characterize the profit ratio of the corresponding historical transaction; screening media information in an open source platform based on the transaction time corresponding to each historical transaction to obtain X historical events, where X is a positive integer and the X historical events are historical events that are relevant to the financial products and / or transaction objects involved in the P historical transactions; and training an initial neural network model based on the transaction information of the P historical transactions, event information corresponding to the X historical events, and the profit label corresponding to each historical transaction to obtain a target model.

[0007] Optionally, the step of training an initial neural network model based on transaction information of P historical transactions, event information corresponding to X historical events, and a profit label corresponding to each historical transaction to obtain a target model includes: performing feature extraction on the transaction information of the P historical transactions to obtain P historical transaction features, wherein the P historical transactions correspond to the P historical transaction features in a one-to-one manner; performing feature extraction on the event information corresponding to the X historical events to obtain X historical event features, wherein the X historical events correspond to the X historical event features in a one-to-one manner; training the initial neural network model based on the P historical transaction features, the X historical event features, and the profit label corresponding to each historical transaction to obtain a first sub-model and a second sub-model, wherein the first sub-model is used to store at least L first features and M second features, and the second sub-model is used to determine a transaction risk value corresponding to the target transaction; and using the first sub-model and the second sub-model as the target model.

[0008] Optionally, in the process of training the initial neural network model based on P historical transaction features, X historical event features, and the profit label corresponding to each historical transaction, the transaction signal generation method further includes: using historical transaction features whose profit label is a target label among the P historical transaction features as positive transaction features to obtain Q positive transaction features, where Q is a positive integer less than or equal to P, and the target label is used to indicate that the profit ratio of the corresponding historical event is greater than or equal to a preset ratio; using historical event features among the X historical event features that are correlated with Q historical transactions corresponding to the Q positive transaction features as typical event features to obtain Y typical event features, where Y is less than or equal to a positive integer; determining L first features based on the Q positive transaction features, where the L first features are sub-features of each positive transaction feature; and determining M second features based on the Y typical event features, where the M second features are sub-features of each typical event feature.

[0009] Optionally, the method for generating a transaction signal also includes: extracting features of the first information through the first sub-model in the target model to obtain target transaction features corresponding to the target transaction; extracting features of the second information through the first sub-model to obtain target event features corresponding to the target event; and determining the transaction risk value based on the target transaction features, the target event features, L first features, and M second features.

[0010] Optionally, the step of determining the transaction risk value based on the target transaction feature, the target event feature, L first features and M second features includes: performing weighted summation of the similarities between the target transaction feature and each of the L first features to obtain a first similarity; performing weighted summation of the similarities between the target event feature and each of the M second features to obtain a second similarity; and determining the transaction risk value based on the first similarity and the second similarity through a second sub-model in the target model.

[0011] Optionally, after determining the transaction risk value corresponding to the target transaction based on the first information, the second information, L first features and M second features, the method for generating a transaction signal also includes: generating early warning information when the transaction risk value is greater than a preset risk value, wherein the early warning information is used to prompt the user to interrupt the target transaction.

[0012] According to another aspect of the present application, a transaction signal generation device is also provided, including: an information collection unit, used to collect first information and second information corresponding to a target transaction, wherein the target transaction is a financial transaction to be initiated by a target user, the first information includes at least the transaction time, transaction type, financial product type and transaction object, and the second information is information of a target event in the open source platform that is relevant to the financial products and / or transaction objects involved in the target transaction; a first determination unit, used to determine the transaction risk value corresponding to the target transaction based on the first information, the second information, L first features and M second features, wherein the L first features are transaction features of historical transactions with a profit ratio greater than or equal to a preset ratio in different dimensions, and the M second features are event features of historical events with a correlation with the profit ratio of historical transactions in different dimensions, and L and M are both positive integers; a first generation unit, used to generate a target signal when the transaction risk value is less than or equal to the preset risk value, wherein the target signal is used to prompt the user to complete the target transaction.

[0013] In the present application, the first information and second information corresponding to the target transaction are first collected, wherein the target transaction is a financial transaction to be initiated by the target user, the first information includes at least the transaction time, transaction type, financial product type and transaction object, and the second information is the information of the target event in the open source platform that is relevant to the financial products and / or transaction objects involved in the target transaction. Afterwards, the present application determines the transaction risk value corresponding to the target transaction based on the first information, the second information, L first features and M second features, wherein the L first features are transaction features of historical transactions with a profit ratio greater than or equal to a preset ratio in different dimensions, and the M second features are event features of historical events with a correlation with the profit ratio of historical transactions in different dimensions, and both L and M are positive integers. Finally, the present application generates a target signal when the transaction risk value is less than or equal to the preset risk value, wherein the target signal is used to prompt the user to complete the target transaction.

