Transaction truth risk assessment method and transaction truth risk assessment device

Abstract

The invention provides a transaction truth risk assessment method and a transaction truth risk assessment device. The method comprises the steps of obtaining standardized transaction information data; and inputting the standardized transaction information data into a target risk assessment model, and outputting a risk index corresponding to the risk type, the target risk assessment model being a convergence model obtained by using multiple groups of training data to perform iterative training by using an unsupervised learning method or a supervised learning method, each group of training data comprises sample standardized transaction data and a corresponding sample risk index, the risk type at least comprises one of the following: a transaction user risk, a financial product risk, a transaction product risk and a quotation information risk, and the transaction user risk is a credit risk of a user initiating a transaction application; and when the risk index is greater than the corresponding set index threshold, sending out prompt information. The problem of low transaction efficiency caused by incapability of accurately evaluating the authenticity and transaction risk of the transaction user in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of bank transaction risk assessment, and in particular to a transaction authenticity risk assessment method, a transaction authenticity risk assessment device and a transaction authenticity risk assessment system. Background Art

[0002] With the continuous development of information technology and Internet technology, the traditional offline bargaining transaction model in the interbank financial field is transitioning to online, digital and platform-based. In the process of bargaining and communication between traders based on the Internet platform, inquiry, quotation and bargaining are the most common communication scenarios. In this process, many important transaction elements and privacy information are involved, such as price, interest rate, term and contact information. Traders need to carefully judge whether various situations can be disclosed to trading users. In the process of inquiry, quotation and bargaining, existing financial interbank traders often determine the financial product element information in the inquiry, quotation and bargaining process based on the familiarity of the trading users and their own judgment on the transaction price of financial products. It is usually necessary to judge whether the trading users have real trading intentions, or whether they are financial intermediaries that collect quotation information. In addition, in the bargaining process, for the quotation information given by the trading users, traders must also judge the authenticity of the quotation, whether it is a transaction that cannot be concluded, or just to interfere with the trader's judgment of the market.

[0003] Existing methods or systems only analyze user behaviors and financial product elements, and judge the authenticity and rationality of interbank financial transaction quotations by setting guiding prices, price fluctuation thresholds and other parameters. On the one hand, they ignore the real-time changes in the financial market, and transaction quotations cannot reflect the real situation of the market in real time, resulting in traders being unable to accurately judge the accuracy of quotations; on the other hand, they do not store and analyze the behavioral data of users in the inquiry, quotation and bargaining stages, and do not use models to analyze abnormal transaction elements or transaction behaviors to alert traders. Summary of the invention

[0004] The main purpose of the present application is to provide a transaction authenticity risk assessment method, a transaction authenticity risk assessment device and a transaction authenticity risk assessment system, so as to at least solve the problem in the prior art that the authenticity and transaction risks of transaction users cannot be accurately assessed, resulting in low transaction efficiency.

[0005] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a transaction authenticity risk assessment method is provided, comprising: obtaining standardized transaction information data, the standardized transaction information data at least including standardized transaction user information, financial product element information, financial product price index, transaction data and external data, the financial product price index at least including Shibor interest rate, DR interest rate and recent transaction records of corresponding financial products; inputting the standardized transaction information data into a target risk assessment model, and outputting a risk index corresponding to the risk type, one risk type corresponds to at least one risk index, the target risk assessment model is a convergence model obtained by iterative training using multiple groups of training data using an unsupervised learning method or a supervised learning method, each group of the training data includes sample standardized transaction data and a corresponding sample risk index, the risk type includes at least one of the following: transaction user risk, financial product risk, transaction product risk and quotation information risk, the transaction user risk is the credit risk of the user who initiates the transaction application; when the risk index is greater than the corresponding set index threshold, a prompt message is issued, and the prompt message is used to prompt that there is a risk in the current transaction.

[0006] Optionally, obtaining standardized transaction information data includes: obtaining original transaction information data, wherein the original transaction information data is the transaction user information, the financial product element information, the financial product price index, the transaction data and the external data that have not been standardized; performing a data cleaning operation on the original transaction information data to obtain clean transaction information data, wherein the data cleaning operation at least includes operations of filling missing values ​​and removing outliers; and performing standardization on the clean transaction information data to obtain the standardized transaction information data.

[0007] Optionally, after obtaining standardized transaction information data, the method further includes: performing feature selection and classification based on the standardized transaction information data to obtain multiple risk features, and setting a corresponding feature score for each risk feature, wherein the risk features include at least transaction behavior features, geographic location features, financial product features, and customer behavior features.

[0008] Optionally, the risk type includes the transaction user risk, the standardized transaction information data is input into a target risk assessment model, and a risk index corresponding to the risk type is output, including: when the target risk assessment model is a convergence model obtained by iterative training using the supervised learning method, the product between the feature score corresponding to the relevant risk feature in the standardized transaction information data and the set weight coefficient is calculated to obtain the risk index of the transaction user risk, the relevant risk feature is a feature in the risk feature that is relevant to the transaction user, and the risk feature at least includes the behavioral characteristics, geographic location characteristics, device characteristics and time characteristics of the transaction user; when the target risk assessment model is a convergence model obtained by iterative training using the unsupervised learning method, the risk information of the transaction user is obtained from the standardized transaction information data, and the risk information is identified as abnormal to obtain the risk index of the transaction user risk, the risk information at least includes the region where the transaction user is located, the product browsing record and transaction record of the transaction user within a set historical period.

[0009] Optionally, the risk information is identified as abnormal, including: judging whether the region where the transaction user is located is marked as high risk, whether the transaction user uses a VPN or a proxy, and whether the geographical location of the transaction user deviates from historical browsing visits based on the region where the transaction user is located in the risk information; analyzing whether small transactions exceed a set number of transactions within a set time, or whether the total number of transactions within the set time exceeds the set number of transactions based on the transaction records in the risk information; judging that the transaction user has transaction abnormality problems if abnormal rules are met, and the abnormal rules are that the region where the transaction user is located is marked as high risk, the transaction user uses a VPN or a proxy, the geographical location of the transaction user deviates from historical browsing visits, small transactions exceed a set number of transactions within a set time, and the total number of transactions within the set time exceeds the set number of transactions to meet a set number; judging that the transaction user does not have transaction abnormality problems if the abnormal rules are not met.

[0010] Optionally, the risk type includes the financial product risk, and the standardized transaction information data is input into a target risk assessment model, and a risk index corresponding to the risk type is output, including: when the target risk assessment model is a convergence model obtained by iterative training using the supervised learning method, the financial product characteristics of the financial product and the feature scores corresponding to the financial product characteristics are obtained from the standardized transaction information data; the feature scores corresponding to each of the financial product characteristics are multiplied by a set weight coefficient to obtain the risk index of the financial product risk, the financial product characteristics at least include the rate of return, the reputation and qualifications of the issuing institution, the risk rating of the financial product, the market environment and the macroeconomic conditions, the risk index of the financial product risk at least includes the capital cost risk index, the liquidity risk index and the price risk index of the financial product, and the liquidity risk index represents the risk index of the liquidity conditions of the financial product in the market; when the target risk assessment model is a convergence model obtained by iterative training using the unsupervised learning method, the transaction data of the financial product is obtained from the standardized transaction information data, and risk assessment is performed based on the transaction data of the financial product to obtain the risk index of the financial product risk.

[0011] Optionally, when the risk index is less than the corresponding set index threshold, a prompt message is issued, including: when the risk type is the financial product risk, the price risk index, the liquidity risk index, and the capital cost risk index of the financial product are compared with the set index threshold; when the risk prompt rule is triggered, the prompt message is issued, and the risk prompt rule is that one or two of the risk indexes among the price risk index, the liquidity risk index, and the capital cost risk index are higher than the first set index threshold; when the prohibition trading rule is triggered, the current transaction is terminated, and the prohibition trading rule is that the price risk index, the liquidity risk index, and the capital cost risk index are all higher than the first set index threshold; in case the risk type is the transaction user risk, comparing the risk index corresponding to the transaction user risk with a second set index threshold; in case the risk index corresponding to the transaction user risk is higher than the second set index threshold, determining that the transaction user has a transaction credit risk and issuing the prompt message; in case the risk type is the quotation information risk, receiving the quotation information of the current financial product from the transaction user, querying the current market price of the current financial product, and comparing the quotation information with the current market price; in case the absolute value of the difference between the quotation information and the current market price is greater than the price index threshold, determining that the authenticity of the current transaction is low, issuing the prompt message and terminating the current transaction.

[0012] Optionally, after inputting the standardized transaction information data into the target risk assessment model and outputting a risk index corresponding to the risk type, the method further includes: when the risk index is greater than or equal to the corresponding set index threshold, continuing the current transaction until the transaction is completed, and correcting the characteristic score of the risk characteristics in real time, the risk characteristics including at least transaction behavior characteristics, geographic location characteristics, financial product characteristics and customer behavior characteristics.

