Bank transaction data processing method and system
By automating checksum abnormal tracking of bank transaction data, the problem of low efficiency of bank transaction data inspection is solved, and more efficient and accurate data inspection is achieved.
Patent Information
- Application Number
- CN202411855102.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-17
- Publication Date
- 2025-03-25
AI Technical Summary
The inspection efficiency of bank transaction data is low and the accuracy is poor, mainly due to the reliance on manual inspection of codes and logs, which leads to waste of manpower and time resources and is prone to omissions.
By obtaining transaction messages from the bank transaction log, automatically check based on preset verification rules, identify matching abnormalities, and generate abnormal transaction messages tracking orders.
It improves the inspection efficiency of bank transaction data, reduces the errors of manual intervention, and achieves higher inspection accuracy and efficiency.
Smart Images
Figure CN119693112A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of data processing technology, and in particular to a method and system for processing bank transaction data. Background Art
[0002] In the process of operating and managing the banking business system, there are many complex and difficult technical challenges. The interaction logic between the systems within the business system is intricate, which directly leads to difficulties in the actual operation of the architecture control. Different project teams have their own tasks and considerations, and different personnel have great differences in their understanding of architecture topics, which makes the system control difficult and seriously affects the efficiency of the business. In terms of system inspection, it currently mainly relies on manual inspection of codes and logs. This method not only requires a lot of manpower and time resources, but also because it is difficult for people to maintain a high degree of concentration for a long time, it is easy to miss something, resulting in low efficiency and poor accuracy of bank transaction data inspection.
[0003] Therefore, how to improve the verification efficiency of bank transaction data is an urgent problem to be solved. Summary of the invention
[0004] In view of this, the purpose of the present application is to provide a bank transaction data processing method and system to improve the verification efficiency of bank transaction data.
[0005] In conjunction with the first aspect of the present application, a method for processing bank transaction data is provided, the method comprising: Obtaining transaction messages corresponding to the transaction details of each bank from the bank transaction log, wherein the transaction message includes a request serial number, a global serial number, a system number, a transaction channel number, a transaction return code, and transaction return information, wherein the system number includes a requester system number corresponding to the request serial number and an original requester system number corresponding to the global serial number; Based on the preset verification rules, the request serial number, the global serial number, the system number, the transaction channel number, the transaction return code, and the transaction return information in the transaction message are verified to determine whether there is a matching anomaly; If the matching abnormality exists, obtaining abnormal content corresponding to the matching abnormality, wherein the abnormal content includes the type of the abnormal content; Generate an abnormal transaction message tracking sheet according to the type of the abnormal content and the source of the transaction message; Output the abnormal transaction message tracking sheet.
[0006] In a possible implementation of the first aspect, the verifying, based on a preset verification rule, the request serial number, the global serial number, the system number, the transaction channel number, the transaction return code, and the transaction return information in the transaction message to determine whether there is a matching abnormality includes: Based on the comparison between the transaction channel number and the reference channel number included in the reference channel number list, determining whether the transaction channel number has an anomaly of a first anomaly type; Determine whether the request serial number has an exception of a second exception type based on the format of the request serial number, the first identification code, and the requester system number included in the request serial number; Determine whether the global serial number has an anomaly of a third anomaly type based on the format of the global serial number, the second identification code, and the original requester system number included in the global serial number; Based on the format of the transaction return code and the system number included in the transaction return code, determining whether the transaction return code has an exception of a fourth exception type; Based on the friendliness of the content included in the transaction return information, determining whether the transaction return information has an anomaly of a fifth anomaly type; If at least one of the first exception type, the second exception type, the third exception type, the fourth exception type, and the fifth exception type exists, it is determined that the transaction message has the matching exception situation.
[0007] In a possible implementation manner of the first aspect, determining whether the transaction return information has an anomaly of a fifth anomaly type based on the friendliness of the content included in the transaction return information includes: Pre-building a sample vocabulary library, the sample vocabulary library includes sample vocabulary and the friendliness of the sample vocabulary, the sample vocabulary library is used to compare the content included in the transaction return information to determine the friendliness of the content, the sample vocabulary comes from the content of the historical transaction return information included in the historical transaction return information library, and the friendliness of the sample vocabulary is pre-marked; Performing word segmentation processing on the content included in the transaction return information to obtain a transaction return information word segmentation set corresponding to the content, wherein the transaction return information word segmentation set includes at least one transaction return information word segmentation; According to the sample vocabulary library and the transaction return information segmentation set, obtaining the friendliness of the transaction return information segmentation in the transaction return information segmentation set; If the friendliness of the content is classified as low friendliness, or if the friendliness of the content is less than or equal to a preset friendliness threshold, it is determined that the transaction return information has an anomaly of the fifth anomaly type.
[0008] In a possible implementation of the first aspect, acquiring the friendliness of the transaction return information segmentation in the transaction return information segmentation set according to the sample vocabulary library and the transaction return information segmentation set includes: Processing the transaction return information word set using a pre-trained natural language processing model, and converting the word sets in the transaction return information word set into word vectors; Acquire the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library by semantic feature extraction; Calculating the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library, wherein the similarity is used to characterize the similarity between the word segment corresponding to the word vector and the sample vocabulary; Determine a first sample vocabulary whose semantic feature similarity with the word vector is greater than or equal to a preset similarity threshold, and obtain the number of the first sample vocabulary; If the number of the first sample words is greater than or equal to a preset number threshold, the friendliness of the transaction return information segmentation words in the transaction return information segmentation set is determined by the friendliness of the first sample words and a weighted operation of the similarity.
