Transaction route matching method and device based on natural language processing, and electronic device

Abstract

The application discloses a transaction clearing route matching method and device based on natural language processing and electronic equipment, and relates to the field of financial technology, and the method comprises the following steps: in the case that a new financial transaction result and a transaction abstract are detected, it is detected whether a transaction deal sheet is contained in the financial transaction result, and when the transaction deal sheet is not detected, a clearing route self-matching mechanism is started, a natural language processing model is used to extract transaction counterparty information, transaction financial institution information and beneficiary object information in the transaction abstract, a transaction element set is obtained, each transaction element in the transaction element set is matched with a cash clearing route mapping table, a clearing route with the highest matching saturation degree and meeting preset business rules is selected as a transaction clearing route, and the transaction clearing process is completed by using the transaction clearing route. The application solves the technical problem that, in the related art, the financial transaction without a deal sheet cannot be automatically matched with a clearing route, which affects the automatic processing efficiency of the transaction, and causes the transaction clearing risk.

Description

Technical Field

[0001] The present invention relates to the field of financial technology or other related fields, and in particular to a transaction remittance matching method and device, and electronic equipment based on natural language processing. Background Art

[0002] In the processing of financial market transactions, once a transaction is concluded, subsequent clearing and remittance operations require precise remittance route matching. In related technologies, transaction remittance route matching relies on detailed information provided in the transaction transaction document, such as the account line, account number, and account name. For transactions that are not accompanied by a transaction document or whose transaction document information is incomplete, automatic remittance route matching cannot be achieved without maintaining a mapping between remittance routes and remittance routes. Existing financial market back-end systems require business personnel to manually maintain remittance route parameters and remittance route mappings during the transaction pre-processing phase to facilitate automatic matching by the system.

[0003] For transactions without a transaction order, if the front-end transaction is completed late and close to the settlement time, the back-end system will not be able to automatically match the remittance route when processing, which will affect the automatic processing efficiency of the transaction and the subsequent settlement process. If it is not processed in time, the settlement window may be missed, and the resulting settlement risk is unpredictable.

[0004] To address the above-mentioned problems, no effective solutions have been proposed so far. Summary of the Invention

[0005] The embodiments of the present invention provide a transaction remittance matching method, device, and electronic device based on natural language processing, so as to at least solve the technical problem in the related art that for financial transactions without transaction orders, the remittance cannot be automatically matched, which affects the automatic processing efficiency of the transaction and leads to transaction settlement risks.

[0006] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a transaction remittance matching method based on natural language processing is provided, including: when a new financial transaction result and a transaction summary are detected, detecting whether the financial transaction result contains a transaction completion note, and when the transaction completion note is not detected, starting a remittance self-matching mechanism; when the remittance self-matching mechanism is started, using a natural language processing model to extract counterparty information, transaction financial institution information and beneficiary information in the transaction summary to obtain a transaction element set, wherein the natural language processing model is a pre-trained model for parsing summary text, and the transaction financial institution information includes at least: transaction account bank, transaction intermediary bank; matching each transaction element in the transaction element set with a cash remittance mapping table, selecting the remittance with the highest matching saturation and meeting preset business rules as the transaction remittance, and using the transaction remittance to complete the transaction settlement process.

[0007] Optionally, before detecting a new financial transaction result and a transaction summary, the method further includes: when detecting the new financial transaction result, retrieving the front-end transaction information table, the associated system parameter table and the transaction details corresponding to each financial product in the financial transaction result; based on the first type of transaction element set, using a key element sorter to extract the transaction information in the front-end transaction information table to obtain a first transaction information set, wherein the first type of transaction element set includes at least one of the following: counterparty characters, transaction type characters, transaction amount characters, and transaction date characters; based on the second type of transaction element set, using a key element sorter to extract the transaction information in the associated system parameter table to obtain a second transaction information set, wherein the second type of transaction element set includes at least one of the following: account bank characters, intermediate bank characters, and account name characters; based on the third type of transaction element set, using a key element sorter to extract the transaction information in the transaction details to obtain a third transaction information set, wherein the third type of transaction element set includes at least one of the following: counterparty abbreviation characters, currency characters, and transaction settlement category characters; and generating the transaction summary based on the first transaction information set, the second transaction information set, and the third transaction information set.

[0008] Optionally, the step of using a natural language processing model to extract counterparty information, transaction financial institution information and beneficiary information in the transaction summary to obtain a set of transaction elements includes: using the natural language processing model to identify the transaction subject element entity in the transaction summary, obtain the transaction account and account name involving the counterparty element, and obtain the counterparty information; using the natural language processing model to identify the institution element entity in the transaction summary, obtain the account bank and intermediary bank of the associated financial institution element, and obtain the transaction financial institution information.

[0009] Optionally, the step of using a natural language processing model to extract the counterparty information, transaction financial institution information and beneficiary information in the transaction summary to obtain a transaction element set also includes: detecting whether the counterparty information in the transaction element set contains the account bank's customer code SWI FTCODE; when it is detected that the counterparty information does not contain the account bank's customer code SWI FTCODE, generating beneficiary information based on the counterparty's abbreviation information, and storing the beneficiary information in a cash remittance route mapping table; or, when it is detected that the account bank's customer code SWI FTCODE exists in the counterparty information, using the account bank's customer code as the beneficiary information.

[0010] Optionally, the step of matching each transaction element in the transaction element set in the cash remittance route mapping table, and selecting the remittance route with the highest matching saturation and complying with preset business rules as the transaction remittance route, further includes: when multiple candidate remittance routes are generated by matching each transaction element in the transaction element set with the cash remittance route mapping table, calculating the matching saturation of each candidate remittance route based on the matched element fields in the cash remittance route mapping table; sorting all the matching saturations to obtain a sorting result; and selecting the candidate remittance route with the highest matching saturation and complying with preset business rules in the sorting result as the transaction remittance route.

[0011] Optionally, the steps of using the transaction remittance route to complete the transaction settlement process include: sending the remittance route information of the transaction remittance route to the business end; if the business end has not confirmed the transaction remittance route, receiving the remittance route adjustment instruction transmitted by the business end; re-matching the transaction remittance route according to the transaction-related business type and business restriction information in the remittance route adjustment instruction, obtaining a new transaction remittance route, and sending the new transaction remittance route to the business end; if the remittance route confirmation instruction is received from the business end, using the transaction remittance route to complete the transaction settlement process.

[0012] Optionally, after detecting whether the financial transaction result includes the transaction completion note, it also includes: when it is detected that the financial transaction result includes the transaction completion note, extracting the transaction account information, transaction financial institution information and transaction party information in the transaction completion note; generating corresponding transaction remittances and remittance mappings based on the transaction account information, the transaction financial institution information and the transaction party information, and using the transaction remittance to authorize the transaction; when the transaction authorization is obtained, starting the payment and collection flow based on the transaction confirmation information, generating a clearing message based on the payment and collection flow, and sending the clearing message to the transaction counterparty to complete the transaction clearing process.