[0013] According to another aspect of the present application, a transaction authenticity risk assessment device is provided, the device comprising: an acquisition unit, used to acquire standardized transaction information data, the standardized transaction information data at least comprising standardized transaction user information, financial product element information, financial product price index, transaction data and external data, the financial product price index at least comprising Shibor interest rate, DR interest rate and recent transaction records of corresponding financial products; an input-output unit, used to input the standardized transaction information data into a target risk assessment model, and output a risk index corresponding to a risk type, one risk type corresponds to at least one risk index, the target risk assessment model is a convergence model obtained by iterative training using multiple groups of training data using an unsupervised learning method or a supervised learning method, each group of the training data comprises sample standardized transaction data and a corresponding sample risk index, the risk type comprises at least one of the following: transaction user risk, financial product risk, transaction product risk and quotation information risk; a prompt unit, used to issue a prompt message when the risk index is greater than the corresponding set index threshold, the prompt message is used to prompt that there is a risk in the current transaction and the current transaction needs to be terminated.

[0014] According to another aspect of the present application, a transaction authenticity risk assessment system is provided, comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include methods for executing any one of the described methods.

[0015] Applying the technical solution of the present application, in a transaction authenticity risk assessment method, first, standardized transaction information data is obtained, the standardized transaction information data at least includes standardized transaction user information, financial product element information, financial product price index, transaction data and external data, the financial product price index at least includes Shibor interest rate, DR interest rate and recent transaction records of corresponding financial products; then, the standardized transaction information data is input into a target risk assessment model, and a risk index corresponding to the risk type is output, one risk type corresponds to at least one risk index, the target risk assessment model is a convergence model obtained by iterative training using multiple groups of training data using an unsupervised learning method or a supervised learning method, each group of the training data includes sample standardized transaction data and a corresponding sample risk index, the risk type includes at least one of the following: transaction user risk, financial product risk, transaction product risk and quotation information risk, the transaction user risk is the credit risk of the user who initiates the transaction application; finally, when the risk index is less than or greater than the corresponding set index threshold, a prompt message is issued, and the prompt message is used to prompt that there is a risk in the current transaction. This application analyzes the behavioral data of transaction users and the key elements of financial products, and analyzes them in combination with various price indices in the financial market, inputs various transaction data into the target risk assessment model, obtains a risk index, and compares the risk index with the corresponding risk threshold to provide risk warnings for the interbank financial transaction process and avoid leakage of transaction elements. The transaction authenticity of transaction users is evaluated and risk warnings are provided to help traders reduce communication costs, reduce transaction risks, improve transaction efficiency, and more efficiently judge various inquiries and quotations that are not for the purpose of real transactions. This application solves the problem of low transaction efficiency caused by the inability to accurately evaluate the authenticity and transaction risks of transaction users in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A hardware structure block diagram of a mobile terminal for executing a transaction authenticity risk assessment method provided in an embodiment of the present application is shown;

[0017] Figure 2 A schematic diagram of a process of a transaction authenticity risk assessment method provided according to an embodiment of the present application is shown;

[0018] Figure 3 A structural block diagram of a transaction authenticity risk assessment device provided according to an embodiment of the present application is shown.

[0019] The above drawings include the following reference numerals:

[0020] 102, processor; 104, memory; 106, transmission device; 108, input and output devices. DETAILED DESCRIPTION

[0021] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0022] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0023] 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 interchanged where appropriate, so that the embodiments of the present application described here. 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.

[0024] As introduced in the background technology, the existing methods or systems in the prior art only analyze user behaviors and financial product elements, and judge the authenticity and rationality of interbank financial transaction quotations by setting guide prices, price fluctuation thresholds and other parameters. On the one hand, they ignore the real-time changes in the financial market, and the transaction quotations cannot reflect the real situation of the market in real time, resulting in traders being unable to accurately judge the accuracy of the quotations; on the other hand, they do not store and analyze the behavioral data of users in the inquiry, quotation and bargaining links, and analyze abnormal transaction elements or transaction behaviors through models to remind traders. In order to solve the problem that the authenticity and transaction risks of transaction users cannot be accurately assessed in the prior art, resulting in low transaction efficiency, the embodiments of the present application provide a transaction authenticity risk assessment method, a transaction authenticity risk assessment device and a transaction authenticity risk assessment system.

[0025] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0026] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 11 is a hardware structure block diagram of a mobile terminal for executing a transaction authenticity risk assessment method according to an embodiment of the present invention. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0027] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the transaction authenticity risk assessment method in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, the above method is implemented. The memory 104 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 examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories can be connected to the mobile 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. The transmission device 106 is used to receive or send data via a network. The above-mentioned specific examples of the network may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0028] In this embodiment, a transaction authenticity risk assessment method running on a mobile terminal, a computer terminal or a similar computing device 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.

[0029] Figure 2 FIG. 1 is a flow chart of a transaction authenticity risk assessment method according to an embodiment of the present application. Figure 2 As shown, the method comprises the following steps:

[0030] Step S201, obtaining standardized transaction information data, the standardized transaction information data at least includes standardized transaction user information, financial product element information, financial product price index, transaction data and external data, the financial product price index at least includes Shibor interest rate, DR interest rate and recent transaction records of corresponding financial products.

[0031] Specifically, data collection includes user information, financial product element information, price index of financial products, transaction data, and external data. Customer data includes basic customer information, user region, recent product browsing records, and historical transaction records; financial product element information includes interest rate, price, term, amount, counterparty category, etc.; financial product price index includes: Shibor interest rate, DR interest rate and recent transaction records of corresponding products; transaction data includes transaction amount, transaction time, transaction product, counterparty information, etc. External data mainly includes customer evaluation data within the bank, etc.

[0032] Step S202: input the standardized transaction information data into a target risk assessment model, and output a risk index corresponding to the risk type. One risk type corresponds to at least one risk index. The target risk assessment model is a convergence model obtained by iterative training using multiple sets of training data using an unsupervised learning method or a supervised learning method. Each set of the training data includes sample standardized transaction data and a corresponding sample risk index. The risk types include at least one of the following: transaction user risk, financial product risk, transaction product risk, and quotation information risk. The transaction user risk is the credit risk of the user who initiates the transaction application.

[0033] Specifically, it supports supervised learning models and unsupervised learning models. Supervised learning models require labeled data for training. Unsupervised learning models are suitable for risk assessment by identifying abnormal behaviors without clear labels. According to the needs of risk assessment, select appropriate unsupervised learning (such as cluster analysis) or supervised learning (such as logistic regression, decision tree, neural network) models. The collected data is cleaned, normalized, feature engineered, etc. to meet the input requirements of the model. Use multiple sets of training data (including sample standardized transaction data and corresponding sample risk index) for iterative training until the model converges. Input the standardized transaction information data into the target risk assessment model and output the corresponding risk index.

[0034] Step S203: When the risk index is greater than the corresponding set index threshold, a prompt message is issued, wherein the prompt message is used to indicate that the current transaction has risks.

[0035] Specifically, a risk index threshold is set for each risk type. When the risk index exceeds the set threshold, the system automatically issues a prompt message to indicate that the current transaction is risky.

[0036] In this embodiment, first, standardized transaction information data is obtained, and the standardized transaction information data at least includes standardized transaction user information, financial product element information, financial product price index, transaction data and external data, and the financial product price index at least includes Shibor interest rate, DR interest rate and recent transaction records of corresponding financial products; then, the standardized transaction information data is input into the target risk assessment model, and the risk index corresponding to the risk type is output, and one risk type corresponds to at least one risk index, and the target risk assessment model is a convergence model obtained by iterative training using multiple groups of training data using an unsupervised learning method or a supervised learning method, and each group of the training data includes sample standardized transaction data and a corresponding sample risk index, and the risk type includes at least one of the following: transaction user risk, financial product risk, transaction product risk and quotation information risk, and the transaction user risk is the credit risk of the user who initiates the transaction application; finally, when the risk index is less than or greater than the corresponding set index threshold, a prompt message is issued, and the prompt message is used to prompt that there is a risk in the current transaction. This application analyzes the behavioral data of transaction users and the key elements of financial products, and analyzes them in combination with various price indices in the financial market, inputs various transaction data into the target risk assessment model, obtains a risk index, and compares the risk index with the corresponding risk threshold to provide risk warnings for the interbank financial transaction process and avoid leakage of transaction elements. The transaction authenticity of transaction users is evaluated and risk warnings are provided to help traders reduce communication costs, reduce transaction risks, improve transaction efficiency, and more efficiently judge various inquiries and quotations that are not for the purpose of real transactions. This application solves the problem of low transaction efficiency caused by the inability to accurately evaluate the authenticity and transaction risks of transaction users in the prior art.

[0037] In order to enable those skilled in the art to more clearly understand the technical solution of the present application, the implementation process of the transaction authenticity risk assessment method of the present application will be described in detail below in conjunction with specific embodiments.

[0038] In order to improve the quality of transaction data, in an optional implementation, the above step S201 includes:

[0039] Step S2011, obtaining original transaction information data, wherein the original transaction information data is the transaction user information, the financial product element information, the financial product price index, the transaction data and the external data that have not been standardized;

[0040] Step S2012, performing a data cleaning operation on the original transaction information data to obtain clean transaction information data, wherein the data cleaning operation at least includes filling missing values ​​and removing outliers;

[0041] Step S2013, standardize the clean transaction information data to obtain the standardized transaction information data.