[0009] In a possible implementation of the first aspect, the calculating the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library includes: Acquire semantic features of the word vector through semantic feature extraction, and hierarchical classification corresponding to the semantic features, wherein the hierarchical classification is determined according to banking transaction business knowledge, and the hierarchical classification includes business type, business status, business subject, business result, etc.; According to the hierarchical classification, obtaining the friendliness influence weight of the hierarchical classification, wherein the friendliness influence weight of different hierarchical classifications is different; Analyzing the hierarchical classification corresponding to the semantic features of the word vector and the hierarchical classification corresponding to the semantic features of the sample vocabulary, and determining the hierarchical classification corresponding to the semantic features of the word vector and the hierarchical classification corresponding to the semantic features of the sample vocabulary; According to the hierarchical classification corresponding to the semantic features of the word vector and the hierarchical classification corresponding to the semantic features of the sample vocabulary, determining the positional relationship between the semantic features of the word vector and the semantic features of the sample vocabulary in the hierarchical classification; Based on the positional relationship and the friendliness influence weight of the hierarchical classification, determining an adjustment factor for calculating the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library; The similarity between the semantic features of the word vector and the semantic features of the sample words in the sample vocabulary library is calculated according to the adjustment factor, the semantic features of the word vector, and the semantic features of the sample words in the sample vocabulary library.
[0010] In a possible implementation of the first aspect, before calculating the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library, the method further includes: Perform semantic expansion on the segmented words of the transaction return information and the sample words in the sample word library according to the semantic expansion resource and the rule engine to generate a first semantic expansion set of the segmented words of the transaction return information and a second semantic expansion set of the sample words; Converting the extended word segments in the first semantic extension set and the extended sample word segments in the second semantic extension set into word vectors; The calculating the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library includes: Calculating the similarity between the semantic features of the word vector and the semantic features of the sample words in the sample vocabulary library; Calculating the similarity between the word vector of the extended word in the first semantic extension set and the word vector of the extended sample word in the second semantic extension set; Based on the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library, and the fusion calculation of the similarity between the word vector of the extended word segmentation in the first semantic extension set and the word vector of the extended sample word segmentation in the second semantic extension set, the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library is generated.
[0011] In a possible implementation of the first aspect, if the matching abnormality exists, the method further includes: storing abnormal transaction messages with the matching abnormality and abnormal contents of the abnormal transaction messages; Summarize the stored abnormal transaction messages according to the preset batch running cycle and generate an abnormal query file; Obtaining, according to the abnormal query file, an abnormal trend and / or an abnormal proportion in a first batch running cycle corresponding to the abnormal query file; The abnormal trend and / or abnormal proportion is output.
[0012] In combination with the second aspect of the present application, a bank transaction data processing system is provided, which includes a machine-readable storage medium and a processor, wherein the machine-readable storage medium stores machine-executable instructions, and when the processor executes the machine-executable instructions, the bank transaction data processing system implements the aforementioned bank transaction data processing method.
[0013] In conjunction with the third aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed, the aforementioned bank transaction data processing method is implemented.
[0014] In conjunction with the fourth aspect of the present application, a computer program product is provided, and when the computer program is executed by a processor, the aforementioned bank transaction data processing method is implemented.
[0015] In combination with any of the above aspects, by obtaining the transaction message corresponding to each bank transaction detail from the bank transaction log, the transaction message includes the request serial number, the global serial number, the system number, the transaction channel number, the transaction return code, and the transaction return information; based on the preset verification rules, the request serial number, the global serial number, the system number, the transaction channel number, the transaction return code, and the transaction return information in the transaction message are verified to determine whether there is a matching abnormality; if there is a matching abnormality, the abnormal content corresponding to the matching abnormality is obtained; according to the type of the abnormal content and the source of the transaction message, an abnormal transaction message tracking sheet is generated and output. This improves the verification efficiency of bank transaction data. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained by combining these drawings without paying creative work.
[0017] Figure 1 A flowchart of a method for processing bank transaction data provided in an embodiment of the present application. DETAILED DESCRIPTION
[0018] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all of the embodiments. According to the embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0019] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, device, product or end that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units inherent to these processes, methods, products or ends.
[0020] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.
[0021] Figure 1 The flowchart of the bank transaction data processing method provided by the embodiment of the present application is shown. It should be understood that in other embodiments, the order of some steps in the bank transaction data processing method of this embodiment can be shared with each other according to actual needs, or some steps can be omitted or maintained. The details of the bank transaction data processing method include: Step S110, obtaining the transaction message corresponding to each bank transaction detail from the bank transaction log.
[0022] Among them, the transaction message includes a request serial number, a global serial number, a system number, a transaction channel number, a transaction return code, and a transaction return information. The system number includes a requesting party system number corresponding to the request serial number and an original requesting party system number corresponding to the global serial number.
[0023] In this embodiment, the bank's server will read transaction data from the bank's transaction log regularly or in real time. The bank's transaction log is a detailed record of various bank transaction activities, including numerous transaction details. When reading, the server will extract the corresponding transaction message for each transaction detail.
[0024] A transaction message is a data structure that contains rich transaction information. Among them, the request serial number is a unique number used to identify each transaction request, which plays a role in tracking requests throughout the entire transaction process. For example, when a user initiates a transfer request through online banking, the system will generate a unique request serial number for this request. Exemplarily, the request serial number is a 35-digit mixed number of numbers and letters, and its format includes the request serial number identifier (1 digit): R, system number (4 digits), transaction date (8 digits, yyyymmdd), transaction time (6 digits, HHMMSS), and a 16-digit unique serial number.
[0025] The global serial number identifies the transaction from a more macro perspective. It is associated with the entire transaction chain. No matter how many systems or links the transaction passes through, the global serial number remains unchanged. The corresponding original requester system number indicates which system originally initiated the transaction. Exemplarily, the global serial number is a mixed number of 35 digits and letters, and its format includes the global serial number identifier (1 digit): R, system number (4 digits), transaction date (8 digits, yyyymmdd), transaction time (6 digits, HHMMSS), and a 16-digit unique serial number.