[0014] From the above content, it can be seen that the present application adopts a comprehensive analysis method of real-time financial transaction information and event information related to the transaction, by pre-constructing typical features (i.e., L first features and M second features), and then determining the transaction risk value corresponding to the target transaction based on the first information, the second information, the L first features and the M second features, and then using the transaction risk value as the judgment condition for whether to generate a transaction signal, thereby achieving the purpose of improving the accuracy and response speed of the transaction signal, thereby achieving the technical effect of providing users with customized, high-precision financial transaction signals, and thus solving the technical problem of low accuracy in the generation of transaction signals in existing financial transaction systems due to lack of understanding of the financial product information and / or transaction object information involved in user transactions. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0016] Figure 1 is a flowchart of an optional method for generating a trading signal according to an embodiment of the present application;

[0017] Figure 2 is a schematic diagram of an optional trading signal system according to an embodiment of the present application;

[0018] Figure 3 is a schematic diagram of an optional trading signal generating device according to an embodiment of the present application;

[0019] Figure 4 This is a structural block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0020] In order to enable those skilled in the art to better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings 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 in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this application.

[0021] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in a sequence other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0022] It should also be noted that the relevant information (including the first information and second information corresponding to the target transaction) and data (including but not limited to data for display and analysis) involved in this application are all information and data authorized by the user or fully authorized by all parties. For example, an interface is set up between this system and the relevant user or organization. Before obtaining relevant information, it is necessary to send an acquisition request to the aforementioned user or organization through the interface, and obtain the relevant information after receiving the consent information fed back by the aforementioned user or organization.

[0023] In addition, the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant information and data involved in this application comply with the relevant laws, regulations and standards of the relevant regions, and necessary confidentiality measures have been taken, and do not violate public order and good morals. In addition, this application provides corresponding operation entrances for users to choose to agree to authorization or refuse authorization. If the user chooses to refuse authorization, he / she will enter the corresponding expert decision-making process.

[0024] According to an embodiment of the present application, an embodiment of a method for generating a trading signal is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0025] This application provides a trading signal generation system (hereinafter referred to as the generation system) for executing the trading signal generation method of this application. Figure 1 is a flow chart of an optional method for generating a trading signal according to an embodiment of the present application, such as Figure 1 As shown, the method includes the following steps:

[0026] Step S101: Collect first information and second information corresponding to the target transaction, wherein the target transaction is a financial transaction to be initiated by the target user, the first information includes at least the transaction time, transaction type, financial product type, and transaction object, and the second information is information about a target event in the open source platform that is relevant to the financial product and / or transaction object involved in the target transaction.

[0027] Optionally, the target transaction refers to a financial transaction that the user plans to or is about to execute, that is, a buying or selling operation of a financial product.

[0028] Optionally, the transaction type is buy or sell, the financial product type includes at least stocks and foreign exchange, and the transaction object is the seller or buyer of the financial product.

[0029] Optionally, the second information is external information from an open source platform, which is related to the financial products or transaction objects involved in the target transaction and can reflect market events that affect the transaction results.

[0030] Optionally, step S101 provides comprehensive data support for subsequent risk assessment and transaction signal generation by collecting internal information (first information) and external environment (second information) of the target transaction in real time.

[0031] Step S102: Determine the transaction risk value corresponding to the target transaction based on the first information, the second information, L first features, and M second features, wherein the L first features are transaction features in different dimensions of historical transactions with a profit ratio greater than or equal to a preset ratio, and the M second features are event features in different dimensions of historical events that are correlated with the profit ratio of historical transactions, and both L and M are positive integers.

[0032] Optionally, the L first features represent transaction features of different dimensions displayed by historical transactions whose profit ratios meet user expectations; the M second features refer to event features associated with the transaction profit ratios within the historical time period.

[0033] Optionally, step S102 utilizes historical data and real-time information to quantify the risk of the target transaction through machine learning or statistical analysis methods. The feature set consisting of L first features and M second features enables the generation system to evaluate from multiple angles, thereby improving the accuracy of risk prediction.

[0034] Step S103: When the transaction risk value is less than or equal to the preset risk value, a target signal is generated, wherein the target signal is used to prompt the user to complete the target transaction.

[0035] Optionally, the preset risk value is a risk threshold pre-set by the generation system to distinguish which transaction risks are acceptable to the user and which are unacceptable. The generation system will set a relatively low preset risk value for users with lower risk preferences.

[0036] Optionally, step S103 ensures that the generation system will recommend users to trade only when the transaction risks are controllable. This step effectively balances profit opportunities and risk control, provides users with a trading decision-making aid based on data and model analysis, and avoids potential asset losses caused by users' blind trading.

[0037] From the above content, it can be seen that the present application adopts a comprehensive analysis method of real-time financial transaction information and event information related to the transaction, by pre-constructing typical features (i.e., L first features and M second features), and then determining the transaction risk value corresponding to the target transaction based on the first information, the second information, the L first features and the M second features, and then using the transaction risk value as the judgment condition for whether to generate a transaction signal, thereby achieving the purpose of improving the accuracy and response speed of the transaction signal, thereby achieving the technical effect of providing users with customized, high-precision financial transaction signals, and thus solving the technical problem of low accuracy in the generation of transaction signals in existing financial transaction systems due to lack of understanding of the financial product information and / or transaction object information involved in user transactions.

[0038] In an optional embodiment, the generation system determines the transaction risk corresponding to the target transaction based on the first information, the second information, the L first features, and the M second features using a target model, and the training step of the target model includes:

[0039] First, the generation system determines a revenue label corresponding to each of the P historical transactions based on transaction information of the P historical transactions, where P is a positive integer and the revenue label is used to represent the revenue ratio of the corresponding historical transaction. Then, the generation system filters media information in the open source platform based on the transaction time corresponding to each historical transaction to obtain X historical events, where X is a positive integer and the X historical events are historical events that are relevant to the financial products and / or transaction objects involved in the P historical transactions. Then, the generation system trains an initial neural network model based on the transaction information of the P historical transactions, the event information corresponding to the X historical events, and the revenue label corresponding to each historical transaction to obtain a target model.