[0013] According to another aspect of an embodiment of the present invention, a transaction remittance matching device based on natural language processing is also provided, including: a transaction order detection unit, which is used to detect whether a new financial transaction result and a transaction summary are included in the financial transaction result, and to start the remittance self-matching mechanism when the transaction order is not detected; a transaction element extraction unit, which is used to use a natural language processing model to extract counterparty information, transaction financial institution information and beneficiary information in the transaction summary when the remittance self-matching mechanism is started, to obtain a transaction element set, wherein the natural language processing model is a pre-trained model for parsing summary text, and the transaction financial institution information includes at least: a transaction account bank and a transaction intermediary bank; a remittance matching unit, which is used to match each transaction element in the transaction element set with a cash remittance mapping table, select the remittance with the highest matching saturation and in compliance with preset business rules as the transaction remittance, and use the transaction remittance to complete the transaction settlement process.

[0014] Optionally, the transaction remittance matching device based on natural language processing further includes: an information table retrieval unit, for retrieving, before detecting a new financial transaction result and a transaction summary, a front-end transaction information table, an associated system parameter table, and transaction details corresponding to each financial product in the financial transaction result when the new financial transaction result is detected; a first information extraction unit, for extracting, based on a first type of transaction element set, transaction information from the front-end transaction information table using a key element sorter to obtain a first transaction information set, wherein the first type of transaction element set includes at least one of the following: counterparty characters, transaction type characters, transaction amount characters, and transaction date characters; a second information extraction unit, for extracting, based on a second type of transaction element set, transaction information from the front-end transaction information table using a key element sorter to obtain a first transaction information set, wherein the first type of transaction element set includes at least one of the following: counterparty characters, transaction type characters, transaction amount characters, and transaction date characters; A transaction element set is configured to use a key element sorter to extract the transaction information in the associated system parameter table to obtain a second transaction information set, wherein the second transaction element set includes at least one of the following: account line characters, middle line characters, and account name characters; a third information extraction unit is configured to use a key element sorter to extract the transaction information in the transaction details based on the third transaction element set to obtain a third transaction information set, wherein the third transaction element set includes at least one of the following: counterparty abbreviation characters, currency characters, and transaction settlement category characters; a transaction summary generation unit is configured to generate the transaction summary based on the first transaction information set, the second transaction information set, and the third transaction information set.

[0015] Optionally, the transaction element extraction unit includes: a counterparty identification module, which is used to use the natural language processing model to identify the transaction subject element entity in the transaction summary, obtain the transaction account and account name involving the counterparty element, and obtain the counterparty information; an institution entity identification module, which is used to use the natural language processing model to identify the institution element entity in the transaction summary, obtain the account bank and intermediary bank of the associated financial institution element, and obtain the transaction financial institution information.

[0016] Optionally, the transaction element extraction unit also includes: a first detection module, used to detect whether the counterparty information in the transaction element set contains the customer code SWI FTCODE of the account bank; a beneficiary information generation module, used to generate beneficiary information based on the counterparty's abbreviation information when it is detected that the customer code SWI FTCODE of the account bank does not exist in the counterparty information, and store the beneficiary information in the cash remittance route mapping table; or, when it is detected that the customer code SWI FTCODE of the account bank exists in the counterparty information, use the customer code of the account bank as the beneficiary information.

[0017] Optionally, the remittance route matching unit includes: a matching saturation calculation module, which is used to calculate the matching saturation of each candidate remittance route based on the matched element fields in the cash remittance route mapping table when multiple candidate remittance routes are generated by matching each transaction element in the transaction element set with the cash remittance route mapping table; a sorting module, which is used to sort all the matching saturations to obtain a sorting result; and a remittance route selection module, which is used to select the candidate remittance route with the highest matching saturation in the sorting result and that meets the preset business rules as the transaction remittance route.

[0018] Optionally, the remittance matching unit includes: a remittance sending module, used to send the remittance information of the transaction remittance to the business end; an adjustment indication receiving module, used to receive the remittance adjustment indication transmitted by the business end when the business end has not confirmed the transaction remittance; a remittance rematching module, used to re-match the transaction remittance according to the transaction-related business type and business restriction information in the remittance adjustment indication, obtain a new transaction remittance, and send the new transaction remittance to the business end; a clearing module, used to use the transaction remittance to complete the transaction clearing process when receiving the remittance confirmation instruction from the business end.

[0019] Optionally, the transaction route matching device based on natural language processing also includes: a transaction order information extraction unit, which is used to, after detecting whether the financial transaction result contains a transaction transaction order, generate a route generation unit for extracting the transaction account information, transaction financial institution information and transaction party information in the transaction transaction order when it is detected that the financial transaction result contains the transaction transaction order; generate a corresponding transaction route and route mapping based on the transaction account information, the transaction financial institution information and the transaction party information, and use the transaction route for transaction authorization; a clearing unit, which is used to, when the transaction authorization is obtained, start the payment and collection flow based on the transaction confirmation information, generate a clearing message based on the payment and collection flow, and send the clearing message to the transaction counterparty to complete the transaction clearing process.

[0020] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is further provided, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the above-mentioned transaction route matching methods based on natural language processing.

[0021] According to another aspect of an embodiment of the present invention, an electronic device is also provided, comprising one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement any one of the above-mentioned transaction route matching methods based on natural language processing.

[0022] According to another aspect of an embodiment of the present invention, a computer program product is provided, including a computer program, which, when executed by a processor, implements the steps of any one of the above-mentioned transaction route matching methods based on natural language processing.

[0023] In the present disclosure, when a new financial transaction result and a transaction summary are detected, it is possible to detect whether the financial transaction result contains a transaction completion note, and when no transaction completion note is detected, the remittance self-matching mechanism is started. When the remittance self-matching mechanism is started, a natural language processing model is used to extract the counterparty information, transaction financial institution information and beneficiary information in the transaction summary to obtain a transaction element set, wherein the natural language processing model is a pre-trained model for parsing summary text, and the transaction financial institution information includes at least: transaction account bank, transaction intermediary bank, and each transaction element in the transaction element set is matched with the cash remittance mapping table, and the remittance with the highest matching saturation and in compliance with preset business rules is selected as the transaction remittance, and the transaction remittance is used to complete the transaction settlement process.

[0024] Based on the above-mentioned public content, in the absence of a transaction order or incomplete transaction order information, key information in the transaction summary can be automatically extracted, and transaction remittance routes can be automatically matched to improve transaction processing efficiency. At the same time, it can ensure that the transaction clearing process is completed before the clearing deadline, avoiding missing the clearing window due to untimely manual processing, and reducing the occurrence of transaction clearing risks, thereby solving the technical problem in related technologies that for financial transactions without transaction orders, remittance routes cannot be automatically matched, affecting the automatic processing efficiency of transactions, and leading to transaction clearing risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0026] Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing a transaction path matching method based on natural language processing is shown;

[0027] Figure 2 This is a flowchart of an optional transaction route matching method based on natural language processing according to an embodiment of the present invention;

[0028] Figure 3 This is a flowchart of an optional automatic matching of transaction routes based on natural language processing according to an embodiment of the present invention;

[0029] Figure 4 is a schematic diagram of an optional transaction route matching device based on natural language processing according to an embodiment of the present invention;

[0030] Figure 5 This is a structural block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0032] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention 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 numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes 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.

[0033] To facilitate those skilled in the art to understand the present invention, some of the terms or nouns involved in the embodiments of the present invention are explained below:

[0034] Transaction: A contract that sets out a set of terms and key elements agreed upon with a counterparty.