[0042] In the above embodiment, all relevant transaction information is collected, including transaction user information, financial product element information, financial product price index, transaction data and external data. These data may come from different sources, such as transaction records, user profiles, market data, etc. These data may contain inconsistencies, errors, missing values ​​or outliers when collected. Data cleaning is an important part of data preprocessing, the purpose of which is to improve data quality and make it more suitable for subsequent analysis and processing. Data cleaning includes at least the following two operations: Filling missing values: For missing data, a variety of methods can be used to fill, such as using the mean, median, mode or model-based predicted values. Removing outliers: Outliers may affect the results of data analysis and need to be identified and processed. Common methods include statistical methods (such as Z-score, IQR), visualization methods (such as box plots) or methods based on business logic. Standardization is the conversion of data into a unified format or scale for easy comparison and analysis. This usually involves the following operations: Normalization: Scale the data to a specific range, such as 0 to 1 or -1 to 1. Standardization: Convert the data to a distribution with a mean of 0 and a standard deviation of 1, which is usually achieved by subtracting the mean and dividing by the standard deviation. Encoding: For categorical data, one-hot encoding or label encoding may be required. It should be noted that the acquisition, storage, use, and processing of data in the embodiments of this application are in compliance with the relevant provisions of national laws and regulations.

[0043] In order to improve the efficiency of transaction risk assessment, in an optional implementation, after the above step S201, the method further includes:

[0044] Step S301, feature selection and classification are performed based on the above-mentioned standardized transaction information data to obtain multiple risk features, and a corresponding feature score is set for each risk feature. The above-mentioned risk features at least include transaction behavior features, geographic location features, financial product features and customer behavior features.

[0045] In the above embodiment, features related to the authenticity of the transaction are found by analyzing the data, and new features are constructed. For example: transaction frequency, transaction amount fluctuation, device usage frequency, etc. Secondly, the new features are classified into different feature classification libraries, including but not limited to transaction behavior features, geographic location features, device features, time features, and customer behavior features. Each feature corresponds to a score. Transaction behavior features: Extract transaction frequency, transaction amount size, transaction time pattern, etc. to analyze the user's transaction frequency, transaction amount fluctuation, transaction product type, transaction time interval, etc. These features can reflect the user's transaction habits and behavior patterns. Geographic location features: Extract the geographic location information where the transaction occurs, such as country, city, longitude and latitude, etc., extract the user's region, device usage location, etc., and combine the user's historical behavior to analyze whether the change in geographic location is abnormal. Financial product features: Extract the type of financial products involved in the transaction, interest rate, term, etc., analyze the element information of financial products, such as interest rate, price, term, etc., and compare these elements with market prices. Customer behavior features: Extract the customer's transaction history, credit score, account behavior, etc., including the customer's browsing history, historical transaction records, customer feedback, etc. These features help understand the customer's preferences and behavior. Based on the results of the above feature selection, new features are constructed, such as the rate of change of transaction frequency, the deviation rate of product price and market index, etc. These new features can capture transaction risks more accurately. The selected features are classified and a score is set for each feature. This score can be based on statistical analysis of historical data or set through expert experience. For example, if a transaction behavior feature has a strong correlation with historical fraud, the score of this feature may be higher. The automated feature selection and classification process can reduce the workload of manual analysis and improve the efficiency of transaction risk assessment. By identifying and quantifying risk features, potential risks can be discovered in advance, so that measures can be taken to reduce transaction risks.

[0046] In order to effectively identify the risks of transaction users, in an optional implementation, the above step S202 includes:

[0047] Step S2021, when the target risk assessment model is a convergence model obtained by iterative training using the supervised learning method, the product of the feature score corresponding to the relevant risk feature in the standardized transaction information data and the set weight coefficient is calculated to obtain the risk index of the transaction user risk, where the relevant risk feature is a feature in the risk feature that is relevant to the transaction user, and the risk feature at least includes the behavior feature, geographic location feature, device feature and time feature of the transaction user;

[0048] Step S2022, when the target risk assessment model is a convergence model obtained by iterative training using the unsupervised learning method, the risk information of the transaction user is obtained from the standardized transaction information data, and the risk information is identified as abnormal to obtain the risk index of the transaction user. The risk information at least includes the region where the transaction user is located, the product browsing records and transaction records of the transaction user within a set historical period.

[0049] In the above embodiment, the above risk type includes the above transaction user risk. If the supervised learning model type is selected, the risk index of each transaction user is obtained by multiplying the feature score corresponding to the feature related to the transaction user by the configured weight coefficient. If the unsupervised learning model is selected, the risk index is obtained by identifying abnormal data using specific transaction data. First, the target risk assessment model is iteratively trained using a supervised learning method until the model converges. In this process, the model learns the relationship between the behavior characteristics, geographic location characteristics, device characteristics and time characteristics of the transaction user and the risk index. The risk characteristics related to the transaction user are extracted from the standardized transaction information data, and these characteristics include at least the user's behavior characteristics, geographic location characteristics, device characteristics and time characteristics. The feature score corresponding to the extracted risk feature is multiplied by the preset weight coefficient to obtain the risk index of the transaction user. The weight coefficient is set based on historical data and expert knowledge to reflect the contribution of different features to the risk index. Through supervised learning, the model can more accurately predict the risk index of the transaction user because it can learn the relationship between the risk characteristics and the risk results from historical data. The target risk assessment model is iteratively trained using an unsupervised learning method until the model converges. Unsupervised learning does not require labeled data, and the model learns by identifying patterns and anomalies in the data. The risk information of trading users is obtained from standardized trading information data, including at least the user's region, product browsing history and transaction history within a set historical period. The acquired risk information is identified for anomalies, abnormal behaviors that are significantly different from the normal pattern are identified, and the risk index of the trading user is calculated. Unsupervised learning methods are good at discovering abnormal patterns in data, which is very useful for identifying potential abnormal trading activities. Unsupervised learning does not require labeled data, which means that trading data without clear risk labels can be processed, expanding the scope of application of the model. The risks of trading users are assessed using supervised learning and unsupervised learning methods. Both methods have their own advantages. The appropriate risk assessment model can be selected according to the actual situation and data availability to effectively identify and manage risks in the transaction process and improve the security and efficiency of financial transactions.

[0050] In order to improve transaction efficiency and security, in an optional implementation, the above step S2022 includes:

[0051] Step S20221, judging whether the region where the transaction user is located is marked as high risk according to the region where the transaction user is located in the risk information, whether the transaction user uses a VPN or proxy, and whether the geographical location of the transaction user deviates from the historical browsing access;

[0052] Step S20222, analyzing whether the number of small transactions exceeds the set number of transactions within the set time, or whether the total number of transactions within the set time exceeds the set number of transactions according to the transaction records in the risk information;

[0053] Step S20223, if the abnormal rules are met, it is determined that the transaction user has transaction abnormalities, and the abnormal rules are that the region where the transaction user is located is marked as high risk, the transaction user uses a VPN or proxy, the geographical location of the transaction user deviates from the historical browsing access, the number of small transactions within a set time exceeds the set number of transactions, and the total number of transactions within the set time exceeds the set number of transactions to meet the set number;

[0054] Step S20224: When the above-mentioned abnormal rules are not met, it is determined that there is no abnormal transaction problem for the above-mentioned transaction user.

[0055] In the above embodiment, when the unsupervised learning model is selected, the risk check of the transaction user is carried out through the region where the transaction user is located, the recent product browsing history and the recent transaction history. The region where the transaction user is located is mainly used to determine whether the region where the transaction user is located is marked with high-risk characteristics, whether VPN or proxy is used, and whether the user's geographical location deviates significantly from the region during the previous regular visit. The transaction history is mainly analyzed to determine whether there are frequent small transactions in a short period of time, or whether there is a sudden increase in transactions in the recent period. At the same time, the region, browsing history and transaction history of the transaction user are combined.

[0056] Identify whether there are potential risk factors in the behavior of trading users. By checking whether the geographical location of trading users is marked as a high-risk area and whether trading users use VPN or proxy services, the system can preliminarily determine the authenticity of transactions. In addition, by comparing the user's current geographical location with historical browsing access records, the system can detect abnormal changes in user behavior. Analyze the transaction frequency and amount of trading users within a set time. By checking whether too many small transactions have been made in a short period of time or the total number of transactions exceeds the normal range, the system can identify possible abnormal transaction patterns, effectively improving the authenticity assessment of financial transactions, reducing transaction risks, and improving transaction efficiency and security.

[0057] In order to improve the efficiency of financial product risk management, in an optional implementation, the above step S202 includes:

[0058] Step S2023, when the target risk assessment model is a convergence model obtained by iterative training using the supervised learning method, obtaining financial product features of the financial product and feature scores corresponding to the financial product features from the standardized transaction information data;

[0059] Step S2024, multiplying the feature score corresponding to each of the above-mentioned financial product features by the set weight coefficient to obtain the above-mentioned risk index of the above-mentioned financial product risk, wherein the above-mentioned financial product features at least include the rate of return, the reputation and qualifications of the issuing institution, the risk rating of the above-mentioned financial product, the market environment and the macroeconomic conditions, and the above-mentioned risk index of the above-mentioned financial product risk at least includes the capital cost risk index, the liquidity risk index and the price risk index of the financial product, and the above-mentioned liquidity risk index represents the risk index of the above-mentioned financial product in the market liquidity conditions;

[0060] Step S2025, when the above-mentioned target risk assessment model is a convergence model obtained by iterative training using the above-mentioned unsupervised learning method, the transaction data of the above-mentioned financial product is obtained from the above-mentioned standardized transaction information data, and risk assessment is performed based on the transaction data of the above-mentioned financial product to obtain the above-mentioned risk index of the risk of the above-mentioned financial product.