[0040] Optionally, the P historical transactions refer to a set of past transaction records extracted from a database.

[0041] Optionally, the profit ratio is calculated based on the opening and closing prices of the transaction, for example, profit ratio = (closing price - opening price) / opening price.

[0042] Optionally, if the profit ratio is greater than or equal to a preset ratio, the historical transaction is marked as a positive profit label (ie, a target label); otherwise, it is marked as a negative profit label.

[0043] Optionally, media information in the open source platform refers to news reports, comments, policy announcements and other information related to financial products and transaction objects that are publicly available on the Internet.

[0044] Optionally, by comparing the transaction time of historical transactions with the timeline of media information, the generation system can screen out historical events that occurred before and after the transaction and had an impact on the transaction results. These historical events constitute the external environmental factors of the transaction.

[0045] Optionally, the initial neural network model is a pre-set network structure comprising multiple layers of nodes, capable of processing and analyzing complex data inputs to learn the nonlinear relationship between input data and output results.

[0046] Optionally, the generation system uses transaction information of historical transactions, related historical event information, and the profit label of each transaction as input to the neural network model. The neural network model learns this data, adjusts its internal parameters, and ultimately forms a target model that can predict transaction risks and returns based on the characteristics of new transactions and current market event information. Through iterative training, the target model has the ability to identify profit opportunities and risks in a real-time trading environment, thereby generating more accurate trading signals for users.

[0047] In an optional embodiment, the generation system first performs feature extraction on transaction information of P historical transactions to obtain P historical transaction features, where the P historical transactions correspond one-to-one to the P historical transaction features. Then, the generation system performs feature extraction on event information corresponding to X historical events to obtain X historical event features, where the X historical events correspond one-to-one to the X historical event features. Then, the generation system trains an initial neural network model based on the P historical transaction features, the X historical event features, and the profit label corresponding to each historical transaction to obtain a first sub-model and a second sub-model, where the first sub-model is used to store at least L first features and M second features, and the second sub-model is used to determine the transaction risk value corresponding to the target transaction. Finally, the generation system uses the first sub-model and the second sub-model as target models.

[0048] Optionally, the P historical transaction features are features extracted from historical transaction information that can describe transaction behavior and results, such as transaction volume, price volatility, duration, and transaction frequency.

[0049] Optionally, the generation system extracts features from the transaction information of P historical transactions, thereby converting the historical data of each historical transaction into a feature vector. These feature vectors contain key attributes of the historical transactions for use in the subsequent model training process. Through feature extraction, the system can focus on variables that are critical to the success rate of transactions, thereby improving the predictive performance of the model.

[0050] Optionally, the X historical event features are features extracted from event information that can describe the potential impact of the event on the financial market, such as the level of influence of the event, positive or negative impact, and type of event. These features are also used to enhance the model's understanding ability.

[0051] Optionally, similar to the processing of historical transaction information, the generation system is also committed to converting event information into feature vectors to facilitate recognition and understanding by the neural network model. The extraction of event features ensures that the model can not only learn based on the attributes of the transaction itself, but also take into account market fluctuations caused by external events.

[0052] Optionally, through model training, the generation system can automatically learn and identify which combinations of transaction features and event features correspond to higher probability of returns and which combinations will lead to risks. The first sub-model focuses on understanding the intrinsic connection between L first features and M second features, while the second sub-model quantifies transaction risks on this basis and provides algorithmic support for risk assessment.

[0053] Optionally, the generation system first extracts key transaction features from a large amount of historical transaction data, then extracts event features from relevant historical events, and then uses these features and profit labels to train the neural network model, ultimately obtaining a first sub-model and a second sub-model that can predict transaction risks. The entire process not only focuses on learning historical transaction behaviors, but also incorporates the impact analysis of market event dynamics, making the transaction risk value predicted by the target model more accurate, helping users make more informed trading decisions and avoid market risks.

[0054] In an optional embodiment, the generation system first uses the historical transaction features with the profit label as the target label among the P historical transaction features as positive transaction features to obtain Q positive transaction features, where Q is a positive integer less than or equal to P, and the target label is used to characterize that the profit ratio of the corresponding historical event is greater than or equal to a preset ratio. Then, the generation system uses the historical event features among the X historical event features that are correlated with the Q historical transactions corresponding to the Q positive transaction features as typical event features to obtain Y typical event features, where Y is less than or equal to a positive integer. Then, the generation system determines L first features based on the Q positive transaction features, where the L first features are sub-features of each positive transaction feature. Subsequently, the generation system determines M second features based on the Y typical event features, where the M second features are sub-features of each typical event feature.

[0055] Optionally, the Q positive transaction features are further screened from the P historical transaction features, and only those transaction features whose transaction results are marked as positive returns (target labels) are retained. The generation system focuses on successful transaction cases, that is, those transactions that have brought sufficient profits. By screening out transaction features with positive returns, the generation system is able to establish a target model that is more focused on learning positive transaction patterns, thereby improving the accuracy of signal generation.