[0035] Payment and collection flow: In the financial market backend system, it is an information flow of payment or collection generated according to specific rules after a transaction is completed. It usually includes information such as the counterparty, delivery date, transaction currency, delivery amount, etc., which serves as the basis for subsequent settlement.

[0036] Remittance path: In the financial market backend system, it is the account number, opening bank, account number and other information of the clearing parties involved used after the transaction is completed. It is used to define the clearing path required for the clearing to be transferred from one party to another.

[0037] It should be noted that the transaction remittance matching method and device based on natural language processing in the present disclosure can be used in the field of financial technology. In the case of automatic matching of transaction remittances based on natural language processing, it can also be used in any field other than the field of financial technology. In the case of automatic matching of transaction remittances based on natural language processing, the present disclosure does not limit the application field of the transaction remittance matching method and device based on natural language processing.

[0038] It should be noted that the information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) collected by this disclosure 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 comply with the relevant laws, regulations and standards of the relevant regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation portals for users to choose to authorize or refuse. For example, an interface is set up between this system and relevant users or institutions. Before obtaining relevant information, it is necessary to send an acquisition request to the aforementioned user or institution through the interface, and obtain relevant information after receiving the consent information fed back by the aforementioned user or institution.

[0039] It should be noted that in this disclosure, when collecting and analyzing customer information, the corresponding operation entrance is provided for users to choose to agree or reject the automated decision-making results; if the user chooses to reject, the expert decision-making process will be entered.

[0040] The following embodiments of the present invention can be applied to various systems / applications / devices for transaction routing matching based on natural language processing. The present invention automatically extracts and parses transaction information through natural language processing technology, eliminating the need for manual maintenance of routing mappings. This significantly reduces transaction preprocessing time and significantly improves transaction processing efficiency, particularly when processing transactions without transaction orders. Automatic routing ensures that transactions enter the clearing process in a timely manner, effectively avoiding clearing risks caused by routing matching delays, especially for transactions close to the clearing deadline.

[0041] The present invention also provides a mechanism for adjusting the sink mapping saturation according to business feedback, which can gradually optimize the matching logic and improve the matching accuracy.

[0042] It should be noted that financial transactions are concluded at the terminal front desk. After the transaction is concluded, it flows into the back-end system of the financial institution. After the back-end system accesses the transaction, it will pre-process the transaction. During the pre-processing, it will match the remittance route, confirmation routing rules, clearing routing rules, accounting entity and other information. After the matching is completed, the transaction is authorized, the corresponding accounts are calculated, and the corresponding confirmation is generated. After the confirmation is sent to the counterparty and the confirmation sent by the counterparty is matched, the payment and collection flow will be started, and then a clearing message will be generated and sent to the counterparty. The calculated accounts and the cleared accounts will be bidirectionally checked to complete the entire transaction process.

[0043] The present invention will be described in detail below with reference to various embodiments.

[0044] Example 1

[0045] According to an embodiment of the present invention, an embodiment of a transaction routing matching method based on natural language processing is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0046] The transaction route matching method embodiment based on natural language processing 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 The following is a hardware block diagram of a computer terminal (or mobile device) for implementing a transaction matching method based on natural language processing. Figure 1 As shown, the computer terminal 10 (or mobile device) may include one or more ( Figure 1 The computer 102 includes a processor 102 (the processor 102 may include but is not limited to a microprocessor MCU (Microcontroller Unit) or a programmable logic device FPGA (Field Programmable Gate Array)), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, the computer 102 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), a network interface, a power supply and / or a camera. It will 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 than shown, or with Figure 1 Different configurations shown.

[0047] 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 circuitry". The data processing circuitry may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuitry may be a single independent processing module, or may be incorporated in whole or in part into any of the other components of the computer terminal 10 (or mobile device). As described in the embodiments of the present application, the data processing circuitry serves as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0048] Memory 104 can be used to store software programs and modules for application software, such as the program instructions / data storage device corresponding to the transaction route matching method based on natural language processing in the embodiments of the present application. Processor 102 executes the software programs and modules stored in memory 104 to perform various functional applications and data processing, thereby implementing the aforementioned transaction route matching method based on natural language processing. Memory 104 may include high-speed random access memory and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, memory 104 may further include memory remotely located relative to processor 102, and such remote memory may be connected to computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0049] The transmission device 106 is used to receive or send data via a network. A specific example of the network may include a wireless network provided by the communications provider of the computer terminal 10. In one embodiment, the transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to enable communication with the Internet. In another embodiment, the transmission device 106 may be a radio frequency (RF) module for wireless communication with the Internet.

[0050] The display may be, for example, a touch screen liquid crystal display (LCD), which enables a user to interact with a user interface of the computer terminal 10 (or mobile device).

[0051] It should be noted that, compared to existing technologies, in some cases, financial transactions may not be accompanied by detailed transaction notes, or the information on the transaction notes may not be sufficient to automatically match the remittance route. The present invention extracts key information from the transaction summary and combines it with the existing counterparty and currency information in the system to automatically match or generate cash remittance routes, allowing transactions to enter the clearing process in a timely and accurate manner. At the same time, in a clearing environment with huge transaction volumes and time-sensitive conditions, the traditional manual remittance matching method may be inefficient and unable to meet the needs. The present invention uses an automated remittance matching mechanism to not only increase the speed of transaction processing, but also reduce errors caused by human factors, thereby improving the efficiency and accuracy of the entire clearing process.

[0052] Furthermore, this embodiment can also quickly identify key information in the transaction summary when the transaction is completed close to the settlement deadline, automatically generate and match the remittance path, so that the transaction enters the settlement process in time and avoids missing the settlement window.

[0053] It should be noted that this embodiment can be applied to the backend system / equipment of a financial institution.

[0054] Under the above operating environment, this application provides Figure 2 The transaction routing matching method based on natural language processing is shown. Figure 2 is a flow chart of an optional transaction path matching method based on natural language processing according to an embodiment of the present invention. Figure 2 As shown, the method includes the following steps:

[0055] Optionally, before detecting a new financial transaction result and a transaction summary, when a new financial transaction result is detected, the front-end transaction information table, the associated system parameter table and the transaction details corresponding to each financial product in the financial transaction result are retrieved; based on the first type of transaction element set, a key element sorter is used to extract the transaction information in the front-end transaction information table to obtain a first transaction information set, wherein the first type of transaction element set includes at least one of the following: counterparty characters, transaction type characters, transaction amount characters, and transaction date characters; based on the second type of transaction element set, a key element sorter is used to extract the transaction information in the associated system parameter table to obtain a second transaction information set, wherein the second type of transaction element set includes at least one of the following: account line characters, intermediate line characters, and account name characters; based on the third type of transaction element set, a key element sorter is used to extract the transaction information in the transaction details to obtain a third transaction information set, wherein the third type of transaction element set includes at least one of the following: counterparty abbreviation characters, currency characters and transaction settlement category characters; based on the first transaction information set, the second transaction information set and the third transaction information set, a transaction summary is generated.