[0061] In the above embodiment, the financial product risk check determines whether the elements of the financial product are true and reliable through the price index of the financial product. The yield characteristic score promised by the financial product is divided by the market capital cost characteristic score to obtain the capital cost risk index score of the product. The market liquidity status characteristic score is evaluated by the liquidity price index, and the liquidity risk index of the financial product is obtained based on the configured weight coefficient. At the same time, the financial product price risk index of the financial product is obtained based on the credit and qualification characteristic score of the issuing institution, the risk rating characteristic score of the product, and the market environment and macroeconomic status characteristic score multiplied by the configured weight coefficient. If the unsupervised learning model is selected, the transaction data of the financial product is used to obtain the various risk indexes of the financial product. The specific features of the financial product are extracted from the standardized transaction information data. These features include the yield, the credit and qualification of the issuing institution, the risk rating of the financial product, the market environment and the macroeconomic conditions. For each extracted feature, the corresponding feature score is obtained. The feature score is a quantitative score for each feature based on historical data and model training results. The feature score corresponding to each financial product feature is multiplied by the pre-set weight coefficient. The weight coefficient is determined based on expert knowledge, historical data and risk preference, which reflects the importance of each feature in risk assessment. Through the above calculation, the risk index of financial products is obtained, including the capital cost risk index, liquidity risk index and price risk index. These indexes respectively represent the risk level of financial products in different aspects. The capital cost risk index reflects the risk level of the difference between the yield of financial products and the market capital cost. Liquidity risk index: a risk index that represents the market liquidity status of financial products, that is, the risk that financial products may face when market liquidity is limited. Price risk index: reflects the risk of price fluctuations of financial products. If the target risk assessment model is iteratively trained based on unsupervised learning, the transaction data of financial products is extracted from the standardized transaction information data. Using these transaction data, risk assessment is performed through unsupervised learning methods to identify abnormal transaction behaviors or patterns, and then the risk index of financial products is calculated. Through the quantitative scoring of financial product characteristics and the calculation of risk index, this solution can effectively identify the risk points of financial products and provide risk warnings for investors and financial institutions. The automated risk assessment process improves the efficiency of financial product risk management and reduces the impact of human errors and subjective judgments.

[0062] In order to improve the security and efficiency of financial transactions, in an optional implementation, the above step S203 includes:

[0063] Step S2031, when the risk type is the financial product risk, comparing the price risk index, the liquidity risk index, and the funding cost risk index of the financial product with the set index threshold;

[0064] Step S2032, in the case where the risk warning rule is triggered, the warning information is issued, and the risk warning rule is that one or two of the above-mentioned risk indexes among the above-mentioned price risk index, the above-mentioned liquidity risk index, and the above-mentioned capital cost risk index are higher than the first set index threshold;

[0065] Step S2033, terminating the current transaction when a prohibited transaction rule is triggered, wherein the prohibited transaction rule is that the price risk index, the liquidity risk index, and the funding cost risk index are all higher than the first set index threshold;

[0066] Step S2034, when the risk type is the transaction user risk, comparing the risk index corresponding to the transaction user risk with a second set index threshold;

[0067] Step S2035, when the risk index corresponding to the risk of the transaction user is higher than the second set index threshold, it is determined that the transaction user has the transaction credit risk and the prompt information is issued;

[0068] Step S2036, when the risk type is the quotation information risk, receiving the quotation information of the current financial product from the trading user, querying the current market price of the current financial product, and comparing the quotation information with the current market price;

[0069] Step S2037: When the absolute value of the difference between the quotation information and the current market price is greater than the price index threshold, it is determined that the authenticity of the current transaction is low, the prompt information is issued and the current transaction is terminated.

[0070] In the above embodiment, the risk prompt of trading products, when the user browses or applies for financial product information, the system compares the price risk index, liquidity risk index, and capital cost risk index of the financial product with the threshold and risk prompt rules configured by the user, and satisfies the triggering rules to give the user the corresponding risk prompt, prompting the user that the current financial product has certain risks, please be cautious in trading, if the risk index triggers the prohibition of trading rules (such as: higher than the set threshold), the current transaction cannot be carried out and the trader needs to apply to the superior. The risk prompt of quotation information, when the trader receives the inquiry, quotation, and bargaining information of the trading user, checks the current market price and compares the quotation of the counterparty with the current market price. If the quotation is significantly different from the current market price, especially far higher or far lower than the market level, then the judgment is unreliable, and subsequent transactions are not allowed to be applied to the superior. At this time, the above prompt information is issued and the above current transaction is terminated. In the case of identifying the risk of financial products, the system compares the price risk index, liquidity risk index, and capital cost risk index of the financial product with the pre-set index threshold. If one or two of the price risk index, liquidity risk index or funding cost risk index exceed the first set index threshold, the risk warning rule is triggered and the system will issue a risk warning message. Reduce bad trading decisions caused by information asymmetry and improve the security of transactions. If all three risk indexes (price, liquidity, funding cost) exceed the first set index threshold, the prohibition of trading rule is triggered, and the system will terminate the current transaction to prevent the execution of high-risk transactions, protect both parties from potential economic losses, effectively avoid the occurrence of high-risk transactions, and reduce the risk exposure of financial institutions. In the case of identifying the risk of the trading user, the system compares the risk index of the trading user with the second set index threshold, which is used to assess the credit risk of the trading user and identify possible fraud or default. By assessing the credit risk of the trading user, bad credit behavior can be prevented in advance to protect transaction security. If the risk index of the trading user exceeds the second set index threshold, it is determined that the trading user has a trading credit risk and a risk warning message is issued. Timely identification and warning of trading credit risks prompt traders or decision makers to take corresponding risk control measures. Reduce the risk of transaction failure caused by trading user credit problems and improve the reliability of transactions. When the quotation information risk is identified, the system will receive the quotation information of the current financial product from the trading user, query the market price of the current financial product, and then compare the quotation information with the market price. The purpose of this step is to evaluate the authenticity and rationality of the quotation and identify possible fraudulent quotations. By comparing the quotation and the market price, abnormal quotations can be discovered in time to avoid trading losses caused by false quotations.If the absolute value of the difference between the quotation information and the market price is greater than the price index threshold, the authenticity of the current transaction is judged to be low, a risk warning message is issued and the current transaction is terminated, preventing the execution of transactions that are obviously deviated from the market price, and protecting both parties from fraud or misleading. This improves the authenticity and transparency of transactions and reduces transaction disputes and losses caused by false quotations.

[0071] In order to improve subsequent transaction efficiency, in an optional implementation, after the above step S202, the method further includes:

[0072] Step S401, when the above-mentioned risk index is greater than or equal to the corresponding above-mentioned set index threshold, continue the current transaction until the transaction is completed, and correct the characteristic score of the risk characteristics in real time. The above-mentioned risk characteristics at least include transaction behavior characteristics, geographic location characteristics, financial product characteristics and customer behavior characteristics.

[0073] In the above embodiment, when the absolute value of the difference between the above quotation information and the above current market price is less than or equal to the above price index threshold, it is determined that the authenticity of the current transaction is high and the above current transaction is continued. When the above risk index corresponding to the risk of the above transaction user is less than or equal to the second set index threshold, it is determined that the above transaction user has no transaction credit risk and the above current transaction is continued. After the transaction parties reach a transaction through bargaining in the system, the system will extract the user information, financial product element information, price index of financial products, and transaction data in this transaction, and update the price index feature score of the financial product and the user's transaction risk-related feature score as part of the model input parameters. By summarizing the transaction prices of financial products in real time and correcting the feature scores of various risk characteristics of the financial market, it is more efficient to judge various inquiries and quotations that are not for the purpose of real transactions, reduce the ineffective communication costs of traders, and improve transaction efficiency.

[0074] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0075] The embodiment of the present application also provides a transaction authenticity risk assessment device. It should be noted that the transaction authenticity risk assessment device of the embodiment of the present application can be used to execute the transaction authenticity risk assessment method provided by the embodiment of the present application. The device is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware for a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0076] The following is an introduction to the transaction authenticity risk assessment device provided in an embodiment of the present application.

[0077] Figure 3 is a structural block diagram of a transaction authenticity risk assessment device according to an embodiment of the present application. Figure 3 As shown, the device comprises:

[0078] The acquisition unit 10 is used to acquire standardized transaction information data, wherein the standardized transaction information data at least includes standardized transaction user information, financial product element information, financial product price index, transaction data and external data, and the financial product price index at least includes Shibor interest rate, DR interest rate and recent transaction records of corresponding financial products.

[0079] Specifically, data collection includes user information, financial product element information, price index of financial products, transaction data, and external data. Customer data includes basic customer information, user region, recent product browsing records, and historical transaction records; financial product element information includes interest rate, price, term, amount, counterparty category, etc.; financial product price index includes: Shibor interest rate, DR interest rate and recent transaction records of corresponding products; transaction data includes transaction amount, transaction time, transaction product, counterparty information, etc. External data mainly includes customer evaluation data within the bank, etc.

[0080] The input-output unit 20 is used to input the above-mentioned standardized transaction information data into the target risk assessment model, and output a risk index corresponding to the risk type. One of the above-mentioned risk types corresponds to at least one of the above-mentioned risk indexes. The above-mentioned target risk assessment model is a convergence model obtained by iterative training using multiple groups of training data using an unsupervised learning method or a supervised learning method. Each group of the above-mentioned training data includes sample standardized transaction data and a corresponding sample risk index. The above-mentioned risk types include at least one of the following: transaction user risk, financial product risk, transaction product risk and quotation information risk. The above-mentioned transaction user risk is the credit risk of the user who initiates the transaction application.