[0056] Optionally, the Y typical event features are event features selected from the X historical event features, and are directly or indirectly correlated with the transactions corresponding to the Q positive transaction features. The generation system focuses on historical events that are closely related to the transactions, thereby being able to more accurately identify which historical event information has a positive impact on transaction results, which helps to generate future trading signals more accurately consider the impact of market dynamics and external events.

[0057] Optionally, the generation system learns from historical data to identify and predict trading opportunities. First, trading features with positive returns are screened out from a large amount of historical trading data, which is equivalent to establishing a successful trading case library. Next, typical event features are extracted from market events related to these successful transactions, further clarifying which external information is valuable. Subsequently, the generation system refines positive trading features into first features and typical event features into second features. These features constitute the core input of model training. Through such a process, the final generated model can accurately evaluate future trading opportunities based on the attributes of the transaction itself and the impact of external events, and provide effective trading signals.

[0058] In an optional embodiment, the generation system first extracts features of the first information through the first sub-model in the target model to obtain target transaction features corresponding to the target transaction. Then, the generation system extracts features of the second information through the first sub-model to obtain target event features corresponding to the target event. Then, the generation system determines the transaction risk value based on the target transaction features, target event features, L first features and M second features.

[0059] Optionally, the generation system performs an in-depth analysis of the first information of the target transaction through the first sub-model to extract features that are helpful in understanding the transaction behavior and predicting its results, thereby ensuring that the model can focus on the intrinsic attributes of the transaction and conduct more accurate risk analysis.

[0060] Optionally, through processing by the first sub-model, the generating system extracts features from the second information that can describe the impact of the target event on the financial market, namely, target event features, thereby helping the model understand the correlation between the target event and the transaction and the market fluctuations caused by the event, thereby ensuring that the target model not only considers the characteristics of the transaction itself when assessing risks, but also incorporates the impact of the external market environment, thereby enhancing the comprehensiveness and accuracy of risk assessment.

[0061] In an optional embodiment, the generation system first performs a weighted sum of the similarities between the target transaction feature and each of the L first features to obtain a first similarity. Then, the generation system performs a weighted sum of the similarities between the target event feature and each of the M second features to obtain a second similarity. Then, the generation system determines the transaction risk value based on the first similarity and the second similarity through the second sub-model in the target model.

[0062] Optionally, the generation system compares the similarity between the current transaction features and the features of past successful transactions, and considers the importance of each feature. The system can preliminarily evaluate whether the target transaction conforms to the pattern of previous successful transactions, thereby calculating a first similarity reflecting the inherent risk of the transaction.

[0063] Optionally, the generation system provides a quantitative method to evaluate whether external conditions are conducive to the current trading plan by calculating the similarity between the target event feature and the M second features, thereby enabling a comprehensive assessment of trading risks.

[0064] Optionally, the generation system controls the second sub-model to take the first similarity and the second similarity as input, and comprehensively analyzes the impact of internal trading characteristics and external market events on trading risks. In this way, the system can provide a comprehensive and quantified trading risk value to help users understand the risk level of their upcoming transactions under current market conditions, thereby ensuring that the decisions of the trading signal system are based on an in-depth understanding of existing trading patterns and market dynamics, thereby improving the accuracy of risk prediction.

[0065] In an optional embodiment, after obtaining the transaction risk value corresponding to the target transaction, if the transaction risk value is greater than a preset risk value, the generation system generates a warning message, wherein the warning message is used to prompt the user to interrupt the target transaction.

[0066] Optionally, the early warning information is a warning message automatically generated by the system when the transaction risk value exceeds a preset risk value, which is used to prompt the user to interrupt or reconsider the upcoming target transaction to avoid high-risk financial transactions.

[0067] Optionally, the above steps are risk control links in the trading signal generation system. After the system analyzes the transaction risk value, it will compare it with the preset risk value. If the transaction risk value is greater than the preset risk value, it indicates that the potential risk of this transaction exceeds the tolerance of the user or institution. At this time, the system will immediately generate an early warning message to remind the user to interrupt the transaction or take other risk management measures. This mechanism ensures that users can understand the potential risk level in a timely manner when making trading decisions, so that they can make more cautious and rational choices and prevent unnecessary asset losses.

[0068] From the above content, it can be seen that the present application adopts a comprehensive analysis method of real-time financial transaction information and event information related to the transaction, by pre-constructing typical features (i.e., L first features and M second features), and then determining the transaction risk value corresponding to the target transaction based on the first information, the second information, the L first features and the M second features, and then using the transaction risk value as the judgment condition for whether to generate a transaction signal, thereby achieving the purpose of improving the accuracy and response speed of the transaction signal, thereby achieving the technical effect of providing users with customized, high-precision financial transaction signals, and thus solving the technical problem of low accuracy in the generation of transaction signals in existing financial transaction systems due to lack of understanding of the financial product information and / or transaction object information involved in user transactions.

[0069] According to another aspect of the embodiment of the present application, a trading signal system is also provided. Figure 2 is a schematic diagram of an optional trading signal system according to an embodiment of the present application, such as Figure 2As shown, the system includes: a message extraction module, a message analysis module, a customer model reverse engineering module and a model generation module.