[0056] The steps for retrieving the front-end transaction information table, associated system parameter table, and transaction details corresponding to each financial product in a financial transaction result involve automatically accessing three key data sources related to the transaction after a new financial transaction result is received or generated. The front-end transaction information table may include, but is not limited to, detailed terms and conditions of the transaction, such as the transaction type, counterparty, transaction amount, currency, and transaction date. This information is crucial for subsequent transaction processing and remittance mapping. The associated system parameter table records parameters related to transaction settlement, such as the SWIFT code (customer code) and full English name of the account bank and intermediary bank. This information forms the basis for generating cash remittance mapping. The transaction details record more detailed transaction information, including the counterparty's abbreviation, transaction currency, and specific clearing category. This information may not be fully included in the transaction note, but is crucial for generating the transaction summary and remittance mapping.

[0057] In this embodiment, a key element sorter can be used to extract information such as the transaction type, counterparty, transaction amount, currency, and transaction date from the front-end transaction information table to form a first transaction information set. This information can be used to directly or indirectly generate the remittance information required for the transaction. Furthermore, the key element sorter can be used to extract information such as the SWIFT code (customer code) and full English name of the account bank and intermediary bank from the associated system parameter table to form a second transaction information set. This information is used to generate complete remittance information, especially when the counterparty has maintained a SWIFT code. Furthermore, in this embodiment, the key element sorter can be used to extract information such as the counterparty's abbreviation, currency, and transaction settlement category from the transaction details to form a third transaction information set.

[0058] A transaction summary can then be generated based on the first, second, and third transaction information sets. Specifically, all key transaction information extracted from the three data sources is integrated to create a comprehensive transaction summary. The transaction summary includes core transaction elements, such as the counterparty, transaction currency, clearing category, and the SWIFT code and name of the financial institution. The generation of the transaction summary provides a clear and structured information foundation for subsequent automatic remittance matching. Even with incomplete transaction information or the absence of a trade order, accurate remittance matching can be performed using natural language processing technology. By automatically extracting and integrating key transaction information from multiple data sources, more comprehensive and accurate data support is provided for remittance matching. The key element sorter automatically identifies and extracts specific information related to the transaction remittance, such as the account bank, intermediary bank, SWIFT code, account name, counterparty abbreviation, and transaction currency, enabling the system to generate a structured transaction summary based on this information. This structured information integration method not only simplifies the transaction processing process and reduces the need for manual intervention, but also improves the accuracy and efficiency of information processing. This is especially true for transactions with incomplete information. By combining key information in the transaction summary with natural language processing technology, it can automatically complete remittance matching, avoiding delays and errors in manual processing, thereby reducing the risk of missing the clearing window and ensuring the smooth progress of transactions.

[0059] It should be noted that the generation of transaction summaries requires consideration of the characteristics and specific details of various transactions, as well as any missing or uncertain information. In this embodiment, a key element sorter can accurately identify and extract different transaction elements. Furthermore, it possesses certain intelligent processing capabilities, such as handling ambiguous information in text and identifying synonyms or abbreviations. This ensures that the generated transaction summary fully and accurately reflects the key transaction information, providing a solid data foundation for subsequent remittance matching and transaction processing.

[0060] Step S201: When a new financial transaction result and a transaction summary are detected, it is detected whether the financial transaction result contains a transaction completion note, and if no transaction completion note is detected, the remittance self-matching mechanism is started.

[0061] In the financial market backend system, whenever a new financial transaction result and transaction summary are identified or received, the processing flow of the present invention will be started first. The system will automatically detect whether the financial transaction result contains a detailed transaction order. The transaction order usually contains all the detailed information of the transaction, such as the two parties to the transaction, the transaction amount, the currency, the settlement path, etc. If the transaction result contains a transaction order, the system will automatically match the remittance route based on the information in the transaction order; however, if the transaction result does not contain a transaction order or the transaction order information is incomplete, the system will automatically start the remittance route self-matching mechanism based on natural language processing, automatically parse the key elements of the transaction through the transaction summary information, and then generate or match the remittance route, which is crucial to improving the processing efficiency of transactions without transaction orders and reducing manual intervention.

[0062] In step S202, when the Huilu self-matching mechanism is activated, a natural language processing model is used to extract counterparty information, transaction financial institution information, and beneficiary information from the transaction summary to obtain a set of transaction elements. The natural language processing model is a pre-trained model for parsing summary text, and the transaction financial institution information includes at least: the transaction account bank and the transaction intermediary bank.

[0063] After the remittance self-matching mechanism is activated, a pre-trained natural language processing model can be used to deeply analyze the transaction summary. This model extracts a set of transaction elements by identifying key entities in the summary text, such as the counterparty, the financial institutions involved in the transaction (including the correspondent bank and intermediary bank), and the beneficiary. It should be noted that counterparty information may include the name of the counterparty; financial institution information includes at least the names of the account bank and intermediary bank, which play a key role in transaction clearing; and beneficiary information refers to the entity to which the transaction funds ultimately flow. Through this step, the system is able to extract specific key information from the abstract transaction summary, which can be used for remittance matching.

[0064] Optionally, step S202 includes: using a natural language processing model to identify the transaction subject element entity in the transaction summary, obtaining the transaction account and account name involved in the counterparty element, and obtaining the counterparty information; using a natural language processing model to identify the institution element entity in the transaction summary, obtaining the account bank and intermediary bank of the associated financial institution element, and obtaining the transaction financial institution information.

[0065] It should be noted that the natural language processing model selected in this embodiment can be a BERT model, which is used to extract counterparty information and transaction financial institution information from the transaction summary. Here, the transaction subject element entity includes key information such as the name and account of the counterparty. By analyzing the text, the model can identify and extract the counterparty's transaction account and account name, thereby obtaining detailed information about the counterparty. This process ensures that the counterparty can be identified as accurately as possible even when the information is incomplete by parsing the semantic relationships in the summary. Similarly, the natural language processing model is used to identify the institutional element entities in the transaction summary. These entities involve financial institutions used in the transaction, such as the correspondent bank and the intermediary bank. The model will identify and extract the name information of these financial institutions, and then obtain the account bank and intermediary bank information involved in the transaction.

[0066] After extracting the transaction element set, the counterparty information is further processed to generate beneficiary information. Optionally, step S202 further includes: detecting whether the counterparty information in the transaction element set contains the account bank's customer code SWIFTCODE; if it is detected that the counterparty information does not contain the account bank's customer code SWIFTCODE, generating beneficiary information based on the counterparty's short name information and storing the beneficiary information in the cash remittance route mapping table; or, if it is detected that the counterparty information contains the account bank's customer code SWIFTCODE, using the account bank's customer code as the beneficiary information.

[0067] When detecting the SWIFT code in counterparty information, the system automatically checks whether it contains the SWIFT code, the account bank's customer code. The SWIFT code is widely used internationally for interbank communications and clearing, and its presence directly impacts the efficiency and accuracy of remittance matching. If the counterparty information contains the SWIFT code, it is directly used as the beneficiary information and stored in the cash remittance mapping table for subsequent remittance matching. The presence of the SWIFT code ensures the accuracy and efficiency of remittance information. Conversely, if the SWIFT code is not detected, the system generates beneficiary information based on the counterparty's short name and stores it in the cash remittance mapping table. This process may require intelligent matching based on historical data, transaction habits, and other factors to generate the most accurate beneficiary information possible, ensuring that transactions proceed smoothly even with incomplete information.