[0026] The system number includes the requester system number corresponding to the request serial number and the original requester system number corresponding to the global serial number. Different system numbers represent different business systems within the bank, such as the core business system, online banking system, etc. Exemplarily, the system number consists of 4 characters, including 0-9, and is numbered sequentially starting from 0001.
[0027] The transaction channel number reflects the way in which the transaction occurred, such as whether the transaction was conducted through an ATM (automatic teller machine), online banking, mobile banking or a bank counter.
[0028] The transaction return code is a code returned by the system after the transaction is completed, which is used to indicate the result status of the transaction. Different codes correspond to different transaction results, such as successful transaction, failed transaction (may be due to insufficient balance, account freezing, etc.). For example, when the transaction is successful, the transaction return code is 000000, and the other error code rules are: E system number-error category (1 letter)-error number (6 letters / numbers), for example: E0138-B-A30003.
[0029] The transaction return information is a text description of the transaction results, which informs users or relevant staff of the transaction status in a more intuitive way.
[0030] The server obtains and analyzes these transaction messages through a dedicated data reading interface and data analysis module. The data reading interface is responsible for obtaining raw data from the database or storage system that stores bank transaction logs, and the data analysis module analyzes the obtained raw data according to the pre-defined message format and extracts key information such as the request serial number, global serial number, system number, transaction channel number, transaction return code, and transaction return information.
[0031] Step S120, based on preset verification rules, the request serial number, the global serial number, the system number, the transaction channel number, the transaction return code, and the transaction return information in the transaction message are verified to determine whether there is any matching anomaly.
[0032] In this embodiment, the server can verify the request serial number, the global serial number, the system number, the transaction channel number, the transaction return code, and the transaction return information by using different verification rules for the request serial number, the global serial number, the system number, the transaction channel number, the transaction return code, and the transaction return information to determine whether there are any matching anomalies in the above contents.
[0033] How the server performs verification based on the preset verification rules can be referred to the detailed description below.
[0034] Step S130: If the matching abnormality exists, obtain the abnormal content corresponding to the matching abnormality.
[0035] The abnormal content includes the type of the abnormal content.
[0036] In this embodiment, after the server determines in step S120 that there is a matching anomaly, it further analyzes the specific content of the anomaly and determines its type.
[0037] If there is an abnormality in the transaction channel number, the server will record the value of the abnormal transaction channel number, such as "UNKNOWN". The type of abnormal content is marked as "invalid channel number". At the same time, the server will query the relevant transaction records and obtain other transaction information related to this channel number, such as transaction time, request serial number, etc., in order to more fully understand the abnormal situation.
[0038] For abnormal request serial numbers, the server will record the error content of the request serial number in detail, such as format error (where the format requirement is not met) or requester system number error (what is the specific wrong system number). The type of abnormal content is marked as "request serial number abnormality", which may be further subdivided into "format error" or "system number error". Then, the server will trace the transaction origin corresponding to the request serial number to check whether there is other related abnormal information or operation records.
[0039] When there is an exception in the global serial number, the server will record the error of the global serial number, such as the specific content of the format not matching or the original requester's system number error. The type of abnormal content is marked as "global serial number abnormality", which can also be subdivided into corresponding subtypes. In addition, the server will check all links related to this global serial number along the transaction link to find out the possible causes of the abnormality.
[0040] When the transaction return code is abnormal, the server will record the value of the incorrect transaction return code and the error type (such as out of range or format error) determined based on the system number. The type of abnormal content is marked as "transaction return code abnormal". At the same time, the server will check the transaction operation corresponding to this transaction return code and analyze which business link has the problem.
[0041] For abnormal transaction return information, the server will record the content of the transaction return information and the calculated friendliness. The type of abnormal content is marked as "unfriendly transaction return information" or further refined according to the specific situation, such as "ambiguous information meaning" and "negative information content". The server will also check the transaction business type corresponding to the transaction return information to see whether this type of abnormality often occurs in a specific business type.
[0042] Step S140: generating an abnormal transaction message tracking sheet according to the type of the abnormal content and the source of the transaction message.
[0043] In this embodiment, the server first determines the basic framework of the abnormal transaction message tracking sheet, which includes some fixed information fields, such as the tracking sheet number (which can be a unique number generated according to certain rules, such as "20230915001" generated according to the date and sequence number), the generation time (accurate to seconds, such as "2023-09-1515:30:00"), etc.
[0044] According to the type of abnormal content, the server will fill in the abnormal content obtained in step S130 in detail in the tracking sheet. For example, if the abnormal content type is "illegal transaction channel number", the tracking sheet will clearly state "abnormal type: illegal transaction channel number, abnormal channel number: UNKNOWN, relevant transaction time: 2023-09-1514:00:00, request serial number: REQ2023091500001" and other information.
[0045] If the transaction return information is abnormal, such as "the transaction return information is unfriendly - the information is negative", the "Exception type: The transaction return information is unfriendly - the information is negative, the transaction return information: system failure, the transaction cannot be carried out, please try again, friendliness: 30, corresponding business type: online payment business" and other contents will be filled in the tracking form.
[0046] The server will record the source of the transaction message in detail. If the transaction message comes from the online banking system, the tracking sheet will indicate "Message source: online banking system, possible source IP address: XXX.XXX.X.XXX, related system module: transfer module" and other information. This helps to locate whether the problem is at the message generation end or during the transmission process.
[0047] The server can also find other transaction messages or transaction information related to the abnormal transaction message according to the bank's internal transaction association rules, and associate them with the tracking sheet. For example, if the abnormal transaction is part of a transfer business, it may be associated with the transfer recipient information, transfer amount and other related information to more comprehensively analyze the abnormal situation.