[0070] Optionally, the construction process of the trading signal system includes:

[0071] Step S201: select expert customers with stable positive profit and loss output from the customer transaction record library, and derive the initial transaction model through the customer model reverse engineering module by combining the message analysis module with the transaction indicator formula library and the product historical price library.

[0072] In step S202, the message extraction module obtains relevant event information of the corresponding transaction from the external news site according to the time point of the important news, and the message analysis module determines the impact of the event on the revenue of the corresponding transaction product based on the relevant event information and the corpus.

[0073] Step S203: Based on the feedback result in step S202 and the expert client's trading behavior before and after the time period, an event-driven expert trading model is formed.

[0074] Step S204 , a model generation module forms a complete transaction model of the expert client based on the expert client's initial transaction model and the event-driven expert transaction model.

[0075] Step S205: Aggregate multiple expert client trading models to form a complete trading signal system that can be recommended externally.

[0076] Optionally, during the derivation of the initial trading model, first obtain information such as the time, trading type, and trading price of each transaction of the expert customer from the customer transaction record library. Then, obtain the full set of technical indicator formulas from the trading indicator formula library. Then, extract one formula from the full set of indicator formulas, substitute the customer's transaction data into it, and traverse the dynamic parameter calculation results in the formula to obtain the dynamic parameter values ​​applicable to the transactions in the table. For example:

[0077] The extracted formula is: F(c, x)=Y, where C=transaction price, x=dynamic parameter built into the formula (assuming possible values ​​are 1-1000), and Y=calculation result (0=no trading signal, 1=trading signal).

[0078] The traversal results are shown in List 1 below:

[0079] Table 1

[0080] X Y 1 0 2 0 3 0 4 0 ........ 0 200 1 201 1 202 1 203 0 204 0 ........ 0 999 0 1000 0

[0081] As shown in Table 1 above, after traversal, Y = 1 only when the dynamic parameters X = 200, 201, and 202 are present, and Y = 0 for the remaining parameters. Therefore, it can be determined that if the expert client uses this formula as its trading model, its dynamic parameters may be (200, 201, 202).

[0082] Optionally, after traversing the full formula, the possible dynamic parameter set of the full formula when used as an expert client transaction is shown in Table 2 below:

[0083] Table 2

[0084] formula Dynamic parameter values F1 200，201，202 F2 1，2 F3 100 F4 120，121

[0085] Optionally, the trading signal system expands the data of each transaction record of the expert client to obtain a dynamic parameter set of the formula applicable to each transaction, as shown in List 3 below:

[0086] Table 3

[0087]

[0088]

[0089] Next, the trading signal system divides the obtained set into two groups based on the trading direction data. It traverses the two sets and sorts the formulas using the same dynamic parameters from high to low. The more formulas using the same parameters, the closer the formula and parameters are to the expert client's trading model. For example, the buy direction group is shown in Table 4 below:

[0090] Table 4

[0091]

[0092] The sell direction group is shown in Table 5 below.

[0093] Table 5

[0094]

[0095]

[0096] Optionally, the two sets are traversed separately to sort the formulas using the same dynamic parameters from high to low. The sorting results of the buy group are shown in Table 6 below:

[0097] Table 6

[0098] formula Dynamic parameters Occurrences F1 200 2 F1 201 1 F1 202 1 F1 198 1 F1 199 1 F2 1 1 F2 2 1 F2 10 1 F3 100 1 F3 200 1 F4 120 1 F4 121 1 F4 200 1 F4 201 1

[0099] Optionally, the selling group sorting result is shown in the following table 7:

[0100] Table 7

[0101] formula Dynamic parameters Occurrences F4 300 2 F4 301 1 F4 299 1 F1 100 1 F1 101 1 F1 90 1 F1 91 1 F1 92 1 F2 9 1 F2 10 1 F2 11 1 F2 300 1 F3 200 1 F3 100 1

[0102] From the above table, we can see that formula F1 (200) in the two transaction samples of the buying group both meet the trading conditions of the expert client and is the trading model that best restores the expert client; similarly, formula F4 (300) in the selling group meets the trading conditions of the expert client.

[0103] Based on this principle, there may be multiple formulas that satisfy the expert customer trading model in more customer trading samples. The system can select from them to form an expert customer model set based on a specific principle (such as proportion or ranking), for example, {model 1, model 2, model 3,…, model N}. Afterwards, this set will serve as one of the components of the subsequent trading signals sent to the majority of customers.

[0104] According to another aspect of the embodiment of the present application, a device for generating a transaction signal is provided. Figure 3 is a schematic diagram of an optional trading signal generating device according to an embodiment of the present application, such as Figure 3 As shown, the transaction signal generating device includes: an information collecting unit 301 , a first determining unit 302 and a first generating unit 303 .

[0105] Optionally, the information collection unit is used to collect first information and second information corresponding to the target transaction, wherein the target transaction is a financial transaction to be initiated by the target user, the first information includes at least the transaction time, transaction type, financial product type and transaction object, and the second information is information of a target event in the open source platform that is relevant to the financial products and / or transaction objects involved in the target transaction; the first determination unit is used to determine the transaction risk value corresponding to the target transaction based on the first information, the second information, L first features and M second features, wherein the L first features are transaction features of historical transactions with a profit ratio greater than or equal to a preset ratio in different dimensions, and the M second features are event features of historical events with a correlation with the profit ratio of historical transactions in different dimensions, and both L and M are positive integers; the first generation unit is used to generate a target signal when the transaction risk value is less than or equal to the preset risk value, wherein the target signal is used to prompt the user to complete the target transaction.