[0068] Natural language processing (NLP) technology extracts transaction elements, particularly information about counterparties and financial institutions, enabling accurate identification of the entities and relationships involved in transactions. The aforementioned natural language processing model is pre-trained. During the model training phase, a large amount of annotated financial transaction data, including transaction summaries of different types, is used to ensure the model's generalization and accuracy.

[0069] Step S203 : Match each transaction element in the transaction element set with the cash remittance route mapping table, select the remittance route with the highest matching saturation and in compliance with the preset business rules as the transaction remittance route, and use the transaction remittance route to complete the transaction settlement process.

[0070] After extracting the transaction element set, the system uses an intelligent matching algorithm to compare and match these transaction elements with the system's internal cash remittance route mapping table. The cash remittance route mapping table is a database that stores various remittance route information and matching rules. It contains information about account banks, intermediary banks, counterparties, and clearing routes. Based on the degree of match between each element in the transaction element set and the mapping table, the system calculates the matching saturation—the degree to which the transaction element matches the preset remittance route information. Generally, a higher matching saturation indicates a more accurate match between the transaction element and the remittance route. The system selects the remittance route with the highest matching saturation that also meets preset business rules (such as transaction type and currency restrictions) as the transaction remittance route. Once the transaction remittance route is determined, the system automatically completes the transaction clearing process according to the selected remittance route, including generating the clearing message, sending it to the counterparty, and performing account reconciliation until the transaction is successfully cleared. This ensures that even with incomplete transaction information, the system can efficiently and accurately complete remittance route matching and transaction clearing, thereby improving the efficiency and security of financial market transactions.

[0071] Optionally, step S203 further includes: when each transaction element in the transaction element set is matched with the cash remittance route mapping table to generate multiple candidate remittance routes, calculating the matching saturation of each candidate remittance route based on the matched element fields in the cash remittance route mapping table; sorting all matching saturations to obtain a sorting result; and selecting the candidate remittance route with the highest matching saturation in the sorting result and meeting the preset business rules as the transaction remittance route.

[0072] When matching each transaction element in a transaction element set against the cash remittance route mapping table, the remittance route with the highest matching saturation and compliance with preset business rules must be selected as the transaction remittance route. The matching process and remittance route selection specifically include generating multiple candidate remittance routes and calculating matching saturation. Specifically, when matching the information in the transaction element set with records in the cash remittance route mapping table, multiple candidate remittance routes may be generated. For each candidate remittance route, the system calculates its matching saturation based on the number and quality of matched element fields (such as counterparty, currency, and account line SWIFT code). A higher matching saturation indicates a higher degree of matching between the remittance route and the transaction elements, and thus a greater ability to meet transaction clearing requirements. Then, matching saturation ranking and remittance route selection are implemented. The system ranks all candidate remittance routes by matching saturation, generating a ranking result. Ranking is typically performed in descending order of matching saturation. From the ranking result, the system selects the candidate remittance route with the highest matching saturation and compliance with preset business rules as the transaction remittance route. Preset business rules may include but are not limited to transaction type, currency, business scope of the account bank, etc., to ensure that the selected remittance route not only technically matches the transaction elements but is also reasonable in terms of business logic.

[0073] Through the above implementation steps, the system can automatically identify the most appropriate remittance channel, reducing the need for manual intervention and improving transaction processing efficiency and accuracy. The matching saturation calculation and sorting logic ensure that even in cases of incomplete information or multiple matching options, the system can accurately make decisions and select the most appropriate remittance channel for transaction settlement.

[0074] Through the above steps, when a new financial transaction result and transaction summary are detected, it is possible to detect whether the financial transaction result contains a transaction completion note, and when no transaction completion note is detected, the remittance self-matching mechanism is started. When the remittance self-matching mechanism is started, a natural language processing model is used to extract the counterparty information, transaction financial institution information and beneficiary information in the transaction summary to obtain a transaction element set, wherein the natural language processing model is a pre-trained model for parsing summary text, and the transaction financial institution information includes at least: transaction account bank, transaction intermediary bank, and each transaction element in the transaction element set is matched with the cash remittance mapping table, and the remittance route with the highest matching saturation and that meets the preset business rules is selected as the transaction remittance route, and the transaction remittance route is used to complete the transaction settlement process. In this embodiment, when there is no transaction order or the transaction order information is incomplete, key information in the transaction summary can be automatically extracted, and the transaction remittance route can be automatically matched to improve transaction processing efficiency. At the same time, it can ensure that the transaction clearing process is completed before the clearing deadline, avoid missing the clearing window due to untimely manual processing, and reduce the occurrence of transaction clearing risks, thereby solving the technical problem in related technologies that for financial transactions without transaction orders, the remittance route cannot be automatically matched, which affects the automatic processing efficiency of the transaction and leads to transaction clearing risks.

[0075] Optionally, when using the transaction remittance route to complete the transaction settlement process, this embodiment includes: sending the remittance route information of the transaction remittance route to the business end; if the business end has not confirmed the transaction remittance route, receiving the remittance route adjustment instruction transmitted by the business end; re-matching the transaction remittance route according to the transaction-related business type and business restriction information in the remittance route adjustment instruction, obtaining a new transaction remittance route, and sending the new transaction remittance route to the business end; if the remittance route confirmation instruction is received from the business end, using the transaction remittance route to complete the transaction settlement process.

[0076] The steps involved in completing the transaction clearing process using a transaction remittance route include interaction with the business end. First, the system sends the automatically matched transaction remittance route information to the business end, where personnel conduct final confirmation. This confirmation process is a crucial step in the transaction clearing process, ensuring the accuracy of the transaction remittance route and transaction compliance. If the business end fails to confirm the automatically matched transaction remittance route, the system will receive remittance route adjustment instructions from the business end, which may include additional conditions such as the transaction-related business type and business restrictions. Based on these adjustment instructions, the system will re-match the transaction remittance route to obtain a new transaction remittance route that better meets business requirements. This process may involve re-querying and matching the cash remittance route mapping table, as well as recalculating and ranking the matching saturation. Once the business end confirms the transaction remittance route, the system automatically initiates the transaction clearing process based on the confirmed transaction remittance route information. This includes generating a clearing message, executing payment and collection operations, and sending the clearing message to the counterparty, until the transaction is successfully cleared.

[0077] It should be noted that in actual operations, the confirmation and adjustment of transaction routing is a dynamic process that may require multiple iterations before finalization. The system can quickly respond to feedback and instructions from the business side and adjust matching strategies to ensure that the selection of transaction routing meets both technical matching requirements and the business side's operational practices and compliance requirements. Furthermore, the system can record and analyze the reasons for each routing adjustment, continuously optimize matching algorithms, and improve the accuracy and efficiency of subsequent automatic matching of transaction routings. This improves the level of transaction processing automation, enhances the system's flexibility and business adaptability, and improves the speed and accuracy of financial market transaction processing.

[0078] Optionally, after detecting whether the financial transaction result contains a transaction completion note, it also includes: when it is detected that the financial transaction result contains a transaction completion note, extracting the transaction account information, transaction financial institution information and transaction party information in the transaction completion note; generating corresponding transaction remittances and remittance mappings based on the transaction account information, transaction financial institution information and transaction party information, and using the transaction remittance to authorize the transaction; when the transaction authorization is obtained, starting the payment and collection flow based on the transaction confirmation information, generating a clearing message based on the payment and collection flow, and sending the clearing message to the counterparty to complete the transaction clearing process.

