Transaction type determination method and device, storage medium and electronic equipment

Abstract

The invention discloses a transaction type determination method and device, a storage medium and electronic equipment, and relates to the field of artificial intelligence, and the method comprises the steps: obtaining transaction information of a target account, the transaction information being used for representing data information generated when the target account carries out a transaction behavior; predicting the transaction information by using a prediction model to obtain a prediction result, the prediction model being obtained by training a neural network model by using a transaction information sample and a prediction result sample, and the prediction result being used for representing a risk level of the transaction information; and determining the transaction type of the target object based on the prediction result, the transaction type including a first transaction type and a second transaction type, and the risk level corresponding to the first transaction type being higher than the risk level corresponding to the second transaction type. Through the method and the device, the problem of high misjudgment rate of a transaction recognition algorithm in related technologies is solved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence, and in particular, to a method, device, storage medium and electronic device for determining a transaction type. Background Art

[0002] With the rapid development of my country's economy and the increasing wealth of residents, the frequency of high-risk transactions (such as high-risk transactions) in transactions has gradually increased. Since high-risk transactions are highly professional and hidden, it is particularly important to accurately locate high-risk transaction behaviors.

[0003] Currently, there are a variety of algorithm models for high-risk transactions to help identify high-risk transactions, but inaccurate identification still occurs, resulting in ordinary people's accounts being mistakenly frozen due to inaccurate model identification, causing trouble for ordinary people.

[0004] There is currently no effective solution to the problem of low accuracy in identifying high-risk models in existing technologies. Summary of the invention

[0005] The main purpose of the present application is to provide a method, device, storage medium and electronic device for determining a transaction type, so as to solve the problem of high misjudgment rate of transaction identification algorithms in related technologies.

[0006] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a method for determining a transaction type is provided, comprising: obtaining transaction information of a target account, wherein the transaction information is used to represent data information generated when the target account performs a transaction; predicting the transaction information using a prediction model to obtain a prediction result, wherein the prediction model is obtained by training a neural network model using transaction information samples and prediction result samples, and the prediction result is used to represent the risk level of the transaction information; based on the prediction result, determining the transaction type, wherein the transaction type includes a first transaction type and a second transaction type, and the risk level corresponding to the first transaction type is higher than the risk level corresponding to the second transaction type.

[0007] Optionally, the method also includes: obtaining transaction information samples; inputting the transaction information samples into the neural network model to obtain an initial value of the prediction result; and training the neural network model based on the initial value of the prediction result and the prediction result samples to obtain a prediction model.

[0008] Optionally, using a prediction model to predict transaction information to obtain a prediction result includes: preprocessing the transaction information to obtain preprocessed transaction information; inputting the preprocessed transaction information into the prediction model, and using the prediction model to predict the preprocessed transaction information to obtain a prediction result.

[0009] Optionally, the transaction information is preprocessed to obtain preprocessed transaction information, including: performing data cleaning on the transaction information to obtain cleaned transaction information; performing feature extraction on the cleaned transaction information to obtain transaction features; encoding the transaction features to obtain encoded transaction features; and generating preprocessed transaction information based on the encoded transaction features.

[0010] Optionally, the transaction information includes: time information of the transaction behavior, transaction funds of the transaction behavior, transaction channels of the transaction behavior, transaction regions of the transaction behavior, and fund flows of the transaction behavior; transaction features include time features, amount features, currency features, and account features; wherein feature extraction is performed on the cleaned transaction information to obtain transaction features, including: feature extraction of the time information in the cleaned transaction information to obtain time features, and adjustment and extraction of the transaction channels, transaction regions, transaction funds, and fund flows in the cleaned transaction information to obtain amount features, currency features, and account features.

[0011] Optionally, encoding the transaction features to obtain the encoded transaction features includes: encoding the transaction features using a one-hot encoding strategy to obtain the encoded transaction features, wherein the one-hot encoding features are used to represent rules for encoding the transaction features.

[0012] In order to achieve the above-mentioned purpose, according to another aspect of the present application, a device for determining a transaction type is provided. The device comprises: an acquisition module, the acquisition module is used to acquire the transaction information of the target account, wherein the transaction information is used to represent the data information generated when the target account performs a transaction; a prediction module, the prediction module is used to predict the transaction information using a prediction model to obtain a prediction result, wherein the prediction model is obtained by training a neural network model using transaction information samples and prediction result samples, and the prediction result is used to represent the risk level of the transaction information; a determination module, the determination module is used to determine the transaction type based on the prediction result, wherein the transaction type includes a first transaction type and a second transaction type, and the risk level corresponding to the first transaction type is higher than the risk level corresponding to the second transaction type.

[0013] According to another aspect of the present application, a computer-readable storage medium is provided, the computer-readable storage medium including a stored executable program, wherein when the executable program runs, the device where the computer-readable storage medium is located is controlled to execute the above-mentioned transaction type determination method.

[0014] According to another aspect of the present application, an electronic device is provided, comprising: a memory storing an executable program; and a processor for running the program, wherein the above method is executed when the program is running.