[0081] Specifically, it supports supervised learning models and unsupervised learning models. Supervised learning models require labeled data for training. Unsupervised learning models are suitable for risk assessment by identifying abnormal behaviors without clear labels. According to the needs of risk assessment, select appropriate unsupervised learning (such as cluster analysis) or supervised learning (such as logistic regression, decision tree, neural network) models. The collected data is cleaned, normalized, feature engineered, etc. to meet the input requirements of the model. Use multiple sets of training data (including sample standardized transaction data and corresponding sample risk index) for iterative training until the model converges. Input the standardized transaction information data into the target risk assessment model and output the corresponding risk index.

[0082] The issuing unit 30 is used to issue a prompt message when the risk index is greater than the corresponding set index threshold, and the prompt message is used to prompt that the current transaction has risks.

[0083] Specifically, a risk index threshold is set for each risk type. When the risk index exceeds the set threshold, the system automatically issues a prompt message to indicate that the current transaction is risky.

[0084] In this embodiment, an acquisition unit is used to acquire standardized transaction information data, wherein the standardized transaction information data at least includes standardized transaction user information, financial product element information, financial product price index, transaction data and external data, wherein the financial product price index at least includes Shibor interest rate, DR interest rate and recent transaction records of corresponding financial products; an input-output unit is used to input the standardized transaction information data into a target risk assessment model, and output a risk index corresponding to a risk type, wherein one risk type corresponds to at least one risk index, and the target risk assessment model is a convergence model obtained by iterative training using multiple sets of training data using an unsupervised learning method or a supervised learning method, wherein each set of the training data includes sample standardized transaction data and a corresponding sample risk index, and the risk type includes at least one of the following: transaction user risk, financial product risk, transaction product risk and quotation information risk; a prompt unit is used to issue a prompt message when the risk index is greater than the corresponding set index threshold, and the prompt message is used to prompt that there is a risk in the current transaction and the current transaction needs to be terminated. This application analyzes the behavioral data of transaction users and the key elements of financial products, and analyzes them in combination with various price indices in the financial market, inputs various transaction data into the target risk assessment model, obtains a risk index, and compares the risk index with the corresponding risk threshold to provide risk warnings for the interbank financial transaction process and avoid leakage of transaction elements. The transaction authenticity of transaction users is evaluated and risk warnings are provided to help traders reduce communication costs, reduce transaction risks, improve transaction efficiency, and more efficiently judge various inquiries and quotations that are not for the purpose of real transactions. This application solves the problem of low transaction efficiency caused by the inability to accurately evaluate the authenticity and transaction risks of transaction users in the prior art.

[0085] In order to improve the quality of transaction data, in an optional implementation manner, the acquisition unit includes:

[0086] A first acquisition module is used to acquire original transaction information data, where the original transaction information data is the transaction user information, the financial product element information, the financial product price index, the transaction data and the external data that have not been standardized;

[0087] A data cleaning module, used to perform a data cleaning operation on the original transaction information data to obtain clean transaction information data, wherein the data cleaning operation at least includes an operation of filling missing values ​​and removing outliers;

[0088] The standardization module is used to perform standardization processing on the above-mentioned clean transaction information data to obtain the above-mentioned standardized transaction information data.

[0089] In the above embodiment, all relevant transaction information is collected, including transaction user information, financial product element information, financial product price index, transaction data and external data. These data may come from different sources, such as transaction records, user profiles, market data, etc. These data may contain inconsistencies, errors, missing values ​​or outliers when collected. Data cleaning is an important part of data preprocessing, the purpose of which is to improve data quality and make it more suitable for subsequent analysis and processing. Data cleaning includes at least the following two operations: Filling missing values: For missing data, a variety of methods can be used to fill, such as using the mean, median, mode or model-based predicted values. Removing outliers: Outliers may affect the results of data analysis and need to be identified and processed. Common methods include statistical methods (such as Z-score, IQR), visualization methods (such as box plots) or methods based on business logic. Standardization is the conversion of data into a unified format or scale for easy comparison and analysis. This usually involves the following operations: Normalization: Scale the data to a specific range, such as 0 to 1 or -1 to 1. Standardization: Convert the data to a distribution with a mean of 0 and a standard deviation of 1, which is usually achieved by subtracting the mean and dividing by the standard deviation. Encoding: For categorical data, one-hot encoding or label encoding may be required. It should be noted that the acquisition, storage, use, and processing of data in the embodiments of this application are in compliance with the relevant provisions of national laws and regulations.

[0090] In order to improve the efficiency of transaction risk assessment, in an optional implementation manner, the device further includes:

[0091] The feature selection unit is used to select and classify features based on the standardized transaction information data after obtaining the standardized transaction information data, obtain multiple risk features, and set a corresponding feature score for each risk feature. The above risk features include at least transaction behavior features, geographic location features, financial product features and customer behavior features.

[0092] In the above embodiment, features related to the authenticity of the transaction are found by analyzing the data, and new features are constructed. For example: transaction frequency, transaction amount fluctuation, device usage frequency, etc. Secondly, the new features are classified into different feature classification libraries, including but not limited to transaction behavior features, geographic location features, device features, time features, and customer behavior features. Each feature corresponds to a score. Transaction behavior features: Extract transaction frequency, transaction amount size, transaction time pattern, etc. to analyze the user's transaction frequency, transaction amount fluctuation, transaction product type, transaction time interval, etc. These features can reflect the user's transaction habits and behavior patterns. Geographic location features: Extract the geographic location information where the transaction occurs, such as country, city, longitude and latitude, etc., extract the user's region, device usage location, etc., and combine the user's historical behavior to analyze whether the change in geographic location is abnormal. Financial product features: Extract the type of financial products involved in the transaction, interest rate, term, etc., analyze the element information of financial products, such as interest rate, price, term, etc., and compare these elements with market prices. Customer behavior features: Extract the customer's transaction history, credit score, account behavior, etc., including the customer's browsing history, historical transaction records, customer feedback, etc. These features help understand the customer's preferences and behavior. Based on the results of the above feature selection, new features are constructed, such as the rate of change of transaction frequency, the deviation rate of product price and market index, etc. These new features can capture transaction risks more accurately. The selected features are classified and a score is set for each feature. This score can be based on statistical analysis of historical data or set through expert experience. For example, if a transaction behavior feature has a strong correlation with historical fraud, the score of this feature may be higher. The automated feature selection and classification process can reduce the workload of manual analysis and improve the efficiency of transaction risk assessment. By identifying and quantifying risk features, potential risks can be discovered in advance, so that measures can be taken to reduce transaction risks.

[0093] In order to effectively identify the risks of transaction users, in an optional implementation, the input and output unit includes:

[0094] A first calculation module is used to calculate the product between the feature score corresponding to the relevant risk feature in the standardized transaction information data and the set weight coefficient to obtain the risk index of the transaction user risk, when the target risk assessment model is a convergence model obtained by iterative training using the supervised learning method, wherein the relevant risk feature is a feature in the risk feature that is relevant to the transaction user, and the risk feature at least includes the behavior feature, geographic location feature, device feature and time feature of the transaction user;

[0095] The anomaly identification module is used to obtain the risk information of the above-mentioned transaction user from the above-mentioned standardized transaction information data, and to identify anomalies in the above-mentioned risk information, so as to obtain the above-mentioned risk index of the risk of the above-mentioned transaction user, when the above-mentioned target risk assessment model is a convergence model obtained by iterative training using the above-mentioned unsupervised learning method. The above-mentioned risk information at least includes the region where the above-mentioned transaction user is located, and the product browsing records and transaction records of the above-mentioned transaction user within a set historical period.

[0096] In the above embodiment, the above risk type includes the above transaction user risk. If the supervised learning model type is selected, the risk index of each transaction user is obtained by multiplying the feature score corresponding to the feature related to the transaction user by the configured weight coefficient. If the unsupervised learning model is selected, the risk index is obtained by identifying abnormal data using specific transaction data. First, the target risk assessment model is iteratively trained using a supervised learning method until the model converges. In this process, the model learns the relationship between the behavior characteristics, geographic location characteristics, device characteristics and time characteristics of the transaction user and the risk index. The risk characteristics related to the transaction user are extracted from the standardized transaction information data, and these characteristics include at least the user's behavior characteristics, geographic location characteristics, device characteristics and time characteristics. The feature score corresponding to the extracted risk feature is multiplied by the preset weight coefficient to obtain the risk index of the transaction user. The weight coefficient is set based on historical data and expert knowledge to reflect the contribution of different features to the risk index. Through supervised learning, the model can more accurately predict the risk index of the transaction user because it can learn the relationship between the risk characteristics and the risk results from historical data. The target risk assessment model is iteratively trained using an unsupervised learning method until the model converges. Unsupervised learning does not require labeled data, and the model learns by identifying patterns and anomalies in the data. The risk information of trading users is obtained from standardized trading information data, including at least the user's region, product browsing history and transaction history within a set historical period. The acquired risk information is identified for anomalies, abnormal behaviors that are significantly different from the normal pattern are identified, and the risk index of the trading user is calculated. Unsupervised learning methods are good at discovering abnormal patterns in data, which is very useful for identifying potential abnormal trading activities. Unsupervised learning does not require labeled data, which means that trading data without clear risk labels can be processed, expanding the scope of application of the model. The risks of trading users are assessed using supervised learning and unsupervised learning methods. Both methods have their own advantages. The appropriate risk assessment model can be selected according to the actual situation and data availability to effectively identify and manage risks in the transaction process and improve the security and efficiency of financial transactions.