[0079] When a new financial transaction result containing a transaction order is detected, key transaction account information, transaction financial institution information, and transaction party information can be directly extracted from the transaction order. This information forms the basis for generating the transaction route. Next, based on this information, the corresponding transaction route and route mapping are generated. This is a critical step in the transaction settlement process, defining the specific path for the transfer of transaction funds from one party to another. Once the transaction route is generated, the system will perform transaction authorization, a necessary step before settlement, to confirm the legality and compliance of the transaction. After successful transaction authorization, the payment and collection process is initiated based on the transaction confirmation information. This initiates the specific operations of transaction settlement, including generating and sending the settlement message to the counterparty, ultimately completing the settlement process.

[0080] Specifically, when a financial transaction is detected with a transaction note, the backend system can directly extract complete transaction account information, financial institution information, and information about both parties from the transaction note. This not only simplifies the remittance route generation process but also improves the accuracy and completeness of transaction information. A transaction note typically contains all transaction details, including the transaction amount, currency, transaction date, and counterparty details, and serves as a crucial document for transaction settlement. The system automatically generates a transaction remittance route based on the transaction account, financial institution, and party information. This process may involve querying historical remittance routes, matching clearing rules, and verifying the information of both parties. The generated transaction remittance route is stored in the cash remittance route mapping table, serving as the basis for subsequent automatic remittance route matching. Furthermore, the generated remittance route mapping includes all necessary transaction information to ensure a smooth clearing process.

[0081] Transaction authorization is a critical step before transaction settlement. It ensures the legality and compliance of the transaction and is a prerequisite for transaction settlement. After successful authorization, the system will initiate the payment and collection process based on the transaction confirmation information. The confirmation information is the confirmation of transaction details by both parties, including the transaction amount, currency, and transaction date, and serves as the basis for initiating transaction settlement. The system will generate specific payment and collection instructions based on the confirmation information, and the settlement process will begin.

[0082] Furthermore, after the payment and collection process is initiated, the system generates a clearing message, an electronic document containing detailed information about the transaction and a crucial communication medium in the clearing process. This message is sent to the counterparty to confirm the transaction's clearing details. The generation and transmission of the clearing message also marks the beginning of the clearing process. The entire clearing process, including fund transfers and account reconciliation, proceeds according to the contents of the clearing message until the transaction is cleared.

[0083] In practice, the clearing process may require customization based on factors such as transaction type, counterparty, and clearing rules. Therefore, the system needs to possess learning and optimization capabilities, continuously adjusting transaction routing and clearing message generation strategies based on historical transaction data and clearing results to improve clearing efficiency and accuracy. Furthermore, the system should provide monitoring and analysis tools for the clearing process, helping business personnel understand clearing status and promptly address anomalies to ensure a smooth clearing process.

[0084] The embodiments of the present invention provide an efficient and accurate solution for automatically matching transaction routings and initiating clearing processes based on natural language processing. For both transactions with and without transaction orders, the system intelligently analyzes transaction summary information, generates or matches accurate transaction routings, automatically authorizes transactions, generates and sends clearing messages to counterparties, and completes the entire transaction clearing process. This not only improves the automation level of transaction processing but also reduces clearing risks caused by incomplete information or manual errors, thus providing significant value for optimizing financial market transaction processing systems.

[0085] The following describes in detail another optional specific implementation.

[0086] To address the issue of being unable to automatically match remittance routes, this embodiment provides a solution that automatically extracts summary information from transactions based on a natural language processing model and automatically matches the most accurate remittance route based on the summary information. Specifically, the information extraction components involved in this embodiment are first described.

[0087] 1. Key element sorter.

[0088] By organizing the front-end transaction information tables corresponding to different products, we obtain summary information corresponding to the corresponding fields. In the summary information, we use the Correspondent Bank keyword to extract account bank information and the Intermediary Bank keyword to extract intermediary bank information. We then associate the account bank and intermediary bank information in the clearing counterparty in the system parameter table, obtain the corresponding SWIFT CODE (account bank customer code) or full English name (account bank account name), and store it in the cash remittance route table. Furthermore, we obtain the counterparty from the transaction details, associate the clearing counterparty in the system table, obtain the corresponding SWIFT CODE (account bank customer code) or full English name (account bank account name), and store it in the cash remittance route table. We then generate a remittance route mapping based on the counterparty's English abbreviation, currency, transaction clearing subcategory, and other information.

[0089] 2. Key element parser.

[0090] Natural language processing (NLP) algorithms are often required to identify key information and sentiment in transactions.

[0091] For the implementation mode of the present invention, a named entity recognition method can be used (of course, the naive Bayes algorithm, support vector machine SVM algorithm, etc. can also be used) to obtain key information such as the account line, intermediary bank, account number, etc. in the summary (account line information is required, and intermediary bank and account number information are optional), combined with the existing counterparty, currency and other related information in the transaction (already existing in the transaction information table), generate cash remittance route information and remittance route matching information mapping, and call the transaction preprocessing process to complete the remittance route matching.

[0092] 3. Huilu matching engine.

[0093] Remittance route information is matched based on the summary extraction results of the key element parser, such as account line, account number, counterparty, currency, and other information. On the one hand, the account line, intermediary bank, and beneficiary generated based on the counterparty are extracted from the results and a cash remittance route is generated according to the existing system logic. If the counterparty maintains a SWIFT CODE, the beneficiary information will be the SWIFT type. If not, the counterparty's English abbreviation is used as the beneficiary account name and stored in the cash remittance route. On the other hand, the counterparty, currency, and clearing subcategory in the transaction are extracted from the results as matching conditions to generate a matching mapping and insert it into the existing cash remittance route mapping table.

[0094] When matching transactions to a remittance route, the decision is made based on the matching degree of the matching fields in the mapping table. If multiple matches are possible, they are sorted by matching saturation, with the highest saturation being prioritized and provided to the business. After the match is complete, manual authorization is performed, and the business confirms whether the remittance route is acceptable. If it is not what the business wants, the transaction can be restarted by adjusting the remittance route mapping saturation (such as adding restrictions on business categories and business types) to match the desired remittance route. This gradually optimizes the matching logic and accuracy. If no matching remittance route is found based on the matching information, a pending task is generated for the business to handle.

[0095] Figure 3 This is a flowchart of an optional automatic matching of transaction routes based on natural language processing according to an embodiment of the present invention. Figure 3 Shown, including:

[0096] 31. When a new transaction inflow is detected, check whether the financial transaction is accompanied by a transaction note. If there is a transaction note, the transaction note information already contains detailed information such as the account line and account number, which can be automatically matched or a matching remittance path can be generated. If it does not contain a transaction note and the business has not maintained the mapping in advance, the self-matching mechanism is required to complete the automatic remittance matching.

[0097] 32. Transaction pre-processing requires matching remittance routes and clearing routing rules. If the remittance routes do not match, manual transaction authorization will be required, and subsequent matching of clearing routing rules will also fail. Clearing routing rules match the generated message type based on transaction and remittance route information. Here, a key element sorter is required to classify transaction elements, a key element parser is used to parse transaction elements, and a remittance matching engine is used to achieve transaction remittance matching.