[0015] According to another aspect of the present application, a computer program product is provided, comprising computer instructions, which implement the steps of the above method when executed by a processor.

[0016] In the embodiments of the present application, a trained model is used for prediction to ensure that the model can learn effective patterns from the data, thereby accurately identifying high-risk transactions, thereby effectively reducing the misjudgment rate of high-risk transactions, improving the security and compliance of financial transactions, and solving the problem of high misjudgment rate of transaction identification algorithms in related technologies, thereby achieving the effect of improving the recognition accuracy and efficiency of transaction types. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0018] Figure 1 A hardware structure block diagram of a computer terminal for implementing a method for determining a transaction type is shown;

[0019] Figure 2 is a flow chart of a method for determining a transaction type according to one embodiment of the present application;

[0020] Figure 3 is a flow chart of a method for determining a transaction type according to one embodiment of the present application;

[0021] Figure 4 is a schematic diagram of a device for determining a transaction type according to one embodiment of the present application;

[0022] Figure 5 It is a structural block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0023] 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.

[0024] 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 sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0025] First, some nouns or terms that appear in the description of the embodiments of the present application are subject to the following explanations:

[0026] It should be noted that the collected information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for display, data for analysis, etc.) involved in this application are information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of relevant data are in compliance with relevant laws, regulations and standards, necessary confidentiality measures are taken, and public order and good customs are not violated, and corresponding operation entrances are provided for users to choose to authorize or refuse. For example, an interface is set up between this system and relevant users or institutions to provide users with corresponding operation entrances for users to choose to agree or refuse the results of automated decision-making; if the user chooses to refuse, the expert decision-making process will be entered.

[0027] Example 1

[0028] According to an embodiment of the present application, an embodiment of a method for determining a transaction type is also 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] The method embodiment provided in the first embodiment of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1 FIG. 1 shows a hardware structure block diagram of a computer terminal (or mobile device) for implementing a method for determining a transaction type. Figure 1As shown, the computer terminal 10 (or mobile device) may include one or more (102a, 102b, ..., 102n are used to illustrate) processors 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), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply and / or a camera. 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 above electronic device. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0030] It should be noted that the one or more processors 102 and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the computer terminal 10 (or mobile device). As described in the embodiments of the present application, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0031] The memory 104 can be used to store software programs and modules of application software, such as program instructions / data storage devices corresponding to the method for determining the transaction type in the embodiment of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, the method for determining the transaction type described above 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 may be connected to the computer terminal 10 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 combinations thereof.

[0032] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, 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 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0033] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the computer terminal 10 (or mobile device).

[0034] Under the above operating environment, this application provides Figure 2 The method for determining the transaction type shown. Figure 2 It is a flowchart of the method for determining the transaction type according to Example 1 of the present application.

[0035] Step S102, obtaining transaction information of the target account, wherein the transaction information is used to represent data information generated when the target account performs transaction behavior;

[0036] In step S102, the transaction information includes the time information of the transaction behavior, the transaction funds of the transaction behavior, the transaction channels of the transaction behavior, the transaction areas of the transaction behavior, and the flow of funds of the transaction behavior. By predicting the above transaction information, it is helpful to determine the risk level and transaction type.

[0037] Step S104, using the prediction model to predict the transaction information to obtain a prediction result, wherein the prediction model is obtained by training a neural network model using transaction information samples and prediction result samples, and the prediction result is used to indicate the risk level of the transaction information.

[0038] In step S104, during the prediction phase, detailed information of each transaction will be input into the model, and the model will output a prediction result based on the patterns and rules it has learned. The result is a probability value, indicating the risk level of the transaction information. For example, a prediction result of 0 indicates "low risk"; a prediction result of 1 indicates "high risk".

[0039] Step S106, based on the prediction result, determining the transaction type, wherein the transaction type includes a first transaction type and a second transaction type, and the risk level corresponding to the first transaction type is higher than the risk level corresponding to the second transaction type.

[0040] In step S106, based on the prediction results of the model, the transaction is classified into the first transaction type (high-risk transaction) or the second transaction type (low-risk transaction). The risk level corresponding to the first transaction type is significantly higher than that of the second transaction type. Such classification helps financial institutions or regulatory agencies to prioritize the review and processing of high-risk transactions. For example, for transactions of the first transaction type, the system may trigger further investigation or take measures such as temporarily freezing the account.

[0041] By implementing steps S102 to S106, an efficient and accurate transaction type determination model can be constructed, thereby effectively reducing the misjudgment rate in anti-high-risk activities and improving the security and compliance of financial transactions.

[0042] Optionally, the method for determining the transaction type provided in the embodiment of the present application further includes:

[0043] Step S201, obtaining a transaction information sample.

[0044] In step S201, a large amount of sample data is obtained from historical transaction data. These sample data should include various types of transactions, especially the known first transaction type (high-risk transaction) and the second transaction type (low-risk transaction), so that the model can learn the characteristic patterns of high-risk behavior from these data. The acquired transaction information samples need to cover different aspects of the transaction, such as transaction amount, transaction time, transaction channel, transaction region, amount flow, etc., to ensure that the model can comprehensively analyze the transaction data.