[0106] In an optional embodiment, the transaction signal generating device further includes: a second determination unit, a screening unit, and a training unit.

[0107] Optionally, the second determining unit is configured to determine a revenue label corresponding to each of the P historical transactions based on transaction information of the P historical transactions, where P is a positive integer and the revenue label is used to represent the revenue ratio of the corresponding historical transaction; the screening unit is configured to screen media information in the open source platform based on the transaction time corresponding to each historical transaction to obtain X historical events, where X is a positive integer and the X historical events are historical events that are relevant to the financial products and / or transaction objects involved in the P historical transactions; and the training unit is configured to train an initial neural network model based on the transaction information of the P historical transactions, event information corresponding to the X historical events, and the revenue label corresponding to each historical transaction to obtain a target model.

[0108] In an optional embodiment, the training unit includes: a first extraction subunit, a second extraction subunit, a training subunit, and a first determination subunit.

[0109] Optionally, the first extraction subunit is configured to perform feature extraction on transaction information of P historical transactions to obtain P historical transaction features, wherein the P historical transactions correspond one-to-one to the P historical transaction features; the second extraction subunit is configured to perform feature extraction on event information corresponding to X historical events to obtain X historical event features, wherein the X historical events correspond one-to-one to the X historical event features; the training subunit is configured to train an initial neural network model based on the P historical transaction features, the X historical event features, and the profit label corresponding to each historical transaction to obtain a first submodel and a second submodel, wherein the first submodel is configured to store at least L first features and M second features, and the second submodel is configured to determine a transaction risk value corresponding to a target transaction; and the first determination subunit is configured to use the first submodel and the second submodel as target models.

[0110] In an optional embodiment, the training subunit includes: a first determination module, a second determination module, a third determination module, and a fourth determination module.

[0111] Optionally, the first determination module is configured to use, as positive transaction features, historical transaction features with a profit label as a target label among the P historical transaction features, to obtain Q positive transaction features, where Q is a positive integer less than or equal to P, and the target label is used to indicate that the profit ratio of the corresponding historical event is greater than or equal to a preset ratio; the second determination module is configured to use, as typical event features, historical event features that are correlated with Q historical transactions corresponding to the Q positive transaction features among the X historical event features, to obtain Y typical event features, where Y is less than or equal to a positive integer; the third determination module is configured to determine L first features based on the Q positive transaction features, where the L first features are sub-features of each positive transaction feature; and the fourth determination module is configured to determine M second features based on the Y typical event features, where the M second features are sub-features of each typical event feature.

[0112] In an optional embodiment, the transaction signal generating device further includes: a first extraction unit, a second extraction unit, and a third determination unit.

[0113] Optionally, the first extraction unit is used to perform feature extraction on the first information through the first sub-model in the target model to obtain target transaction features corresponding to the target transaction; the second extraction unit is used to perform feature extraction on the second information through the first sub-model to obtain target event features corresponding to the target event; and the third determination unit is used to determine the transaction risk value based on the target transaction features, the target event features, L first features and M second features.

[0114] In an optional embodiment, the third determining unit further includes: a first summing subunit, a second summing subunit, and a second determining subunit.

[0115] Optionally, the first summation subunit is used to perform weighted summation of the similarities between the target transaction feature and each of the L first features to obtain a first similarity; the second summation subunit is used to perform weighted summation of the similarities between the target event feature and each of the M second features to obtain a second similarity; and the second determination subunit is used to determine the transaction risk value based on the first similarity and the second similarity through the second submodel in the target model.

[0116] In an optional embodiment, the transaction signal generating device further includes: a second generating unit.

[0117] Optionally, the second generating unit is configured to generate warning information when the transaction risk value is greater than a preset risk value, wherein the warning information is used to prompt the user to interrupt the target transaction.

[0118] From the above content, it can be seen that the generating device adopts a comprehensive analysis method of real-time financial transaction information and event information related to the transaction, by pre-constructing typical features (i.e., L first features and M second features). Then, the generating device determines the transaction risk value corresponding to the target transaction based on the first information, the second information, the L first features and the M second features, and then uses the transaction risk value as the judgment condition for whether to generate a transaction signal, thereby achieving the purpose of improving the accuracy and response speed of the transaction signal, thereby achieving the technical effect of providing users with customized, high-precision financial transaction signals, and thus solving the technical problem of low accuracy in the generation of transaction signals in existing financial transaction systems due to lack of understanding of the financial product information and / or transaction object information involved in user transactions.

[0119] According to another aspect of an embodiment of the present application, a computer program product is further provided. The computer program product includes a stored computer program, wherein when the computer program is running, the computer program product is controlled to execute any one of the above-mentioned methods for generating a trading signal.

[0120] According to another aspect of the embodiments of the present application, an electronic device is provided. Figure 4 This is a structural block diagram of an electronic device according to an embodiment of the present application. Figure 4 As shown, the electronic device may include: one or more ( Figure 4 Only one is shown) processor 402, memory 404, storage controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0121] Among them, the memory can be used to store software programs and modules, such as program instructions / modules corresponding to the methods and devices in the embodiments of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, implementing the above-mentioned method. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the terminal via a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network and a combination thereof.