[0097] In order to improve transaction efficiency and security, in an optional implementation, the above-mentioned abnormality identification module includes:

[0098] A judgment submodule, used to judge whether the region where the transaction user is located is marked as high risk according to the region where the transaction user is located in the risk information, whether the transaction user uses a VPN or proxy, and whether the geographical location of the transaction user deviates from the historical browsing access;

[0099] An analysis submodule, for analyzing whether the number of small transactions exceeds the set number of transactions within the set time, or whether the total number of transactions within the set time exceeds the set number of transactions, based on the transaction records in the risk information;

[0100] The first determination submodule is used to determine that the transaction user has transaction abnormality problems when the abnormal rules are met. The abnormal rules are that the region where the transaction user is located is marked as high risk, the transaction user uses a VPN or proxy, the geographical location of the transaction user deviates from the historical browsing access, the number of small transactions exceeds the set number of transactions within the set time, and the total number of transactions within the set time exceeds the set number of transactions to meet the set number;

[0101] The second determination submodule is used to determine that the transaction user does not have any transaction abnormality problem when the abnormal rule is not satisfied.

[0102] In the above embodiment, when the unsupervised learning model is selected, the risk check of the transaction user is mainly to determine whether the region where the transaction user is located is marked with high-risk features, whether VPN or proxy is used, and whether the user's geographical location deviates significantly from the region during the previous regular visit through the region where the transaction user is located. For the transaction record, it is mainly to analyze whether there are frequent small transactions in a short period of time, or whether there is a sudden increase in transactions in the recent period. At the same time, the region, browsing history, and transaction record of the transaction user will be combined. Identify whether there are potential risk factors in the behavior of the transaction user. By checking whether the geographical location of the transaction user is marked as a high-risk area, and whether the transaction user uses VPN or proxy service, the system can preliminarily judge the authenticity of the transaction. In addition, by comparing the user's current geographical location with the historical browsing access record, the system can detect abnormal changes in user behavior. Analyze the transaction frequency and amount of the transaction user within the set time. By checking whether too many small transactions have been made in a short period of time or the total number of transactions exceeds the normal range, the system can identify possible abnormal transaction patterns, effectively improve the authenticity assessment of financial transactions, reduce transaction risks, and improve transaction efficiency and security.

[0103] In order to improve the efficiency of financial product risk management, in an optional implementation, the input and output unit includes:

[0104] A second acquisition module, when the target risk assessment model is a convergence model obtained by iterative training using the supervised learning method, acquires financial product features of the financial product and feature scores corresponding to the financial product features from the standardized transaction information data;

[0105] A second calculation module is used to multiply the feature score corresponding to each of the above-mentioned financial product features by a set weight coefficient to obtain the above-mentioned risk index of the above-mentioned financial product risk, wherein the above-mentioned financial product features at least include yield, the reputation and qualifications of the issuing institution, the risk rating of the above-mentioned financial product, the market environment and the macroeconomic conditions, and the above-mentioned risk index of the above-mentioned financial product risk at least includes the capital cost risk index, the liquidity risk index and the price risk index of the financial product, and the above-mentioned liquidity risk index represents the risk index of the above-mentioned financial product in the market liquidity conditions;

[0106] An evaluation module is used to obtain the transaction data of the above-mentioned financial products from the above-mentioned standardized transaction information data when the above-mentioned target risk assessment model is a convergence model obtained by iterative training using the above-mentioned unsupervised learning method, and to perform risk assessment based on the transaction data of the above-mentioned financial products to obtain the above-mentioned risk index of the risk of the above-mentioned financial product.

[0107] In the above embodiment, the financial product risk check determines whether the elements of the financial product are true and reliable through the price index of the financial product. The yield characteristic score promised by the financial product is divided by the market capital cost characteristic score to obtain the capital cost risk index score of the product. The market liquidity status characteristic score is evaluated by the liquidity price index, and the liquidity risk index of the financial product is obtained based on the configured weight coefficient. At the same time, the financial product price risk index of the financial product is obtained based on the credit and qualification characteristic score of the issuing institution, the risk rating characteristic score of the product, and the market environment and macroeconomic status characteristic score multiplied by the configured weight coefficient. If the unsupervised learning model is selected, the transaction data of the financial product is used to obtain the various risk indexes of the financial product. The specific features of the financial product are extracted from the standardized transaction information data. These features include the yield, the credit and qualification of the issuing institution, the risk rating of the financial product, the market environment and the macroeconomic conditions. For each extracted feature, the corresponding feature score is obtained. The feature score is a quantitative score for each feature based on historical data and model training results. The feature score corresponding to each financial product feature is multiplied by the pre-set weight coefficient. The weight coefficient is determined based on expert knowledge, historical data and risk preference, which reflects the importance of each feature in risk assessment. Through the above calculation, the risk index of financial products is obtained, including the capital cost risk index, liquidity risk index and price risk index. These indexes respectively represent the risk level of financial products in different aspects. The capital cost risk index reflects the risk level of the difference between the yield of financial products and the market capital cost. Liquidity risk index: a risk index that represents the market liquidity status of financial products, that is, the risk that financial products may face when market liquidity is limited. Price risk index: reflects the risk of price fluctuations of financial products. If the target risk assessment model is iteratively trained based on unsupervised learning, the transaction data of financial products is extracted from the standardized transaction information data. Using these transaction data, risk assessment is performed through unsupervised learning methods to identify abnormal transaction behaviors or patterns, and then the risk index of financial products is calculated. Through the quantitative scoring of financial product characteristics and the calculation of risk index, this solution can effectively identify the risk points of financial products and provide risk warnings for investors and financial institutions. The automated risk assessment process improves the efficiency of financial product risk management and reduces the impact of human errors and subjective judgments.

[0108] In order to improve the security and efficiency of financial transactions, in an optional implementation manner, the issuing unit includes:

[0109] A first comparison module is used to compare the price risk index, the liquidity risk index, and the capital cost risk index of the financial product with the set index threshold when the risk type is the financial product risk;

[0110] An issuing module, used for issuing the above-mentioned prompt information in the case where a risk warning rule is triggered, wherein the above-mentioned risk warning rule is that one or two of the above-mentioned risk indexes among the above-mentioned price risk index, the above-mentioned liquidity risk index, and the above-mentioned capital cost risk index are higher than a first set index threshold;

[0111] A termination module, used to terminate the current transaction when a prohibited transaction rule is triggered, wherein the prohibited transaction rule is that the price risk index, the liquidity risk index, and the funding cost risk index are all higher than the first set index threshold;

[0112] A second comparison module is used for comparing the risk index corresponding to the transaction user risk with a second set index threshold when the risk type is the transaction user risk;

[0113] A first determination module is used to determine that the transaction user has the transaction credit risk and issue the prompt information when the risk index corresponding to the transaction user risk is higher than the second set index threshold;

[0114] A comparison module, for receiving the quotation information of the current financial product from the trading user when the risk type is the quotation information risk, querying the current market price of the current financial product, and comparing the quotation information with the current market price;

[0115] The second determination module is used to determine that the authenticity of the current transaction is low, issue the prompt information and terminate the current transaction when the absolute value of the difference between the quotation information and the current market price is greater than the price index threshold.

[0116] In the above embodiment, the risk prompt of trading products, when the user browses or applies for financial product information, the system compares the price risk index, liquidity risk index, and capital cost risk index of the financial product with the threshold and risk prompt rules configured by the user, and satisfies the triggering rules to give the user the corresponding risk prompt, prompting the user that the current financial product has certain risks, please be cautious in trading, if the risk index triggers the prohibition of trading rules (such as: higher than the set threshold), the current transaction cannot be carried out and the trader needs to apply to the superior. The risk prompt of quotation information, when the trader receives the inquiry, quotation, and bargaining information of the trading user, checks the current market price and compares the quotation of the counterparty with the current market price. If the quotation is significantly different from the current market price, especially far higher or far lower than the market level, then the judgment is unreliable, and subsequent transactions are not allowed to be applied to the superior. At this time, the above prompt information is issued and the above current transaction is terminated. In the case of identifying the risk of financial products, the system compares the price risk index, liquidity risk index, and capital cost risk index of the financial product with the pre-set index threshold. If one or two of the price risk index, liquidity risk index or funding cost risk index exceed the first set index threshold, the risk warning rule is triggered and the system will issue a risk warning message. Reduce bad trading decisions caused by information asymmetry and improve the security of transactions. If all three risk indexes (price, liquidity, funding cost) exceed the first set index threshold, the prohibition of trading rule is triggered, and the system will terminate the current transaction to prevent the execution of high-risk transactions, protect both parties from potential economic losses, effectively avoid the occurrence of high-risk transactions, and reduce the risk exposure of financial institutions. In the case of identifying the risk of the trading user, the system compares the risk index of the trading user with the second set index threshold, which is used to assess the credit risk of the trading user and identify possible fraud or default. By assessing the credit risk of the trading user, bad credit behavior can be prevented in advance to protect transaction security. If the risk index of the trading user exceeds the second set index threshold, it is determined that the trading user has a trading credit risk and a risk warning message is issued. Timely identification and warning of trading credit risks prompt traders or decision makers to take corresponding risk control measures. Reduce the risk of transaction failure caused by trading user credit problems and improve the reliability of transactions. When the quotation information risk is identified, the system will receive the quotation information of the current financial product from the trading user, query the market price of the current financial product, and then compare the quotation information with the market price. The purpose of this step is to evaluate the authenticity and rationality of the quotation and identify possible fraudulent quotations. By comparing the quotation and the market price, abnormal quotations can be discovered in time to avoid trading losses caused by false quotations.If the absolute value of the difference between the quotation information and the market price is greater than the price index threshold, the authenticity of the current transaction is judged to be low, a risk warning message is issued and the current transaction is terminated, preventing the execution of transactions that are obviously deviated from the market price, and protecting both parties from fraud or misleading. This improves the authenticity and transparency of transactions and reduces transaction disputes and losses caused by false quotations.