[0045] Step S202, inputting the transaction information sample into the neural network model to obtain the initial value of the prediction result.

[0046] In step S202, the transaction information sample is input into the neural network model for prediction. In the early stage of model training, the model parameters are usually randomly initialized, so the model's prediction results (prediction result initial values) may be significantly different from the actual transaction risk level. The purpose of this step is to evaluate the prediction ability of the model in its current state and provide a benchmark for subsequent model optimization.

[0047] Step S203: training the neural network model based on the prediction result initial value and the prediction result sample to obtain a prediction model.

[0048] In step S203, the core of model training is to adjust the model parameters to minimize the difference between the initial value of the prediction result and the sample of the prediction result. The sample of the prediction result is a known transaction risk level, that is, the correct result that the model should predict. This step is implemented by the back propagation algorithm, which reversely adjusts the weights and biases of the neural network according to the error between the initial value of the prediction result and the sample of the prediction result, so that the model can gradually learn and optimize its prediction ability.

[0049] Furthermore, the neural network is mainly divided into an input layer, an inference layer, and an output layer. Based on the transaction information samples of high-risk transactions, each transaction is vectorized and marked as the input layer, the main attributes of high-risk transactions are sorted out, and each transaction information of a transaction is used as an element, such as transaction time, transaction amount, transaction channel, transaction region, amount flow, etc., and each attribute is given an initial vector value and weight as the inference layer. If the output layer prediction value is significantly different from the comparison between high-risk transactions and low-risk transactions, the weights of each attribute in the inference layer are modified and calculated repeatedly until the error rate between the prediction result and the given value is within 0.0001.

[0050] Furthermore, some transaction information of the transaction is selected to extract transaction features, such as currency, account, amount, time, etc., and the transaction information is vectorized using unique hot encoding to construct an N-dimensional vector. The transaction attribute vectorization is used as the input layer, and the implicit calculation method is used in the reasoning layer. The high-risk transactions and low-risk transactions determined by bank staff are used as positive and negative example data as the output layer results. The initial weight is set for the importance of each attribute in the reasoning layer, such as currency 1.1, account 1, amount 1.2, time 0.7, and one type is used as a training model for one layer. The output layer is calculated through multiple models to obtain the judgment result. In the reasoning layer, it is necessary to constantly adjust the degree of influence of each attribute on a transaction.

[0051] During the training process, steps S202 and S203 are repeated continuously, that is, transaction information samples are continuously input to obtain the initial value of the prediction result, and then the model parameters are adjusted based on the prediction error until the prediction accuracy of the model reaches a predetermined threshold or the training process converges. This process may involve data preprocessing (such as feature selection, feature scaling), adjustment of model structure (such as adding or reducing neural network layers), and optimization of training strategies (such as learning rate adjustment, using different optimization algorithms).

[0052] Based on steps S201 to S203, the resulting prediction model will be able to accurately predict the risk level of a transaction based on the input transaction information, helping financial institutions or regulators to promptly identify and handle potential high-risk transactions and improve the efficiency and accuracy of transaction monitoring.

[0053] Optionally, in the method for determining the transaction type provided in the embodiment of the present application, the transaction information is predicted using a prediction model to obtain a prediction result, including:

[0054] Step S211, preprocessing the transaction information to obtain preprocessed transaction information.

[0055] In step S211, preprocessing is a key step in the machine learning process, the purpose of which is to convert the original transaction information into a format that can be effectively processed by the model. The preprocessing step may include the following:

[0056] Data cleaning: Remove outliers, erroneous data or missing values ​​in transaction information to ensure data quality. For example, correct the timestamp format, process abnormally large values ​​in the amount, etc.

[0057] Feature engineering: Extract useful features from raw data that can help the model better understand the characteristics of transactions. This may include converting transaction times into time features (such as time periods, weekdays / weekends, holidays, etc.), converting transaction locations into geocodes, and processing text data (such as transaction descriptions).

[0058] Encoding: Encoding non-numeric features so that they can be processed by the neural network model. Common encoding methods include one-hot encoding, label encoding or embedding encoding, especially for categorical features such as transaction channels, currencies, etc.

[0059] Standardization or normalization: Standardize or normalize numerical features to avoid the impact of feature magnitude differences during model training. For example, convert transaction amounts to the same scale (such as [0, 1] interval).

[0060] Feature selection: Feature selection may also be required to retain only the features that have a significant impact on the prediction results in order to reduce the complexity of the model and improve the prediction speed and accuracy.

[0061] The preprocessed transaction information should contain all necessary and valid features, have a unified format and high data quality, and can be directly input into the prediction model for analysis.

[0062] Step S212, input the preprocessed transaction information into the prediction model, use the prediction model to predict the preprocessed transaction information, and obtain a prediction result.

[0063] In step S212, after the preprocessing stage is completed, the preprocessed transaction information is input into the prediction model in an appropriate format. The prediction model has been fully trained at this stage and can predict the risk level of the transaction based on the input feature data. The model may be a deep neural network, such as a BP neural network, which calculates the prediction results through a forward propagation algorithm.