[0122] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: collecting first information and second information corresponding to the target transaction, wherein the target transaction is a financial transaction to be initiated by the target user, the first information at least includes the transaction time, transaction type, financial product type and transaction object, and the second information is the information of the target event in the open source platform that is relevant to the financial products and / or transaction objects involved in the target transaction; determining the transaction risk value corresponding to the target transaction based on the first information, the second information, L first features and M second features, wherein the L first features are transaction features of historical transactions with a return ratio greater than or equal to a preset ratio in different dimensions, and the M second features are event features of historical events with a return ratio relevant to the return ratio of historical transactions in different dimensions, and L and M are both positive integers; when the transaction risk value is less than or equal to the preset risk value, generating a target signal, wherein the target signal is used to prompt the user to complete the target transaction.

[0123] The processor can call information and applications stored in the memory through the transmission device to perform the following steps: determining a profit label corresponding to each of the P historical transactions based on transaction information of the P historical transactions, where P is a positive integer and the profit label is used to represent the profit ratio of the corresponding historical transaction; filtering media information in the open source platform based on the transaction time corresponding to each historical transaction to obtain X historical events, where X is a positive integer and the X historical events are historical events that are relevant to the financial products and / or transaction objects involved in the P historical transactions; and training an initial neural network model based on the transaction information of the P historical transactions, event information corresponding to the X historical events, and the profit label corresponding to each historical transaction to obtain a target model.

[0124] The processor can call information and applications stored in the memory through the transmission device to perform the following steps: extracting features from transaction information of P historical transactions to obtain P historical transaction features, wherein the P historical transactions correspond one-to-one to the P historical transaction features; extracting features from event information corresponding to X historical events to obtain X historical event features, wherein the X historical events correspond one-to-one to the X historical event features; training an initial neural network model based on the P historical transaction features, the X historical event features, and a profit label corresponding to each historical transaction to obtain a first sub-model and a second sub-model, wherein the first sub-model is used to store at least L first features and M second features, and the second sub-model is used to determine a transaction risk value corresponding to a target transaction; and using the first sub-model and the second sub-model as target models.

[0125] The processor can call information and applications stored in the memory through the transmission device to perform the following steps: using historical transaction features with a profit label as a target label among P historical transaction features as positive transaction features to obtain Q positive transaction features, where Q is a positive integer less than or equal to P, and the target label is used to indicate that the profit ratio of the corresponding historical event is greater than or equal to a preset ratio; using historical event features that are correlated with Q historical transactions corresponding to the Q positive transaction features among X historical event features as typical event features to obtain Y typical event features, where Y is less than or equal to a positive integer; determining L first features based on the Q positive transaction features, where the L first features are sub-features of each positive transaction feature; and determining M second features based on the Y typical event features, where the M second features are sub-features of each typical event feature.

[0126] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: extract features of the first information through the first sub-model in the target model to obtain target transaction features corresponding to the target transaction; extract features of the second information through the first sub-model to obtain target event features corresponding to the target event; determine the transaction risk value based on the target transaction features, target event features, L first features and M second features.

[0127] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: perform weighted summation of the similarities between the target transaction feature and each of the L first features to obtain a first similarity; perform weighted summation of the similarities between the target event feature and each of the M second features to obtain a second similarity; and determine the transaction risk value based on the first similarity and the second similarity through the second sub-model in the target model.

[0128] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: when the transaction risk value is greater than the preset risk value, generate an early warning message, wherein the early warning message is used to prompt the user to interrupt the target transaction.

[0129] An embodiment of the present application provides a technical solution for a transaction generation method. The present application adopts a comprehensive analysis method of real-time financial transaction information and event information related to the transaction, and pre-constructs typical features (i.e., L first features and M second features). Subsequently, the present application determines the transaction risk value corresponding to the target transaction based on the first information, the second information, the L first features, and the M second features, and then uses the transaction risk value as a judgment condition for whether to generate a transaction signal, thereby achieving the purpose of improving the accuracy and response speed of the transaction signal, thereby achieving the technical effect of providing users with customized, high-precision financial transaction signals, and further solving the technical problem of low accuracy in generating transaction signals in existing financial transaction systems due to lack of understanding of the financial product information and / or transaction object information involved in user transactions.

[0130] It can be understood by those skilled in the art that Figure 4 The structure shown is for illustration only, and the electronic device may also be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a PDA, a mobile Internet device (MID), a PAD, or other terminal devices. Figure 4 It does not limit the structure of the above electronic device. For example, the electronic device may also include Figure 4 More or fewer components (such as network interfaces, display devices, etc.) shown in, or with Figure 4 Different configurations shown.

[0131] A person skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0132] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0133] In the above embodiments of the present application, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0134] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0135] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0136] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0137] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for enabling a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk.

[0138] The above is only a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for generating a trading signal, characterized in that: include: Collecting first information and second information corresponding to a target transaction, wherein the target transaction is a financial transaction to be initiated by a target user, the first information including at least the transaction time, transaction type, financial product type, and transaction object, and the second information is information on a target event on the open source platform that is relevant to the financial product and / or transaction object involved in the target transaction; Determining a transaction risk value corresponding to the target transaction based on the first information, the second information, L first features, and M second features, wherein the L first features are transaction features in different dimensions of historical transactions whose profit ratios are greater than or equal to a preset ratio, and the M second features are event features in different dimensions of historical events that are correlated with the profit ratios of the historical transactions, where L and M are both positive integers; When the transaction risk value is less than or equal to a preset risk value, a target signal is generated, wherein the target signal is used to prompt the user to complete the target transaction.