[0117] In order to improve subsequent transaction efficiency, in an optional implementation manner, the device further includes:

[0118] The correction unit is used to input the above-mentioned standardized transaction information data into the target risk assessment model and output the risk index corresponding to the risk type. If the above-mentioned risk index is greater than or equal to the corresponding above-mentioned set index threshold, continue the current transaction until the transaction is completed, and correct the characteristic score of the risk characteristics in real time. The above-mentioned risk characteristics at least include transaction behavior characteristics, geographic location characteristics, financial product characteristics and customer behavior characteristics.

[0119] In the above embodiment, when the absolute value of the difference between the above quotation information and the above current market price is less than or equal to the above price index threshold, it is determined that the authenticity of the current transaction is high and the above current transaction is continued. When the above risk index corresponding to the risk of the above transaction user is less than or equal to the second set index threshold, it is determined that the above transaction user has no transaction credit risk and the above current transaction is continued. After the transaction parties reach a transaction through bargaining in the system, the system will extract the user information, financial product element information, price index of financial products, and transaction data in this transaction, and update the price index feature score of the financial product and the user's transaction risk-related feature score as part of the model input parameters. By summarizing the transaction prices of financial products in real time and correcting the feature scores of various risk characteristics of the financial market, it is more efficient to judge various inquiries and quotations that are not for the purpose of real transactions, reduce the ineffective communication costs of traders, and improve transaction efficiency.

[0120] The transaction authenticity risk assessment device includes a processor and a memory. The acquisition unit, input / output unit, and issuing unit are stored in the memory as program units, and the processor executes the program units stored in the memory to implement corresponding functions. The modules are all located in the same processor; or, the modules are located in different processors in any combination.

[0121] The processor includes a kernel, which retrieves the corresponding program unit from the memory. One or more kernels can be set, and the problem of low transaction efficiency caused by the inability to accurately assess the authenticity of transaction users and transaction risks in the prior art can be solved by adjusting kernel parameters.

[0122] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0123] An embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the transaction authenticity risk assessment method.

[0124] Specifically, transaction authenticity risk assessment methods include:

[0125] Step S201, obtaining standardized transaction information data, the standardized transaction information data at least includes standardized transaction user information, financial product element information, financial product price index, transaction data and external data, the financial product price index at least includes Shibor interest rate, DR interest rate and recent transaction records of corresponding financial products;

[0126] Step S202, input the above-mentioned standardized transaction information data into the target risk assessment model, and output a risk index corresponding to the risk type. One of the above-mentioned risk types corresponds to at least one of the above-mentioned risk indexes. The above-mentioned target risk assessment model is a convergence model obtained by iterative training using multiple sets of training data using an unsupervised learning method or a supervised learning method. Each set of the above-mentioned training data includes sample standardized transaction data and a corresponding sample risk index. The above-mentioned risk type includes at least one of the following: transaction user risk, financial product risk, transaction product risk and quotation information risk. The above-mentioned transaction user risk is the credit risk of the user who initiates the transaction application;

[0127] Step S203: When the risk index is greater than the corresponding set index threshold, a prompt message is issued, wherein the prompt message is used to indicate that the current transaction has risks.

[0128] An embodiment of the present invention provides a processor, and the processor is used to run a program, wherein the transaction authenticity risk assessment method is executed when the program is running.

[0129] Specifically, transaction authenticity risk assessment methods include:

[0130] Step S201, obtaining standardized transaction information data, the standardized transaction information data at least includes standardized transaction user information, financial product element information, financial product price index, transaction data and external data, the financial product price index at least includes Shibor interest rate, DR interest rate and recent transaction records of corresponding financial products;

[0131] Step S202, input the above-mentioned standardized transaction information data into the target risk assessment model, and output a risk index corresponding to the risk type. One of the above-mentioned risk types corresponds to at least one of the above-mentioned risk indexes. The above-mentioned target risk assessment model is a convergence model obtained by iterative training using multiple sets of training data using an unsupervised learning method or a supervised learning method. Each set of the above-mentioned training data includes sample standardized transaction data and a corresponding sample risk index. The above-mentioned risk type includes at least one of the following: transaction user risk, financial product risk, transaction product risk and quotation information risk. The above-mentioned transaction user risk is the credit risk of the user who initiates the transaction application;

[0132] Step S203: When the risk index is greater than the corresponding set index threshold, a prompt message is issued, wherein the prompt message is used to indicate that the current transaction has risks.

[0133] An embodiment of the present invention provides a transaction authenticity risk assessment system, which includes a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, at least the following steps are implemented:

[0134] Step S201, obtaining standardized transaction information data, the standardized transaction information data at least includes standardized transaction user information, financial product element information, financial product price index, transaction data and external data, the financial product price index at least includes Shibor interest rate, DR interest rate and recent transaction records of corresponding financial products;

[0135] Step S202, input the above-mentioned standardized transaction information data into the target risk assessment model, and output a risk index corresponding to the risk type. One of the above-mentioned risk types corresponds to at least one of the above-mentioned risk indexes. The above-mentioned target risk assessment model is a convergence model obtained by iterative training using multiple sets of training data using an unsupervised learning method or a supervised learning method. Each set of the above-mentioned training data includes sample standardized transaction data and a corresponding sample risk index. The above-mentioned risk type includes at least one of the following: transaction user risk, financial product risk, transaction product risk and quotation information risk. The above-mentioned transaction user risk is the credit risk of the user who initiates the transaction application;

[0136] Step S203: When the risk index is greater than the corresponding set index threshold, a prompt message is issued, wherein the prompt message is used to indicate that the current transaction has risks.

[0137] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program for initializing at least the following method steps:

[0138] Step S201, obtaining standardized transaction information data, the standardized transaction information data at least includes standardized transaction user information, financial product element information, financial product price index, transaction data and external data, the financial product price index at least includes Shibor interest rate, DR interest rate and recent transaction records of corresponding financial products;

[0139] Step S202, input the above-mentioned standardized transaction information data into the target risk assessment model, and output a risk index corresponding to the risk type. One of the above-mentioned risk types corresponds to at least one of the above-mentioned risk indexes. The above-mentioned target risk assessment model is a convergence model obtained by iterative training using multiple sets of training data using an unsupervised learning method or a supervised learning method. Each set of the above-mentioned training data includes sample standardized transaction data and a corresponding sample risk index. The above-mentioned risk type includes at least one of the following: transaction user risk, financial product risk, transaction product risk and quotation information risk. The above-mentioned transaction user risk is the credit risk of the user who initiates the transaction application;

[0140] Step S203: When the risk index is greater than the corresponding set index threshold, a prompt message is issued, wherein the prompt message is used to indicate that the current transaction has risks.

[0141] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0142] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0143] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0144] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0145] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0146] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0147] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash memory (lash RAM). The memory is an example of a computer-readable medium.

[0148] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0149] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0150] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:

[0151] 1) The transaction authenticity risk assessment method of the present application, first, obtain standardized transaction information data, the above-mentioned standardized transaction information data at least includes standardized transaction user information, financial product element information, financial product price index, transaction data and external data, the above-mentioned financial product price index at least includes Shibor interest rate, DR interest rate and recent transaction records of corresponding financial products; then, input the above-mentioned standardized transaction information data into the target risk assessment model, and output the risk index corresponding to the risk type, one of the above-mentioned risk types corresponds to at least one of the above-mentioned risk indexes, the above-mentioned target risk assessment model is a convergence model obtained by iterative training using multiple groups of training data using an unsupervised learning method or a supervised learning method, each group of the above-mentioned training data includes sample standardized transaction data and a corresponding sample risk index, the above-mentioned risk type includes at least one of the following: transaction user risk, financial product risk, transaction product risk and quotation information risk, the above-mentioned transaction user risk is the credit risk of the user who initiates the transaction application; finally, when the above-mentioned risk index is less than or greater than the corresponding set index threshold, a prompt message is issued, and the above-mentioned prompt message is used to prompt that there is a risk in the current transaction. This application analyzes the behavioral data of transaction users and the key elements of financial products, and analyzes them in combination with various price indices in the financial market, inputs various transaction data into the target risk assessment model, obtains a risk index, and compares the risk index with the corresponding risk threshold to provide risk warnings for the interbank financial transaction process and avoid leakage of transaction elements. The transaction authenticity of transaction users is evaluated and risk warnings are provided to help traders reduce communication costs, reduce transaction risks, improve transaction efficiency, and more efficiently judge various inquiries and quotations that are not for the purpose of real transactions. This application solves the problem of low transaction efficiency caused by the inability to accurately evaluate the authenticity and transaction risks of transaction users in the prior art.