[0048] Step S150: outputting the abnormal transaction message tracking sheet.
[0049] Based on the above steps, the transaction message corresponding to each bank transaction detail is obtained from the bank transaction log, wherein the transaction message includes the request serial number, the global serial number, the system number, the transaction channel number, the transaction return code, and the transaction return information; based on the preset verification rules, the request serial number, the global serial number, the system number, the transaction channel number, the transaction return code, and the transaction return information in the transaction message are verified to determine whether there is a matching abnormality; if there is a matching abnormality, the abnormal content corresponding to the matching abnormality is obtained; according to the type of the abnormal content and the source of the transaction message, an abnormal transaction message tracking sheet is generated and output. This improves the verification efficiency of bank transaction data.
[0050] In a possible implementation, the request serial number, the global serial number, the system number, the transaction channel number, the transaction return code, and the transaction return information in the transaction message are verified based on a preset verification rule to determine whether there is a matching anomaly, including: Based on the comparison between the transaction channel number and the reference channel number included in the reference channel number list, determining whether the transaction channel number has an anomaly of a first anomaly type; Determine whether the request serial number has an exception of a second exception type based on the format of the request serial number, the first identification code, and the requester system number included in the request serial number; Determine whether the global serial number has an anomaly of a third anomaly type based on the format of the global serial number, the second identification code, and the original requester system number included in the global serial number; Based on the format of the transaction return code and the system number included in the transaction return code, determining whether the transaction return code has an exception of a fourth exception type; Based on the friendliness of the content included in the transaction return information, determining whether the transaction return information has an anomaly of a fifth anomaly type; If at least one of the first exception type, the second exception type, the third exception type, the fourth exception type, and the fifth exception type exists, it is determined that the transaction message has the matching exception situation.
[0051] In this embodiment, the server obtains the transaction channel number in the transaction message. Then, the server obtains all reference channel numbers from a pre-stored reference channel number list. The server compares the transaction channel number in the transaction message with the reference channel numbers in the reference channel number list one by one. If the transaction channel number is not in the reference channel number list, it is determined that the transaction channel number has an anomaly of the first anomaly type.
[0052] The server obtains the format requirements (such as the specified length, character type, etc.) of the request serial number in the transaction message, the preset first identification code, and the requester system number contained in the request serial number. Then, the server checks whether the request serial number meets the format requirements, whether it contains the correct first identification code, and whether the requester system number is correct. If the format of the request serial number does not meet the requirements, or does not contain the correct first identification code, or the requester system number is wrong, then it is determined that the request serial number has an abnormality of the second abnormality type.
[0053] The server obtains the format requirements of the global serial number in the transaction message, the pre-set second identification code, and the original requester system number contained in the global serial number. Next, the server checks whether the global serial number meets the format requirements, whether it contains the correct second identification code, and whether the original requester system number is correct. If the format of the global serial number does not meet the requirements, does not contain the correct second identification code, or the original requester system number is wrong, it is determined that the global serial number has an exception of the third exception type.
[0054] The server obtains the format requirements of the transaction return code and the system number contained therein in the transaction message. The server checks whether the transaction return code meets the format requirements and whether the system number contained therein is correct. If the transaction return code does not meet the format requirements or the system number is incorrect, it is determined that the transaction return code has an exception of the fourth exception type.
[0055] The server first constructs a sample vocabulary library, which contains sample words obtained from the historical transaction return information library and their pre-marked friendliness. Then, the transaction return information is segmented to obtain a transaction return information segmentation set. Then, the friendliness of each segmentation in the transaction return information segmentation set is determined based on the sample vocabulary library, thereby calculating the friendliness of the transaction return information content. If the friendliness of the content is classified as low friendliness or less than or equal to the preset friendliness threshold, it is determined that the transaction return information has an abnormality of the fifth abnormality type.
[0056] The above-mentioned verification process for different fields can be cross-checked. For example, after verifying the system number, when verifying the global serial number, the system number included in the global serial number can be verified again. This application does not impose any restrictions on this.
[0057] In a possible implementation manner, determining whether the transaction return information has an anomaly of the fifth anomaly type based on the friendliness of the content included in the transaction return information includes: Pre-building a sample vocabulary library, the sample vocabulary library includes sample vocabulary and the friendliness of the sample vocabulary, the sample vocabulary library is used to compare the content included in the transaction return information to determine the friendliness of the content, the sample vocabulary comes from the content of the historical transaction return information included in the historical transaction return information library, and the friendliness of the sample vocabulary is pre-marked; Performing word segmentation processing on the content included in the transaction return information to obtain a transaction return information word segmentation set corresponding to the content, wherein the transaction return information word segmentation set includes at least one transaction return information word segmentation; According to the sample vocabulary library and the transaction return information segmentation set, obtaining the friendliness of the transaction return information segmentation in the transaction return information segmentation set; If the friendliness of the content is classified as low friendliness, or if the friendliness of the content is less than or equal to a preset friendliness threshold, it is determined that the transaction return information has an anomaly of the fifth anomaly type.
[0058] In this embodiment, the server first accesses the historical transaction return information library and extracts various words from the content of the numerous historical transaction return information. For each extracted word, its friendliness is marked according to business experience and understanding of the transaction situation. For example, words such as "success", "normal", "processed", etc., which indicate a smooth transaction, are marked with high friendliness; while words such as "failure", "error", "not found", etc., which indicate a problem with the transaction, are marked with low friendliness. These words and their corresponding friendliness are combined to construct a sample vocabulary library.
[0059] The server uses the pre-selected word segmentation tool to perform word segmentation on the content in the transaction return information. The tool divides the continuous text content into independent words according to certain rules (such as the semantics and part of speech of Chinese characters), and these words constitute the word segmentation set of the transaction return information.