[0064] The prediction result is usually one or more numerical values, indicating the risk level of the transaction. For example, the model may output a value between 0 and 1, indicating the probability that the transaction is a high-risk transaction, where if the output prediction result is 1, it means that the transaction type is "high risk", and if the output prediction result is 0, it means that the transaction type is "low risk". These prediction results will be used to determine the subsequent transaction type, helping financial institutions identify potential high-risk transaction behaviors.

[0065] Based on steps S211 to S212, it can be ensured that the prediction model is analyzed based on high-quality data, thereby improving the accuracy and efficiency of high-risk transaction identification.

[0066] Optionally, in the method for determining the transaction type provided in the embodiment of the present application, preprocessing the transaction information to obtain the preprocessed transaction information includes:

[0067] Step S221, clean the transaction information to obtain cleaned transaction information.

[0068] In step S221, data cleaning is the first step of preprocessing, and its main purpose is to ensure the reliability and consistency of the data set. Specifically, it includes the following steps:

[0069] a. Remove or correct erroneous data entries, such as abnormal values ​​of transaction amounts and incorrect timestamp formats.

[0070] b. To handle missing data, you can fill in the missing values ​​(such as using the mean or median), delete the missing values, or predict the missing values.

[0071] c. Remove duplicate data to avoid bias during model training.

[0072] The cleaned transaction information based on the above steps will be more accurate and complete, providing a solid foundation for subsequent feature extraction and encoding.

[0073] Step S222, extracting features from the cleaned transaction information to obtain transaction features.

[0074] In step S222, feature extraction is to convert the raw data into a format that can be understood and utilized by the machine learning model. For transaction information, feature extraction may include:

[0075] a. Extract transaction time features: Convert timestamps into time periods of the day, weekdays / weekends, months, etc. These features can help the model learn time patterns.

[0076] b. Extract amount features: It may be necessary to classify or group transaction amounts to identify transaction patterns in different amount ranges.

[0077] c. Extract transaction channel features: Convert the channels where transactions occur (such as online banking, ATM, and counter services) into classification features, which are very important for identifying transaction behaviors in specific channels.

[0078] d. Extract transaction area features: Convert transaction locations into geocodes or region identifiers to help the model understand transaction patterns related to geographic locations.

[0079] e. Extract the amount flow characteristics: analyze the source and destination of the transaction amount and identify abnormal fund flow patterns.

[0080] Step S223, encode the transaction features to obtain encoded transaction features.

[0081] In step S223, feature encoding is to convert non-numeric features into numeric ones so that the model can process them. Common encoding methods include:

[0082] One-hot encoding: Convert categorical features into binary vectors, where each category has an independent bit. When the feature appears, the corresponding bit is set to 1, and the other bits are 0. For example, for the transaction channel feature, "online banking" can be encoded as [1, 0, 0], "ATM" can be encoded as [0, 1, 0], and "counter service" can be encoded as [0, 0, 1].

[0083] Label encoding: converts categorical features into integer encoding, which is suitable for ordered classification. However, in transaction type identification, one-hot encoding is usually more applicable because it can avoid introducing unnecessary ordinal relationships.

[0084] Step S224, generating pre-processed transaction information based on the encoded transaction features.

[0085] In step S224, after feature extraction and encoding, all transaction features will be integrated into a unified, pre-processed transaction information format, that is, a vector or data set containing multiple features, each of which is properly encoded and formatted for input into the neural network model. The pre-processed transaction information not only contains the key information of the original transaction, but also presents it in a format that is easy for the machine learning model to understand, providing an optimized data foundation for model training and prediction.

[0086] Based on steps S221 to S224, the original transaction data that may contain errors or non-numerical information can be converted into clean, structured, and numerically pre-processed transaction information, greatly improving the efficiency of model training and the accuracy of prediction results.

[0087] Optionally, in the method for determining the transaction type provided in the embodiment of the present application, the transaction information includes: time information of the transaction behavior, transaction funds of the transaction behavior, transaction channels of the transaction behavior, transaction regions of the transaction behavior, and fund flows of the transaction behavior; the transaction features include time features, amount features, currency features, and account features; wherein feature extraction is performed on the cleaned transaction information to obtain transaction features, including:

[0088] Step S231, extracting features from the time information in the cleaned transaction information to obtain time features, and adjusting and extracting the transaction channels, transaction areas, transaction funds and fund flows in the cleaned transaction information to obtain amount features, currency features and account features.

[0089] In step S231, time information is extracted from the cleaned transaction information as time features, where time features include but are not limited to: time period, weekdays / weekends, and months. Dividing a day into different time periods (such as morning, afternoon, late night) helps identify abnormal transactions that occur frequently during a specific time period. Differentiating whether the transaction is conducted on a weekday or weekend, high-risk traders may be more inclined to trade during non-working hours to avoid immediate detection. Analyzing the seasonal pattern of transactions, some months may become peak periods for high-risk activities due to annual settlements, holidays, etc.