[0098] 33. Verify the existence of the account line information obtained from the summary information. If it does not exist, the remittance route cannot be generated. If it does exist, the counterparty account line and beneficiary information are generated based on the account line information and counterparty information, and the corresponding matching mapping is generated. The transaction is manually authorized to the business. The remittance route information can be displayed to the business on the remittance route sheet page, and the accuracy of the extraction results can be gradually learned and optimized.

[0099] 34. If no useful element information can be extracted based on the summary information, a to-do task will be generated in the system to remind you that the business transaction has not been matched to the remittance route and requires manual maintenance. The transaction will also be manually authorized.

[0100] Based on the above technical solution, the present invention provides an automatic matching device for transaction remittances based on natural language processing, which can solve the problem of untimely manual processing when transactions cannot be automatically matched to remittances. It can automatically generate matching mappings between remittances and remittances based on summary information, complete automatic matching of transaction remittances, and automatically complete subsequent clearing processes, effectively reducing the clearing risk caused by missing the clearing deadline.

[0101] The following describes it in detail with reference to another embodiment.

[0102] Example 2

[0103] A transaction route matching device based on natural language processing provided in this embodiment includes multiple implementation units, each of which corresponds to each implementation step in the above-mentioned embodiment 1. Its specific implementation methods and beneficial effects can be referred to the above-mentioned method embodiments and will not be repeated here.

[0104] Figure 4 is a schematic diagram of an optional transaction path matching device based on natural language processing according to an embodiment of the present invention. Figure 4 As shown, the transaction route matching device based on natural language processing may include: a transaction order detection unit 41, a transaction element extraction unit 42, and a route matching unit 43.

[0105] The transaction order detection unit 41 is used to detect whether the financial transaction result contains a transaction order when a new financial transaction result and transaction summary are detected, and to start the remittance self-matching mechanism if no transaction order is detected.

[0106] The transaction element extraction unit 42 is used to extract the counterparty information, transaction financial institution information, and beneficiary information from the transaction summary using a natural language processing model when the remittance self-matching mechanism is activated to obtain a set of transaction elements, wherein the natural language processing model is a pre-trained model for parsing summary text, and the transaction financial institution information includes at least: the transaction account bank and the transaction intermediary bank.

[0107] The remittance route matching unit 43 is used to match each transaction element in the transaction element set with the cash remittance route mapping table, select the remittance route with the highest matching saturation and in compliance with the preset business rules as the transaction remittance route, and use the transaction remittance route to complete the transaction settlement process.

[0108] The above-mentioned transaction remittance route matching device based on natural language processing can detect whether the financial transaction result contains a transaction transaction document when a new financial transaction result and transaction summary are detected through the transaction document detection unit 41, and start the remittance route self-matching mechanism when the transaction transaction document is not detected. When the remittance route self-matching mechanism is started, the transaction element extraction unit 42 uses a natural language processing model to extract the counterparty information, transaction financial institution information and beneficiary information in the transaction summary to obtain a transaction element set, wherein the natural language processing model is a pre-trained model for parsing summary text, and the transaction financial institution information includes at least: transaction account bank, transaction intermediary bank, and each transaction element in the transaction element set is matched with the cash remittance route mapping table through the remittance route matching unit 43, and the remittance route with the highest matching saturation and in compliance with preset business rules is selected as the transaction remittance route, and the transaction remittance route is used to complete the transaction settlement process. In this embodiment, when there is no transaction order or the transaction order information is incomplete, key information in the transaction summary can be automatically extracted, and the transaction remittance route can be automatically matched to improve transaction processing efficiency. At the same time, it can ensure that the transaction clearing process is completed before the clearing deadline, avoid missing the clearing window due to untimely manual processing, and reduce the occurrence of transaction clearing risks, thereby solving the technical problem in related technologies that for financial transactions without transaction orders, the remittance route cannot be automatically matched, which affects the automatic processing efficiency of the transaction and leads to transaction clearing risks.

[0109] Optionally, the transaction remittance matching device based on natural language processing further includes: an information table retrieval unit, for retrieving the front-end transaction information table, the associated system parameter table and the transaction details corresponding to each financial product in the financial transaction result before detecting the new financial transaction result and the transaction summary; a first information extraction unit, for extracting the transaction information in the front-end transaction information table using a key element sorter based on the first type of transaction element set to obtain the first transaction information set, wherein the first type of transaction element set includes at least one of the following: counterparty characters, transaction type characters, transaction amount characters, and transaction date characters; a second information extraction unit, for extracting the transaction information in the front-end transaction information table based on the first type of transaction element set using a key element sorter to obtain the first transaction information set, wherein the first type of transaction element set includes at least one of the following: counterparty characters, transaction type characters, transaction amount characters, and transaction date characters; The second type of transaction element set uses a key element sorter to extract transaction information from the associated system parameter table to obtain a second transaction information set, wherein the second type of transaction element set includes at least one of the following: account line characters, middle line characters, and account name characters; the third information extraction unit is used to extract transaction information from transaction details based on the third type of transaction element set using a key element sorter to obtain a third transaction information set, wherein the third type of transaction element set includes at least one of the following: counterparty abbreviation characters, currency characters, and transaction settlement category characters; the transaction summary generation unit is used to generate a transaction summary based on the first transaction information set, the second transaction information set, and the third transaction information set.

[0110] Optionally, the transaction element extraction unit includes: a counterparty identification module, which is used to use a natural language processing model to identify the transaction subject element entity in the transaction summary, obtain the transaction account and account name involving the counterparty element, and obtain the counterparty information; an institution entity identification module, which is used to use a natural language processing model to identify the institution element entity in the transaction summary, obtain the account bank and intermediary bank of the associated financial institution element, and obtain the transaction financial institution information.

[0111] Optionally, the transaction element extraction unit also includes: a first detection module, used to detect whether the counterparty information in the transaction element set contains the customer code SWI FTCODE of the account bank; a beneficiary information generation module, used to generate beneficiary information based on the counterparty's abbreviation information when it is detected that the counterparty information does not contain the customer code SWI FTCODE of the account bank, and store the beneficiary information into the cash remittance route mapping table; or, when it is detected that the counterparty information contains the customer code SWI FTCODE of the account bank, use the customer code of the account bank as the beneficiary information.

[0112] Optionally, the remittance route matching unit includes: a matching saturation calculation module, which is used to calculate the matching saturation of each candidate remittance route based on the matched element fields in the cash remittance route mapping table when each transaction element in the transaction element set is matched with the cash remittance route mapping table to generate multiple candidate remittance routes; a sorting module, which is used to sort all matching saturations to obtain a sorting result; and a remittance route selection module, which is used to select the candidate remittance route with the highest matching saturation in the sorting result and that meets the preset business rules as the transaction remittance route.

[0113] Optionally, the remittance matching unit includes: a remittance sending module, which is used to send the remittance information of the transaction remittance to the business end; an adjustment indication receiving module, which is used to receive the remittance adjustment indication transmitted by the business end when the business end has not confirmed the transaction remittance; a remittance rematching module, which is used to re-match the transaction remittance according to the transaction-related business type and business restriction information in the remittance adjustment indication, obtain a new transaction remittance, and send the new transaction remittance to the business end; a clearing module, which is used to complete the transaction clearing process using the transaction remittance when receiving the remittance confirmation instruction from the business end.