[0060] The server traverses each word in the word set of the transaction return information. For each word, it searches the sample vocabulary to see if there is the same word. If so, it obtains the friendliness of the word in the sample vocabulary. If there is no identical word in the sample vocabulary, it may be necessary to determine the friendliness based on rules such as semantic similarity (for example, through a pre-defined semantically similar word mapping relationship).
[0061] After obtaining the friendliness of each word in the transaction return information word set, the server calculates the friendliness of the entire transaction return information content according to a certain algorithm. For example, the friendliness classification of the content can be simply determined based on the number of word segments and the proportion of low-friendliness word segments. If the proportion of low-friendliness word segments is relatively large, the friendliness of the content is classified as low-friendliness; or the friendliness of each word is weighted and summed (the weight is set according to the importance of the word in the content, etc.) to obtain a specific friendliness value. If this value is less than or equal to the preset friendliness threshold, it is determined that the transaction return information has an anomaly of the fifth anomaly type.
[0062] In a possible implementation, acquiring the friendliness of the transaction return information segmentation words in the transaction return information segmentation word set according to the sample vocabulary library and the transaction return information segmentation word set includes: Processing the transaction return information word set using a pre-trained natural language processing model, and converting the word sets in the transaction return information word set into word vectors; Acquire the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library by semantic feature extraction; Calculating the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library, wherein the similarity is used to characterize the similarity between the word segment corresponding to the word vector and the sample vocabulary; Determine a first sample vocabulary whose semantic feature similarity with the word vector is greater than or equal to a preset similarity threshold, and obtain the number of the first sample vocabulary; If the number of the first sample words is greater than or equal to a preset number threshold, the friendliness of the transaction return information segmentation words in the transaction return information segmentation set is determined by the friendliness of the first sample words and a weighted operation of the similarity.
[0063] In this embodiment, the server loads a pre-trained natural language processing model (such as a Word2Vec model, etc.). Then, each word in the transaction return information word set is input into the model in turn. The model processes each input word according to its internal neural network structure and training parameters, and converts it into a word vector of fixed dimension. The word vector can represent the semantic information of the word to a certain extent.
[0064] The server uses a special semantic feature extraction tool or algorithm. For word vectors, the tool extracts its semantic features based on the numerical distribution of the word vectors and pre-defined semantic feature extraction rules. For example, semantic features representing business type, sentiment tendency, etc. may be extracted from word vectors. For sample words in the sample vocabulary library, the server can convert each sample word into a word vector, and then use the semantic feature extraction tool to extract its semantic features.
[0065] The server uses a suitable similarity calculation method, such as the cosine similarity calculation method. For the semantic features of the word vector of each word in the transaction return information word set and the semantic features of each sample word in the sample vocabulary library, the cosine similarity between them is calculated. The value of cosine similarity is between -1 and 1. The closer the value is to 1, the more similar the two semantic features are.
[0066] The server sets a preset similarity threshold (for example, 0.8). For each transaction return information word vector, after calculating its semantic feature similarity with the sample words in the sample vocabulary library, those sample words with similarity greater than or equal to 0.8 are selected. These sample words are the first sample words. Then the number of first sample words is counted.
[0067] The server sets a preset number threshold (e.g., 1). If the number of first sample words is greater than or equal to 1, then for each first sample word, the server obtains its friendliness and the semantic feature similarity between the first sample word and the corresponding transaction return information segmentation. Then, the friendliness of the transaction return information segmentation is calculated according to a weighted operation formula (e.g., friendliness multiplied by the weighted sum of similarity).
[0068] In a possible implementation, the calculating the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library includes: Acquire semantic features of the word vector through semantic feature extraction, and hierarchical classification corresponding to the semantic features, wherein the hierarchical classification is determined according to banking transaction business knowledge, and the hierarchical classification includes business type, business status, business subject, business result, etc.; According to the hierarchical classification, obtaining the friendliness influence weight of the hierarchical classification, wherein the friendliness influence weight of different hierarchical classifications is different; Analyzing the hierarchical classification corresponding to the semantic features of the word vector and the hierarchical classification corresponding to the semantic features of the sample vocabulary, and determining the hierarchical classification corresponding to the semantic features of the word vector and the hierarchical classification corresponding to the semantic features of the sample vocabulary; According to the hierarchical classification corresponding to the semantic features of the word vector and the hierarchical classification corresponding to the semantic features of the sample vocabulary, determining the positional relationship between the semantic features of the word vector and the semantic features of the sample vocabulary in the hierarchical classification; Based on the positional relationship and the friendliness influence weight of the hierarchical classification, determining an adjustment factor for calculating the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library; The similarity between the semantic features of the word vector and the semantic features of the sample words in the sample vocabulary library is calculated according to the adjustment factor, the semantic features of the word vector, and the semantic features of the sample words in the sample vocabulary library.
[0069] In this embodiment, the server uses a pre-built semantic feature extraction model or algorithm. For word vectors, this model will parse the semantic information in the word vector into different hierarchical classifications based on the knowledge of banking transaction business. For example, for the word vector representing "transfer", the model may parse out its business type as "fund transfer", business status as "in progress" (assuming it is in a specific context), business subject as "user", business result as "undetermined" and other hierarchical classifications, and simultaneously extract semantic features related to these hierarchical classifications.
[0070] The server pre-sets the friendliness impact weights of different hierarchical categories according to the bank transaction business rules. For example, the business result hierarchical category may have a higher impact weight on friendliness because it directly reflects the success of the transaction; while the business subject may have a lower impact weight on friendliness because it is mainly about the roles involved in the transaction. These weights can be stored in a configuration file or database table, and the server reads the corresponding weight value when performing calculations.