[0090] Analyze transaction funds and extract amount characteristics. Amount characteristics include but are not limited to: transaction amount size, amount distribution, and amount change trend. Pay attention to large transactions, because high-risk transaction activities usually involve the transfer of large amounts of funds. Analyze the frequency distribution of amounts and identify abnormal amount combinations or distribution patterns. Check the inflow and outflow of account amounts to see if there are frequent large amount changes in funds in a short period of time.

[0091] Extract the currency features involved in the transaction, as some currencies may be more commonly used in high-risk transaction activities, especially in the case of cross-border high-risk transactions. Currency features include common currencies and unusual currencies, focusing on currencies that appear more frequently in transactions, identifying uncommon currency transactions or those that are not commonly used in specific regions, which may be signs of high-risk transaction activities.

[0092] Analyze the information of trading accounts and extract account features, including but not limited to: account type, account activity frequency and account history, distinguish between personal accounts and corporate accounts, etc. Different types of accounts may have different trading risks. Check the activity of the account, including transaction frequency and transaction time. Including account creation time, abnormalities in previous transaction records, and correlation with known high-risk trading accounts.

[0093] Transaction channel and transaction area feature extraction: This involves the way and geographical location of the transaction, including but not limited to: transaction channels and transaction areas, such as online banking, ATM, counter services, etc. Different channels may have different risk levels. Analyze the geographical location where the transaction occurs, including countries, cities, etc. Some areas may become high-risk areas for high-risk transactions due to the regulatory environment or geographical location.

[0094] Based on step S231, complex transaction information can be converted into a series of specific numerical or categorical features, which will be further encoded and integrated to provide input for the machine learning model to identify suspicious high-risk transaction behaviors. The selection and extraction of features should be based on a deep understanding of high-risk transaction behavior patterns and a careful analysis of the data set to ensure that the extracted features have practical value for model training and prediction.

[0095] Optionally, in the method for determining the transaction type provided in the embodiment of the present application, encoding the transaction feature to obtain the encoded transaction feature includes:

[0096] Step S241, encode the transaction features using a one-hot encoding strategy to obtain encoded transaction features, wherein the one-hot encoding features are used to represent a rule for encoding the transaction features.

[0097] It should be noted that one-hot encoding is an encoding method that converts categorical features into a set of binary vectors, where each categorical value corresponds to a vector bit, and when a feature value appears, the corresponding bit is set to 1, and the remaining bits are 0. This method avoids introducing unnecessary ordinal relationships in the model, allowing the model to correctly process categorical information without mistaking it for a continuous variable.

[0098] In step S241, for transaction features, the application of one-hot encoding may include:

[0099] Time features: For example, if trading hours are classified into "morning", "noon", "evening", and "late night", each category will correspond to a one-hot encoded vector. For "morning", the encoding may be [1, 0, 0, 0], for "noon" it is [0, 1, 0, 0], and so on.

[0100] Transaction channel features: Different transaction channels, such as "online banking", "ATM", and "counter service", will be encoded as different one-hot encoding vectors. For example, "online banking" may be encoded as [1, 0, 0], "ATM" as [0, 1, 0], and "counter service" as [0, 0, 1].

[0101] Transaction region features: If the transaction regions are classified into different countries or regions, each region will also be converted into a one-hot encoded vector to ensure that the model can correctly identify and process transactions in different regions.

[0102] Account characteristics: Categorical information such as account type and account history can also be converted through one-hot encoding. For example, "personal account" and "corporate account" will be encoded as [1, 0] and [0, 1] respectively.

[0103] Funds flow features: If the funds flow is classified into different categories (such as "deposit", "withdrawal", "transfer"), then each category will have a one-hot encoded vector.

[0104] Based on step S241, the encoded transaction features will be used as input to the neural network model to help the model understand and learn the importance of different features for identifying high-risk transactions. One-hot encoding enables the model to independently evaluate the impact of each categorical feature without confusing them with numerical features, thereby improving the model's prediction accuracy and explanatory power. After encoding is completed, all features will be integrated into a unified format to provide data support for model training and prediction.

[0105] This application provides Figure 3 The method for determining the transaction type shown comprises the following steps:

[0106] Step S201, obtaining a transaction information sample.

[0107] Step S202, inputting the transaction information sample into the neural network model to obtain the initial value of the prediction result.

[0108] Step S203: training the neural network model based on the prediction result initial value and the prediction result sample to obtain a prediction model.

[0109] Step S211, preprocessing the transaction information to obtain preprocessed transaction information.

[0110] Step S212, input the preprocessed transaction information into the prediction model, use the prediction model to predict the preprocessed transaction information, and obtain a prediction result.

[0111] Step S221, clean the transaction information to obtain cleaned transaction information.

[0112] Step S222, extracting features from the cleaned transaction information to obtain transaction features.

[0113] Step S223, encode the transaction features to obtain encoded transaction features.

[0114] Step S224, generating pre-processed transaction information based on the encoded transaction features.

[0115] Step S231, extracting features from the time information in the cleaned transaction information to obtain time features, and adjusting and extracting the transaction channels, transaction areas, transaction funds and fund flows in the cleaned transaction information to obtain amount features, currency features and account features.