2. The method for generating a trading signal according to claim 1, wherein: Determining the transaction risk corresponding to the target transaction by using a target model based on the first information, the second information, the L first features, and the M second features, wherein the training step of the target model includes: Determining a profit label corresponding to each of the P historical transactions based on transaction information of the P historical transactions, where P is a positive integer and the profit label is used to represent a profit ratio of the corresponding historical transaction; Filtering media information in the open source platform based on the transaction time corresponding to each historical transaction to obtain X historical events, where X is a positive integer, and the X historical events are historical events that are relevant to the financial products and / or transaction objects involved in the P historical transactions; The initial neural network model is trained based on the transaction information of the P historical transactions, the event information corresponding to the X historical events, and the revenue label corresponding to each historical transaction to obtain the target model.

3. The method for generating a trading signal according to claim 2, wherein: The initial neural network model is trained based on the transaction information of the P historical transactions, the event information corresponding to the X historical events, and the revenue label corresponding to each historical transaction to obtain the target model, including: Performing feature extraction on the transaction information of the P historical transactions to obtain P historical transaction features, wherein the P historical transactions correspond one-to-one to the P historical transaction features; Performing feature extraction on event information corresponding to the X historical events to obtain X historical event features, wherein the X historical events correspond one-to-one to the X historical event features; The initial neural network model is trained based on the P historical transaction features, the X historical event features, and the revenue label corresponding to each historical transaction to obtain a first sub-model and a second sub-model, wherein the first sub-model is used to store at least the L first features and the M second features, and the second sub-model is used to determine the transaction risk value corresponding to the target transaction; The first sub-model and the second sub-model are used as the target models.

4. The method for generating a trading signal according to claim 3, wherein: In the process of training the initial neural network model based on the P historical transaction features, the X historical event features, and the profit label corresponding to each historical transaction, the transaction signal generation method further includes: The historical transaction features with the profit label as the target label among the P historical transaction features are used as positive transaction features to obtain Q positive transaction features, where Q is a positive integer less than or equal to P, and the target label is used to indicate that the profit ratio of the corresponding historical event is greater than or equal to the preset ratio; Taking the historical event features of the X historical event features that are correlated with the Q historical transactions corresponding to the Q positive transaction features as typical event features, to obtain Y typical event features, where Y is a positive integer less than or equal to; Determining the L first features based on the Q positive transaction features, wherein the L first features are sub-features of each positive transaction feature; The M second features are determined based on the Y typical event features, wherein the M second features are sub-features of each typical event feature.

5. The method for generating a trading signal according to claim 1, wherein: Determining a transaction risk value corresponding to the target transaction based on the first information, the second information, the L first features, and the M second features includes: Performing feature extraction on the first information using a first sub-model in a target model to obtain a target transaction feature corresponding to the target transaction; Performing feature extraction on the second information using the first sub-model to obtain target event features corresponding to the target event; The transaction risk value is determined based on the target transaction feature, the target event feature, the L first features, and the M second features.

6. The method for generating a trading signal according to claim 5, wherein: Determining the transaction risk value based on the target transaction feature, the target event feature, the L first features, and the M second features includes: Performing a weighted summation of similarities between the target transaction feature and each of the L first features to obtain a first similarity; Performing a weighted summation on the similarity between the target event feature and each of the M second features to obtain a second similarity; The transaction risk value is determined based on the first similarity and the second similarity by a second sub-model in the target model.

7. The method for generating a trading signal according to claim 1, wherein: After determining the transaction risk value corresponding to the target transaction based on the first information, the second information, the L first features, and the M second features, the transaction signal generation method further includes: In a case where the transaction risk value is greater than the preset risk value, a warning message is generated, wherein the warning message is used to prompt the user to interrupt the target transaction.

8. A trading signal generating device, characterized in that: include: an information collection unit, configured to collect first information and second information corresponding to a target transaction, wherein the target transaction is a financial transaction to be initiated by a target user, the first information including at least the transaction time, transaction type, financial product type, and transaction object, and the second information is information on a target event on the open source platform that is relevant to the financial product and / or transaction object involved in the target transaction; a first determining unit configured to determine a transaction risk value corresponding to the target transaction based on the first information, the second information, L first features, and M second features, wherein the L first features are transaction features in different dimensions of historical transactions with a return ratio greater than or equal to a preset ratio, and the M second features are event features in different dimensions of historical events that are correlated with the return ratio of the historical transactions, and L and M are both positive integers; The first generating unit is configured to generate a target signal when the transaction risk value is less than or equal to a preset risk value, wherein the target signal is used to prompt a user to complete the target transaction.

9. A computer program product, characterized in that The computer program product includes a computer program, wherein when the computer program is run, the computer program product is controlled to execute the method for generating a trading signal according to any one of claims 1 to 7.

10. An electronic device, characterized in that: The system comprises one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method for generating a trading signal according to any one of claims 1 to 7.

Images