[0071] The server traverses the hierarchical classification corresponding to each semantic feature of the word vector, and traverses the hierarchical classification corresponding to the semantic features of the sample vocabulary. For example, for the word vector representing "withdrawal" (the hierarchical classification mentioned above) and the sample vocabulary "successful deposit", the server will find out their respective situations under the hierarchical classifications of business type, business status, business subject, business result, etc.
[0072] The server compares the semantic features of the word vector and the sample vocabulary in each hierarchical classification. For example, in the business type hierarchical classification, if the word vector is "fund withdrawal" and the sample vocabulary is "fund deposit", their positional relationship in the business type hierarchical classification is a relationship between different types of fund operations; in the business result hierarchical classification, if the word vector is "unfinished" and the sample vocabulary is "successful", it is a relationship between different business result states.
[0073] The server determines the adjustment factor based on the position relationship of the hierarchical classification and the weight of the friendliness impact according to the pre-defined rules. For example, if in the business result hierarchical classification, the word vector is "unfinished" and the sample word is "successful", the adjustment factor may be a small value (such as 0.2) because the business result weight is high and the difference between the two is large; if the two are the same in the business subject hierarchical classification and the business subject weight is low, the adjustment factor may be close to 1 (such as 0.9).
[0074] The server uses an appropriate similarity calculation method (such as cosine similarity, etc.) and combines it with an adjustment factor to calculate the final similarity. For example, the original similarity between the word vector and the sample vocabulary is first calculated according to the conventional cosine similarity, and then the original similarity is multiplied by the adjustment factor to obtain the final similarity.
[0075] In a possible implementation, before calculating the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library, the method further includes: Perform semantic expansion on the segmented words of the transaction return information and the sample words in the sample word library according to the semantic expansion resource and the rule engine to generate a first semantic expansion set of the segmented words of the transaction return information and a second semantic expansion set of the sample words; Converting the extended word segments in the first semantic extension set and the extended sample word segments in the second semantic extension set into word vectors; The calculating the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library includes: Calculating the similarity between the semantic features of the word vector and the semantic features of the sample words in the sample vocabulary library; Calculating the similarity between the word vector of the extended word in the first semantic extension set and the word vector of the extended sample word in the second semantic extension set; Based on the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library, and the fusion calculation of the similarity between the word vector of the extended word segmentation in the first semantic extension set and the word vector of the extended sample word segmentation in the second semantic extension set, the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library is generated.
[0076] In this embodiment, the semantic extension resource of the server can be a database containing semantic relationships such as synonyms, near-synonyms, and related words. The rule engine determines how to perform semantic extension based on predefined rules. For each word segmentation in the transaction return information, the server searches for related words from the semantic extension resource and determines whether to add these words to the first semantic extension set based on the rule engine. Similarly, for each sample word in the sample vocabulary library, the second semantic extension set is constructed in the same way. For example, if the semantic extension resource contains synonyms of "payment" such as "payment" and "delivery", and the rule engine stipulates that as long as it is a synonym, it will be expanded, then for the word segmentation "payment" in the transaction return information, "payment" and "delivery" will be added to the first semantic extension set.
[0077] The server uses a pre-trained word vector model (such as Word2Vec, GloVe, etc.). For each extended word in the first semantic extension set, it is input into the word vector model, and the model converts the extended word into a word vector of fixed dimension according to its internal parameters and algorithms. Similarly, the same operation is performed on each extended sample word in the second semantic extension set.
[0078] For calculating the similarity between the semantic features of the word vector and the semantic features of the sample words in the sample vocabulary library, the server adopts an appropriate similarity calculation method, such as cosine similarity. For the original word vector (before expansion), the cosine similarity between them is calculated. For the similarity between the word vector of the expanded word in the first semantic expansion set and the word vector of the expanded sample word in the second semantic expansion set, the cosine similarity calculation method is also used to calculate the similarity between the expanded word and the expanded sample word one by one.
[0079] The server determines a fusion calculation method, such as weighted average. First, the weights of the original word vector similarity and the expanded word vector similarity are determined. Assume that the weight of the original word vector similarity is 0.6 and the weight of the expanded word vector similarity is 0.4. Then, the original word vector similarity is multiplied by its weight, and the expanded word vector similarity is multiplied by its weight. Finally, the two are added together to obtain the fused similarity.
[0080] In a possible implementation manner, if the matching abnormality exists, the method further includes: storing abnormal transaction messages with the matching abnormality and abnormal contents of the abnormal transaction messages; Summarize the stored abnormal transaction messages according to the preset batch running cycle and generate an abnormal query file; Obtaining, according to the abnormal query file, an abnormal trend and / or an abnormal proportion in a first batch running cycle corresponding to the abnormal query file; The abnormal trend and / or abnormal proportion is output.
[0081] In this embodiment, when the server detects a transaction message matching an abnormal situation, it will store the entire abnormal transaction message in a dedicated abnormal transaction storage area (such as a specific table in a database or a specific folder in a file system). At the same time, the abnormal content of the abnormal transaction message (including abnormal type, related transaction data fields, etc.) is also stored, and an association relationship with the abnormal transaction message is established for subsequent query and analysis.
[0082] The server pre-sets the batch running cycle, for example, it starts the batch running at 2 am every day. When the batch running cycle arrives, the server reads all abnormal transaction messages stored in the batch running cycle from the area where abnormal transaction messages are stored (such as the database table or folder mentioned above). Then, these abnormal transaction messages and their related abnormal contents are summarized and sorted in a certain format (such as CSV format or a custom format) to generate an abnormal query file. This file can contain key information of the transaction message (such as transaction serial number, transaction time, etc.) and a brief description of the abnormal content.