[0116] Step S241, encode the transaction features using a one-hot encoding strategy to obtain encoded transaction features, wherein the one-hot encoding features are used to represent a rule for encoding the transaction features.

[0117] Based on steps S201 to S241, after data cleaning, feature extraction and coding preprocessing, the trained model is used for prediction to ensure that the model can learn the most effective pattern from the data, thereby accurately identifying high-risk transactions. Through this series of steps, strong risk control support is provided for financial institutions, solving the problem of high misjudgment rate of transaction identification algorithms in related technologies, thereby achieving the effect of improving the recognition accuracy and efficiency of transaction types.

[0118] 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.

[0119] Example 2

[0120] The embodiment of the present application also provides a transaction type determination device. It should be noted that the transaction type determination device of the embodiment of the present application can be used to execute the transaction type determination method provided in the embodiment of the present application. The transaction type determination device provided in the embodiment of the present application is introduced below.

[0121] According to an embodiment of the present application, a device for implementing the above transaction type determination method is also provided, such as Figure 4 As shown, the device comprises:

[0122] The acquisition module 301 is used to acquire the transaction information of the target account, wherein the transaction information is used to represent the data information generated when the target account performs a transaction;

[0123] Prediction module 302, prediction module 302 is used to predict transaction information using a prediction model to obtain a prediction result, wherein the prediction model is obtained by training a neural network model using transaction information samples and prediction result samples, and the prediction result is used to indicate the risk level of the transaction information.

[0124] The determination module 303 is used to determine the transaction type based on the prediction result, wherein the transaction type includes a first transaction type and a second transaction type, and the risk level corresponding to the first transaction type is higher than the risk level corresponding to the second transaction type.

[0125] The transaction type determination device provided in the embodiment of the present application uses a trained model for prediction, ensuring that the model can learn effective patterns from the data, thereby accurately identifying high-risk transactions, thereby effectively reducing the misjudgment rate of high-risk transactions, improving the security and compliance of financial transactions, and solving the problem of high misjudgment rate of transaction identification algorithms in related technologies, thereby achieving the effect of improving the recognition accuracy and efficiency of transaction types.

[0126] Optionally, in the transaction type determination device provided in the embodiment of the present application, the acquisition module 301 is also used to obtain transaction information samples; the prediction module 302 is also used to input the transaction information samples into the neural network model to obtain an initial value of the prediction result; based on the initial value of the prediction result and the prediction result sample, the neural network model is trained to obtain a prediction model.

[0127] Optionally, in the transaction type determination device provided in the embodiment of the present application, the prediction module 302 is also used to preprocess the transaction information to obtain preprocessed transaction information; input the preprocessed transaction information into the prediction model, and use the prediction model to predict the preprocessed transaction information to obtain a prediction result.

[0128] Optionally, in the transaction type determination device provided in the embodiment of the present application, the prediction module 302 is also used to perform data cleansing on the transaction information to obtain cleansed transaction information; perform feature extraction on the cleansed transaction information to obtain transaction features; encode the transaction features to obtain encoded transaction features; and generate preprocessed transaction information based on the encoded transaction features.

[0129] Optionally, in the transaction type determination device provided in the embodiment of the present application, the prediction module 302 is also used to perform feature extraction on the time information in the cleaned transaction information to obtain time features, and to adjust and extract the transaction channels, transaction areas, transaction funds and fund flows in the cleaned transaction information to obtain amount features, currency features and account features.

[0130] It should be noted that the acquisition module 301, prediction module 302 and determination module 303 correspond to steps S201 to S241 in Example 1, and the two modules and the corresponding steps implement the same examples and application scenarios, but are not limited to the contents disclosed in the above-mentioned Example 1. It should be noted that the above-mentioned modules or units may be hardware components or software components stored in a memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n), and the above-mentioned modules may also be part of the device and may be run in the computer terminal 10 provided in Example 1.

[0131] Example 3

[0132] An embodiment of the present application may provide an electronic device, Figure 5 is a structural block diagram of an electronic device according to an embodiment of the present application. Figure 5 As shown, the electronic device may include: one or more ( Figure 5 (only one is shown) processor 1002, memory 1004, storage controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0133] Among them, the memory can be used to store software programs and modules, such as program instructions / modules corresponding to the methods and devices in the embodiments of the present application, and the processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, realizing the above-mentioned method. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0134] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: obtain the transaction information of the target account, wherein the transaction information is used to represent the data information generated when the target account performs transaction behavior; use the prediction model to predict the transaction information to obtain the prediction result, wherein the prediction model is obtained by training the neural network model using the transaction information sample and the prediction result sample, and the prediction result is used to represent the risk level of the transaction information. Based on the prediction result, the transaction type is determined, wherein the transaction type includes a first transaction type and a second transaction type, and the risk level corresponding to the first transaction type is higher than the risk level corresponding to the second transaction type.

[0135] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: obtain transaction information samples; input the transaction information samples into the neural network model to obtain the initial value of the prediction result; based on the initial value of the prediction result and the prediction result sample, train the neural network model to obtain the prediction model.

[0136] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: preprocessing the transaction information to obtain preprocessed transaction information; inputting the preprocessed transaction information into the prediction model, and using the prediction model to predict the preprocessed transaction information to obtain a prediction result.