[0152] 2) The transaction authenticity risk assessment device of the present application comprises an acquisition unit for acquiring standardized transaction information data, wherein the standardized transaction information data at least comprises standardized transaction user information, financial product element information, financial product price index, transaction data and external data, wherein the financial product price index at least comprises Shibor interest rate, DR interest rate and recent transaction records of corresponding financial products; an input-output unit for inputting the standardized transaction information data into a target risk assessment model, and outputting a risk index corresponding to a risk type, wherein one risk type corresponds to at least one risk index, and the target risk assessment model is a convergence model obtained by iteratively training with multiple sets of training data using an unsupervised learning method or a supervised learning method, wherein each set of the training data comprises sample standardized transaction data and a corresponding sample risk index, and the risk type comprises at least one of the following: transaction user risk, financial product risk, transaction product risk and quotation information risk; a prompt unit for issuing a prompt message when the risk index is greater than the corresponding set index threshold, and the prompt message is used to prompt that the current transaction has a risk and needs to terminate the current transaction. This application analyzes the behavioral data of transaction users and the key elements of financial products, and analyzes them in combination with various price indices in the financial market, inputs various transaction data into the target risk assessment model, obtains a risk index, and compares the risk index with the corresponding risk threshold to provide risk warnings for the interbank financial transaction process and avoid leakage of transaction elements. The transaction authenticity of transaction users is evaluated and risk warnings are provided to help traders reduce communication costs, reduce transaction risks, improve transaction efficiency, and more efficiently judge various inquiries and quotations that are not for the purpose of real transactions. This application solves the problem of low transaction efficiency caused by the inability to accurately evaluate the authenticity and transaction risks of transaction users in the prior art.

[0153] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A transaction authenticity risk assessment method, characterized in that: include: Obtaining standardized transaction information data, wherein the standardized transaction information data at least includes standardized transaction user information, financial product element information, financial product price index, transaction data and external data, wherein the financial product price index at least includes Shibor interest rate, DR interest rate and recent transaction records of corresponding financial products; The standardized transaction information data is input into a target risk assessment model, and a risk index corresponding to a risk type is outputted. One risk type corresponds to at least one risk index. The target risk assessment model is a convergence model obtained by iterative training using multiple sets of training data using an unsupervised learning method or a supervised learning method. Each set of training data includes sample standardized transaction data and a corresponding sample risk index. The risk type includes at least one of the following: transaction user risk, financial product risk, transaction product risk, and quotation information risk. The transaction user risk is the credit risk of the user who initiates the transaction application. When the risk index is greater than the corresponding set index threshold, a prompt message is issued, and the prompt message is used to prompt that there is a risk in the current transaction.

2. The method according to claim 1, characterized in that Obtain standardized transaction information data, including: Acquire original transaction information data, wherein the original transaction information data is the transaction user information, the financial product element information, the financial product price index, the transaction data, and the external data that have not been standardized; Performing a data cleaning operation on the original transaction information data to obtain clean transaction information data, wherein the data cleaning operation at least includes an operation of filling missing values ​​and removing outliers; The clean transaction information data is standardized to obtain the standardized transaction information data.

3. The method according to claim 1, characterized in that After obtaining the standardized transaction information data, the method further includes: Feature selection and classification are performed based on the standardized transaction information data to obtain multiple risk features, and a corresponding feature score is set for each risk feature. The risk features at least include transaction behavior features, geographic location features, financial product features, and customer behavior features.

4. The method according to claim 1, characterized in that: The risk type includes the transaction user risk, and the standardized transaction information data is input into the target risk assessment model, and the risk index corresponding to the risk type is output, including: In the case where the target risk assessment model is a convergence model obtained by iterative training using the supervised learning method, the product between the feature score corresponding to the relevant risk feature in the standardized transaction information data and the set weight coefficient is calculated to obtain the risk index of the transaction user risk, wherein the relevant risk feature is a feature in the risk feature that is relevant to the transaction user, and the risk feature at least includes the behavior feature, geographic location feature, device feature and time feature of the transaction user; In the case where the target risk assessment model is a convergent model obtained by iterative training using the unsupervised learning method, the risk information of the transaction user is obtained from the standardized transaction information data, and the risk information is identified for abnormalities to obtain the risk index of the transaction user's risk, and the risk information at least includes the region where the transaction user is located, and the product browsing records and transaction records of the transaction user within a set historical period.

5. The method according to claim 4, characterized in that Performing abnormal identification on the risk information includes: Determine whether the region where the transaction user is located is marked as high risk according to the region where the transaction user is located in the risk information, whether the transaction user uses a VPN or proxy, and whether the geographical location of the transaction user deviates from the historical browsing access; Analyze whether small transactions exceed the set number of transactions within the set time, or whether the total number of transactions within the set time exceeds the set number of transactions, based on the transaction records in the risk information; In the case where the abnormal rules are met, it is determined that the transaction user has transaction abnormalities, and the abnormal rules are that the region where the transaction user is located is marked as high risk, the transaction user uses a VPN or proxy, the transaction user's geographical location deviates from the historical browsing access, the number of small transactions within a set time exceeds the set number of transactions, and the total number of transactions within the set time exceeds the set number of transactions to meet the set number; When the abnormal rule is not satisfied, it is determined that the transaction user has no transaction abnormality problem.

6. The method according to claim 1, characterized in that The risk type includes the financial product risk, and the standardized transaction information data is input into the target risk assessment model, and the risk index corresponding to the risk type is output, including: When the target risk assessment model is a convergent model obtained by iterative training using the supervised learning method, obtaining financial product features of the financial product and feature scores corresponding to the financial product features from the standardized transaction information data; The risk index of the financial product risk is obtained by multiplying the feature score corresponding to each of the financial product features by a set weight coefficient, wherein the financial product features at least include the rate of return, the reputation and qualifications of the issuing institution, the risk rating of the financial product, the market environment and the macroeconomic conditions, and the risk index of the financial product risk at least includes the capital cost risk index, the liquidity risk index and the price risk index of the financial product, and the liquidity risk index represents the risk index of the financial product in the market liquidity conditions; When the target risk assessment model is a convergent model obtained by iterative training using the unsupervised learning method, the transaction data of the financial product is obtained from the standardized transaction information data, and a risk assessment is performed based on the transaction data of the financial product to obtain the risk index of the financial product risk.

7. The method according to claim 6, characterized in that When the risk index is greater than the corresponding set index threshold, a prompt message is issued, including: In the case where the risk type is the financial product risk, comparing the price risk index, the liquidity risk index, and the capital cost risk index of the financial product with the set index threshold; In the case where a risk warning rule is triggered, issuing the warning information, wherein the risk warning rule is that one or two of the risk indexes among the price risk index, the liquidity risk index, and the capital cost risk index are higher than a first set index threshold; When a prohibited transaction rule is triggered, the current transaction is terminated, and the prohibited transaction rule is that the price risk index, the liquidity risk index, and the funding cost risk index are all higher than the first set index threshold; In a case where the risk type is the transaction user risk, comparing the risk index corresponding to the transaction user risk with a second set index threshold; When the risk index corresponding to the transaction user risk is higher than the second set index threshold, determining that the transaction user has a transaction credit risk and issuing the prompt information; In the case where the risk type is the quotation information risk, receiving quotation information of the transaction user on the current financial product, querying the current market price of the current financial product, and comparing the quotation information with the current market price; When the absolute value of the difference between the quotation information and the current market price is greater than the price index threshold, it is determined that the authenticity of the current transaction is low, the prompt information is issued and the current transaction is terminated.

8. The method according to claim 1, characterized in that After inputting the standardized transaction information data into the target risk assessment model and outputting a risk index corresponding to the risk type, the method further includes: When the risk index is less than the corresponding set index threshold, the current transaction is continued until the transaction is completed, and the characteristic score of the risk characteristics is corrected in real time. The risk characteristics include at least transaction behavior characteristics, geographic location characteristics, financial product characteristics and customer behavior characteristics.

9. A transaction authenticity risk assessment device, characterized in that: The device comprises: An acquisition unit, used to acquire standardized transaction information data, wherein the standardized transaction information data at least includes standardized transaction user information, financial product element information, financial product price index, transaction data and external data, wherein the financial product price index at least includes Shibor interest rate, DR interest rate and recent transaction records of corresponding financial products; An input-output unit, used for inputting the standardized transaction information data into a target risk assessment model, and outputting a risk index corresponding to a risk type, wherein one risk type corresponds to at least one risk index, and the target risk assessment model is a convergence model obtained by iteratively training using multiple sets of training data using an unsupervised learning method or a supervised learning method, wherein each set of training data includes sample standardized transaction data and a corresponding sample risk index, and the risk type includes at least one of the following: transaction user risk, financial product risk, transaction product risk, and quotation information risk; The prompt unit is used to issue a prompt message when the risk index is greater than the corresponding set index threshold, wherein the prompt message is used to prompt that there is a risk in the current transaction and the current transaction needs to be terminated.

10. A transaction authenticity risk assessment system, characterized in that: include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include methods for executing any one of claims 1 to 7.

Images