[0083] To obtain the abnormal trend, the server can analyze the distribution of abnormal transaction messages in the abnormal query file over time during the batch cycle. For example, by counting the number of abnormal transaction messages that appear in each hour period, observe whether the number of abnormalities gradually increases, decreases, or fluctuates, and thus determine the abnormal trend. To obtain the abnormal proportion, the server first counts the total number of transaction messages in the batch cycle (which can be obtained from the transaction log or other statistical data sources), and then counts the number of abnormal transaction messages in the abnormal query file, and divides the number of abnormal transaction messages by the total number of transaction messages to obtain the abnormal proportion.
[0084] In the above embodiments, the bank transaction data processing system for executing the above method embodiments has at least one processor, a control module (chip set) coupled to at least one of the (at least one) processors, a memory coupled to the control module, a non-volatile memory (NVM) / storage device coupled to the control module, at least one input / output device coupled to the control module, and a network interface coupled to the control module.
[0085] The processor may include at least one single-core or multi-core processor, and the processor may include any combination of general-purpose processors or special-purpose processors (such as graphics processors, application processors, baseband processors, etc.). For some alternative implementations, the bank transaction data processing system can be used as a bank transaction data processing system device such as a gateway described in the embodiments of the present application.
[0086] For some alternative embodiments, the bank transaction data processing system may include at least one computer-readable medium (e.g., a memory or NVM / storage device) having instructions and at least one processor integrated with the at least one computer-readable medium and configured to execute the instructions to implement a module to perform the actions described in the present disclosure.
[0087] For one embodiment, the control module may include any suitable interface controller to provide any suitable interface to at least one of the processor(s) and / or any suitable device or component in communication with the control module.
[0088] The control module may include a memory controller module to provide an interface to the memory. The memory controller module may be a hardware module, a software module, and / or a firmware module.
[0089] The memory may be used, for example, to load and store data and / or instructions for a banking transaction data processing system. For one embodiment, the memory may include any suitable volatile memory, for example, a suitable DRAM.
[0090] For one embodiment, the control module may include at least one input / output controller to provide an interface to the NVM / storage device and (at least one) input / output device.
[0091] For example, the NVM / storage device may be used to store data and / or instructions. The NVM / storage device may include any suitable non-volatile memory (e.g., flash memory) and / or may include any suitable (at least one) non-volatile storage device (e.g., at least one hard disk drive (HDD), at least one compact disk (CD) drive, and / or at least one digital versatile disk (DVD) drive).
[0092] The NVM / storage device may include storage resources that are physically part of the device on which the bank transaction data processing system is installed, or it may be accessible to the device without being part of the device. For example, the NVM / storage device may be accessed via (at least one) input / output device over a network.
[0093] The (at least one) input / output device may provide an interface for the bank transaction data processing system to communicate with any other appropriate device, and the input / output device may include a communication component, a phonetic component, a sensor component, etc. The network interface may provide an interface for the bank transaction data processing system to communicate according to at least one network, and the bank transaction data processing system may wirelessly communicate with at least one component of a wireless network according to any of at least one wireless network standard and / or protocol, for example, access a wireless network according to a communication standard.
[0094] For one embodiment, at least one of the (at least one) processor may be loaded together with the logic of at least one controller of the control module (e.g., a memory controller module). For one embodiment, at least one of the (at least one) processor may be loaded together with the logic of at least one controller of the control module to form a system level load. For one embodiment, at least one of the (at least one) processor may be integrated on the same die with the logic of at least one controller of the control module. For one embodiment, at least one of the (at least one) processor may be integrated on the same die with the logic of at least one controller of the control module to form a system on chip (SoC).
[0095] The embodiments of the present application are introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core idea of the present application. At the same time, for general technical personnel in this field, according to the idea of the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.
[0096] An embodiment of the present invention discloses a computer-readable storage medium storing a computer program for electronic data exchange, wherein the computer program enables a computer to execute the steps of the bank transaction data processing method described in the above embodiment.
[0097] An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to enable a computer to execute the steps in the bank transaction data processing method described in the above embodiment.
[0098] The device embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, that is, they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.
[0099] Through the specific description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution can be essentially or partly contributed to the prior art in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other medium that can be used to be computer-readable with or store data.
[0100] Finally, it should be noted that what is disclosed above is only a preferred embodiment of the present invention, which is only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.
Claims
1. A method for processing bank transaction data, characterized in that: include: Obtaining transaction messages corresponding to the transaction details of each bank from the bank transaction log, wherein the transaction message includes a request serial number, a global serial number, a system number, a transaction channel number, a transaction return code, and transaction return information, wherein the system number includes a requester system number corresponding to the request serial number and an original requester system number corresponding to the global serial number; Based on the preset verification rules, the request serial number, the global serial number, the system number, the transaction channel number, the transaction return code, and the transaction return information in the transaction message are verified to determine whether there is a matching anomaly; If the matching abnormality exists, obtaining abnormal content corresponding to the matching abnormality, wherein the abnormal content includes the type of the abnormal content; Generate an abnormal transaction message tracking sheet according to the type of the abnormal content and the source of the transaction message; Output the abnormal transaction message tracking sheet.
2. The bank transaction data processing method according to claim 1, characterized in that: The verification of the request serial number, the global serial number, the system number, the transaction channel number, the transaction return code, and the transaction return information in the transaction message based on a preset verification rule to determine whether there is a matching anomaly includes: Based on the comparison between the transaction channel number and the reference channel number included in the reference channel number list, determining whether the transaction channel number has an anomaly of a first anomaly type; Determine whether the request serial number has an exception of a second exception type based on the format of the request serial number, the first identification code, and the requester system number included in the request serial number; Determine whether the global serial number has an anomaly of a third anomaly type based on the format of the global serial number, the second identification code, and the original requester system number included in the global serial number; Based on the format of the transaction return code and the system number included in the transaction return code, determining whether the transaction return code has an exception of a fourth exception type; Based on the friendliness of the content included in the transaction return information, determining whether the transaction return information has an anomaly of a fifth anomaly type; If at least one of the first exception type, the second exception type, the third exception type, the fourth exception type, and the fifth exception type exists, it is determined that the transaction message has the matching exception situation.