[0137] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: perform data cleansing on the transaction information to obtain cleaned transaction information; perform feature extraction on the cleaned transaction information to obtain transaction features; encode the transaction features to obtain encoded transaction features; and generate pre-processed transaction information based on the encoded transaction features.

[0138] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: extract features from the time information in the cleaned transaction information to obtain time features, and adjust and extract the transaction channels, transaction areas, transaction funds and fund flows in the cleaned transaction information to obtain amount features, currency features and account features.

[0139] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: encode the transaction features using a one-hot encoding strategy to obtain encoded transaction features, wherein the one-hot encoding features are used to represent the rules for encoding the transaction features.

[0140] By using the embodiment of the present application, a method for determining the transaction type is provided. The trained model is used for prediction to ensure that the model can learn effective patterns from the data, thereby accurately identifying high-risk transactions, thereby effectively reducing the misjudgment rate of high-risk transactions, improving the security and compliance of financial transactions, and solving the problem of high misjudgment rate of transaction identification algorithms in related technologies, thereby achieving the effect of improving the recognition accuracy and efficiency of transaction types.

[0141] It can be understood by those skilled in the art that Figure 5 The structure shown is for illustration only, and the electronic device may also be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a PDA, a mobile Internet device (Mobile Internet Devices, MID), a PAD, and other terminal devices. Figure 5The structure of the electronic device is not limited. Figure 5 More or fewer components (such as network interfaces, display devices, etc.) shown in, or having Figure 5 Different configurations shown.

[0142] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, and the storage medium may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0143] Example 4

[0144] The embodiment of the present application also provides a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the method for determining the transaction type provided in the first embodiment. The method for determining the transaction type includes the following steps:

[0145] Step S102, obtaining transaction information of the target account, wherein the transaction information is used to represent data information generated when the target account performs transaction behavior;

[0146] Step S104, using the prediction model to predict the transaction information to obtain a prediction result, wherein the prediction model is obtained by training a neural network model using transaction information samples and prediction result samples, and the prediction result is used to indicate the risk level of the transaction information.

[0147] Step S106, based on the prediction result, determining the transaction type, wherein the transaction type includes a first transaction type and a second transaction type, and the risk level corresponding to the first transaction type is higher than the risk level corresponding to the second transaction type.

[0148] In another embodiment of the present application, the method for determining the transaction type further includes the following steps:

[0149] Step S201, obtaining a transaction information sample.

[0150] Step S202, inputting the transaction information sample into the neural network model to obtain the initial value of the prediction result.

[0151] Step S203: training the neural network model based on the prediction result initial value and the prediction result sample to obtain a prediction model.

[0152] Step S211, preprocessing the transaction information to obtain preprocessed transaction information.

[0153] Step S212, input the preprocessed transaction information into the prediction model, use the prediction model to predict the preprocessed transaction information, and obtain a prediction result.

[0154] Step S221, clean the transaction information to obtain cleaned transaction information.

[0155] Step S222, extracting features from the cleaned transaction information to obtain transaction features.

[0156] Step S223, encode the transaction features to obtain encoded transaction features.

[0157] Step S224, generating pre-processed transaction information based on the encoded transaction features.

[0158] Step S231, extracting features from the time information in the cleaned transaction information to obtain time features, and adjusting and extracting the transaction channels, transaction areas, transaction funds and fund flows in the cleaned transaction information to obtain amount features, currency features and account features.

[0159] Step S241, encode the transaction features using a one-hot encoding strategy to obtain encoded transaction features, wherein the one-hot encoding features are used to represent a rule for encoding the transaction features.

[0160] Based on steps S201 to S241, after data cleaning, feature extraction and coding preprocessing, the trained model is used for prediction to ensure that the model can learn the most effective pattern from the data, thereby accurately identifying high-risk transactions. Through this series of steps, strong risk control support is provided for financial institutions, solving the problem of high misjudgment rate of transaction identification algorithms in related technologies, thereby achieving the effect of improving the recognition accuracy and efficiency of transaction types.

[0161] Optionally, in this embodiment, the above storage medium may be located in any computer terminal in a computer terminal group in a computer network, or in any mobile terminal in a mobile terminal group.

[0162] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing the steps of the method for determining the transaction type. The method for determining the transaction type includes the following steps:

[0163] Step S102, obtaining transaction information of the target account, wherein the transaction information is used to represent data information generated when the target account performs transaction behavior;

[0164] Step S104, using the prediction model to predict the transaction information to obtain a prediction result, wherein the prediction model is obtained by training a neural network model using transaction information samples and prediction result samples, and the prediction result is used to indicate the risk level of the transaction information.

[0165] Step S106, based on the prediction result, determining the transaction type, wherein the transaction type includes a first transaction type and a second transaction type, and the risk level corresponding to the first transaction type is higher than the risk level corresponding to the second transaction type.

[0166] In another embodiment of the present application, the method for determining the transaction type further includes the following steps:

[0167] Step S201, obtaining a transaction information sample.

[0168] Step S202, inputting the transaction information sample into the neural network model to obtain the initial value of the prediction result.

[0169] Step S203: training the neural network model based on the prediction result initial value and the prediction result sample to obtain a prediction model.

[0170] Step S211, preprocessing the transaction information to obtain preprocessed transaction information.

[0171] Step S212, input the preprocessed transaction information into the prediction model, use the prediction model to predict the preprocessed transaction information, and obtain a prediction result.

[0172] Step S221, clean the transaction information to obtain cleaned transaction information.

[0173] Step S222, extracting features from the cleaned transaction information to obtain transaction features.

[0174] Step S223, encode the transaction features to obtain encoded transaction features.

[0175] Step S224, generating pre-processed transaction information based on the encoded transaction features.

[0176] Step S231, extracting features from the time information in the cleaned transaction information to obtain time features, and adjusting and extracting the transaction channels, transaction areas, transaction funds and fund flows in the cleaned transaction information to obtain amount features, currency features and account features.

[0177] Step S241, encode the transaction features using a one-hot encoding strategy to obtain encoded transaction features, wherein the one-hot encoding features are used to represent a rule for encoding the transaction features.

[0178] Based on steps S201 to S241, after data cleaning, feature extraction and coding preprocessing, the trained model is used for prediction to ensure that the model can learn the most effective pattern from the data, thereby accurately identifying high-risk transactions. Through this series of steps, strong risk control support is provided for financial institutions, solving the problem of high misjudgment rate of transaction identification algorithms in related technologies, thereby achieving the effect of improving the recognition accuracy and efficiency of transaction types.

[0179] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0180] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0181] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0182] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0183] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0184] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.

[0185] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for determining a transaction type, characterized in that: include: Acquire transaction information of a target account, wherein the transaction information is used to represent data information generated when the target account performs a transaction; Using a prediction model to predict the transaction information to obtain a prediction result, wherein the prediction model is obtained by training a neural network model using transaction information samples and prediction result samples, and the prediction result is used to indicate the risk level of the transaction information; Based on the prediction result, the transaction type is determined, wherein the transaction type includes a first transaction type and a second transaction type, and the risk level corresponding to the first transaction type is higher than the risk level corresponding to the second transaction type.

2. The method according to claim 1, characterized in that The method further comprises: Obtaining the transaction information sample; Inputting the transaction information sample into the neural network model to obtain an initial value of the prediction result; Based on the prediction result initial value and the prediction result sample, the neural network model is trained to obtain the prediction model.

3. The method according to claim 1, characterized in that: Using the prediction model to predict the transaction information to obtain the prediction result includes: Preprocessing the transaction information to obtain the preprocessed transaction information; The preprocessed transaction information is input into the prediction model, and the preprocessed transaction information is predicted using the prediction model to obtain the prediction result.

4. The method according to claim 3, characterized in that Preprocessing the transaction information to obtain the preprocessed transaction information includes: Performing data cleansing on the transaction information to obtain cleansed transaction information; Performing feature extraction on the cleaned transaction information to obtain transaction features; Encoding the transaction feature to obtain the encoded transaction feature; Based on the encoded transaction features, the pre-processed transaction information is generated.

5. The method according to claim 4, characterized in that The transaction information includes: time information of the transaction behavior, transaction funds of the transaction behavior, transaction channels of the transaction behavior, transaction regions of the transaction behavior, and fund flows of the transaction behavior; the transaction features include time features, amount features, currency features, and account features; wherein feature extraction is performed on the cleaned transaction information to obtain the transaction features, including: Feature extraction is performed on the time information in the cleaned transaction information to obtain the time feature, and the transaction channel, the transaction area, the transaction funds and the fund flow direction in the cleaned transaction information are adjusted and extracted to obtain the amount feature, the currency feature and the account feature.

6. The method according to claim 4, characterized in that Encoding the transaction feature to obtain the encoded transaction feature includes: The transaction feature is encoded by adopting a one-hot encoding strategy to obtain the encoded transaction feature, wherein the one-hot encoding feature is used to represent a rule for encoding the transaction feature.

7. A device for determining a transaction type, characterized in that: include: An acquisition module, the acquisition module is used to acquire transaction information of a target account, wherein the transaction information is used to represent data information generated when the target account performs a transaction; A prediction module, the prediction module is used to predict the transaction information using a prediction model to obtain a prediction result, wherein the prediction model is obtained by training a neural network model using transaction information samples and prediction result samples, and the prediction result is used to indicate the risk level of the transaction information; A determination module is used to determine the transaction type based on the prediction result, wherein the transaction type includes a first transaction type and a second transaction type, and the risk level corresponding to the first transaction type is higher than the risk level corresponding to the second transaction type.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored executable program, wherein when the executable program is executed, the device where the computer-readable storage medium is located is controlled to execute the transaction type determination method according to any one of claims 1 to 6.

9. An electronic device, characterized in that: include: A memory storing an executable program; A processor, configured to run the program, wherein the program executes the method according to any one of claims 1 to 6 when running.

10. A computer program product comprising computer instructions, characterized in that: When the computer instructions are executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images