3. The bank transaction data processing method according to claim 2, characterized in that: The determining, based on the friendliness of the content included in the transaction return information, whether the transaction return information has an anomaly of a fifth anomaly type includes: Pre-building a sample vocabulary library, the sample vocabulary library includes sample vocabulary and the friendliness of the sample vocabulary, the sample vocabulary library is used to compare the content included in the transaction return information to determine the friendliness of the content, the sample vocabulary comes from the content of the historical transaction return information included in the historical transaction return information library, and the friendliness of the sample vocabulary is pre-marked; Performing word segmentation processing on the content included in the transaction return information to obtain a transaction return information word segmentation set corresponding to the content, wherein the transaction return information word segmentation set includes at least one transaction return information word segmentation; According to the sample vocabulary library and the transaction return information segmentation set, obtaining the friendliness of the transaction return information segmentation in the transaction return information segmentation set; If the friendliness of the content is classified as low friendliness, or if the friendliness of the content is less than or equal to a preset friendliness threshold, it is determined that the transaction return information has an anomaly of the fifth anomaly type.
4. The bank transaction data processing method according to claim 3, characterized in that: The acquiring, based on the sample vocabulary library and the transaction return information segmentation set, the friendliness of the transaction return information segmentation in the transaction return information segmentation set includes: Processing the transaction return information word set using a pre-trained natural language processing model, and converting the word sets in the transaction return information word set into word vectors; Acquire the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library by semantic feature extraction; Calculating the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library, wherein the similarity is used to characterize the similarity between the word segment corresponding to the word vector and the sample vocabulary; Determine a first sample vocabulary whose semantic feature similarity with the word vector is greater than or equal to a preset similarity threshold, and obtain the number of the first sample vocabulary; If the number of the first sample words is greater than or equal to a preset number threshold, the friendliness of the transaction return information segmentation words in the transaction return information segmentation set is determined by the friendliness of the first sample words and a weighted operation of the similarity.
5. The method for processing bank transaction data according to claim 4, characterized in that: The calculating the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library includes: Acquire semantic features of the word vector through semantic feature extraction, and hierarchical classification corresponding to the semantic features, wherein the hierarchical classification is determined according to banking transaction business knowledge, and the hierarchical classification includes business type, business status, business subject, business result, etc.; According to the hierarchical classification, obtaining the friendliness influence weight of the hierarchical classification, wherein the friendliness influence weight of different hierarchical classifications is different; Analyzing the hierarchical classification corresponding to the semantic features of the word vector and the hierarchical classification corresponding to the semantic features of the sample vocabulary, and determining the hierarchical classification corresponding to the semantic features of the word vector and the hierarchical classification corresponding to the semantic features of the sample vocabulary; According to the hierarchical classification corresponding to the semantic features of the word vector and the hierarchical classification corresponding to the semantic features of the sample vocabulary, determining the positional relationship between the semantic features of the word vector and the semantic features of the sample vocabulary in the hierarchical classification; Based on the positional relationship and the friendliness influence weight of the hierarchical classification, determining an adjustment factor for calculating the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library; The similarity between the semantic features of the word vector and the semantic features of the sample words in the sample vocabulary library is calculated according to the adjustment factor, the semantic features of the word vector, and the semantic features of the sample words in the sample vocabulary library.
6. The method for processing bank transaction data according to claim 4, characterized in that: Before calculating the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library, the method further includes: Perform semantic expansion on the segmented words of the transaction return information and the sample words in the sample word library according to the semantic expansion resource and the rule engine to generate a first semantic expansion set of the segmented words of the transaction return information and a second semantic expansion set of the sample words; Converting the extended word segments in the first semantic extension set and the extended sample word segments in the second semantic extension set into word vectors; The calculating the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library includes: Calculating the similarity between the semantic features of the word vector and the semantic features of the sample words in the sample vocabulary library; Calculating the similarity between the word vector of the extended word in the first semantic extension set and the word vector of the extended sample word in the second semantic extension set; Based on the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library, and the fusion calculation of the similarity between the word vector of the extended word segmentation in the first semantic extension set and the word vector of the extended sample word segmentation in the second semantic extension set, the similarity between the semantic features of the word vector and the semantic features of the sample vocabulary in the sample vocabulary library is generated.
7. The method for processing bank transaction data according to any one of claims 1 to 6, characterized in that: If the matching exception occurs, the method further includes: Storing abnormal transaction messages with the matching abnormality and abnormal contents of the abnormal transaction messages; Summarize the stored abnormal transaction messages according to the preset batch running cycle and generate an abnormal query file; Obtaining, according to the abnormal query file, an abnormal trend and / or an abnormal proportion in a first batch running cycle corresponding to the abnormal query file; The abnormal trend and / or abnormal proportion is output.
8. A banking transaction data processing system, characterized in that: It comprises a processor and a computer-readable storage medium, wherein the computer-readable storage medium stores machine-executable instructions, and when the machine-executable instructions are executed by a computer, the bank transaction data processing method according to any one of claims 1 to 7 is implemented.
9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores machine-executable instructions, and when the machine-executable instructions are executed by a computer, the bank transaction data processing method described in any one of claims 1 to 7 is implemented.
10. A computer program product, characterized in that The invention comprises a computer program, which implements the method according to any one of claims 1 to 7 when being executed by a processor.