[0114] Optionally, the transaction route matching device based on natural language processing also includes: a transaction order information extraction unit, which is used to, after detecting whether the financial transaction result contains a transaction transaction order, generate a route generation unit for extracting the transaction account information, transaction financial institution information and transaction party information in the transaction transaction order when it is detected that the financial transaction result contains the transaction transaction order; generate corresponding transaction routes and route mappings based on the transaction account information, transaction financial institution information and transaction party information, and use the transaction route for transaction authorization; a clearing unit, which is used to, when the transaction authorization is obtained, start the payment and collection flow based on the transaction confirmation information, generate a clearing message based on the payment and collection flow, and send the clearing message to the counterparty to complete the transaction clearing process.

[0115] The above-mentioned transaction route matching device based on natural language processing can also include a processor and a memory. The above-mentioned transaction order detection unit 41, transaction element extraction unit 42, route matching unit 43, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize the corresponding functions.

[0116] The processor includes a kernel, which retrieves the corresponding program unit from the memory. One or more kernels can be configured, and automatic matching of transaction paths based on natural language processing can be achieved by adjusting kernel parameters.

[0117] The above-mentioned memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0118] Example 3

[0119] An embodiment of the present application may provide an electronic device, Figure 5 This 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 502, memory 504, storage controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0120] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the transaction route matching method and device based on natural language processing in the embodiments of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby realizing the above-mentioned transaction route matching method based on natural language processing. The memory may include high-speed random access memory and may also include 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 may further include a memory remotely located 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, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0121] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: when a new financial transaction result and transaction summary are detected, detect whether the financial transaction result contains a transaction completion order, and when no transaction completion order is detected, start the remittance self-matching mechanism; when the remittance self-matching mechanism is started, use the natural language processing model to extract the counterparty information, transaction financial institution information and beneficiary information in the transaction summary to obtain a transaction element set, wherein the natural language processing model is a pre-trained model for parsing summary text, and the transaction financial institution information includes at least: transaction account bank, transaction intermediary bank; match each transaction element in the transaction element set with the cash remittance mapping table, select the remittance with the highest matching saturation and in accordance with the preset business rules as the transaction remittance, and use the transaction remittance to complete the transaction settlement process.

[0122] The processor can also call information and applications stored in the memory through the transmission device to perform the following steps: before detecting a new financial transaction result and transaction summary, when a new financial transaction result is detected, call the front-end transaction information table, the associated system parameter table and transaction details corresponding to each financial product in the financial transaction result; based on the first type of transaction element set, use a key element sorter to extract transaction information in the front-end transaction information table to obtain a first transaction information set, wherein the first type of transaction element set includes at least one of the following: counterparty characters, transaction type characters, transaction amount characters, and transaction date characters; based on the second type of transaction element set, use a key element sorter to extract transaction information in the associated system parameter table to obtain a second transaction information set, wherein the second type of transaction element set includes at least one of the following: account line characters, intermediate line characters, and account name characters; based on the third type of transaction element set, use a key element sorter to extract transaction information in the transaction details to obtain a third transaction information set, wherein the third type of transaction element set includes at least one of the following: counterparty abbreviation characters, currency characters, and transaction settlement category characters; based on the first transaction information set, the second transaction information set, and the third transaction information set, generate a transaction summary.

[0123] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: using a natural language processing model to identify the transaction subject element entity in the transaction summary, obtain the transaction account and account name involving the counterparty element, and obtain the counterparty information; using a natural language processing model to identify the institution element entity in the transaction summary, obtain the account bank and intermediary bank of the associated financial institution element, and obtain the transaction financial institution information.

[0124] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: detecting whether the counterparty information in the transaction element set contains the customer code SWI FTCODE of the account bank; when it is detected that the counterparty information does not contain the customer code SWI FTCODE of the account bank, generating beneficiary information based on the abbreviation information of the counterparty, and storing the beneficiary information in the cash remittance route mapping table; or, when it is detected that the counterparty information contains the customer code SWI FTCODE of the account bank, using the customer code of the account bank as the beneficiary information.

[0125] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: when each transaction element in the transaction element set is matched with the cash remittance route mapping table to generate multiple candidate remittance routes, calculate the matching saturation of each candidate remittance route based on the matched element fields in the cash remittance route mapping table; sort all matching saturations to obtain a sorting result; and select the candidate remittance route with the highest matching saturation in the sorting result and that meets the preset business rules as the transaction remittance route.

[0126] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: sending the transaction remittance information of the transaction remittance to the business end; if the business end has not confirmed the transaction remittance, receiving the remittance adjustment instruction transmitted by the business end; re-matching the transaction remittance according to the transaction-related business type and business restriction information in the remittance adjustment instruction, obtaining a new transaction remittance, and sending the new transaction remittance to the business end; when receiving the remittance confirmation instruction from the business end, using the transaction remittance to complete the transaction settlement process.

[0127] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: after detecting whether the financial transaction result contains a transaction completion order, if it is detected that the financial transaction result contains a transaction completion order, extract the transaction account information, transaction financial institution information and transaction party information in the transaction completion order; based on the transaction account information, transaction financial institution information and transaction party information, generate the corresponding transaction remittance and remittance mapping, and use the transaction remittance to authorize the transaction; if the transaction authorization is obtained, start the payment and collection flow based on the transaction confirmation information, generate a clearing message based on the payment and collection flow, and send the clearing message to the counterparty to complete the transaction clearing process.

[0128] 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 terminal device such as a smart phone, a tablet computer, a PDA, a Mobile Internet Device (MID), or a PAD. Figure 5 It does not limit the structure of the above electronic device. For example, the electronic device may also include Figure 5 More or fewer components (such as network interfaces, display devices, etc.) shown in, or with Figure 5 Different configurations shown.

[0129] Those skilled in the art will appreciate that all or part of the steps in the various natural language processing-based transaction route matching methods of the above-mentioned embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0130] Example 4

[0131] The embodiment of the present application further provides a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the transaction route matching method based on natural language processing provided in the first embodiment.

[0132] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is further provided, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the transaction route matching method based on natural language processing according to any one of the above-mentioned embodiments.

[0133] Optionally, in this embodiment, the 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.

[0134] The present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the transaction route matching method based on natural language processing described in each embodiment of the present application.

[0135] The present application also provides a computer program product, including a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the transaction route matching method based on natural language processing described in each embodiment of the present application are implemented.

[0136] The serial numbers of the above embodiments of the present invention are for description only and do not represent the advantages or disadvantages of the embodiments.

[0137] In the above embodiments of the present invention, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0138] 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 exemplary. For example, the division of the units can be a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0139] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0140] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0141] 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 invention, 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. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, etc. Various media that can store program codes.

[0142] The above is only a preferred embodiment of the present invention. 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 invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A transaction path matching method based on natural language processing, characterized in that: include: Upon detecting a new financial transaction result and transaction summary, detecting whether the financial transaction result contains a transaction order, and activating the Huilu self-matching mechanism if the transaction order is not detected; When the Huilu self-matching mechanism is activated, a natural language processing model is used to extract counterparty information, transaction financial institution information, and beneficiary information from the transaction summary to obtain a transaction element set, wherein the natural language processing model is a pre-trained model for parsing summary text, and the transaction financial institution information includes at least: the transaction account bank and the transaction intermediary bank; Matching each transaction element in the transaction element set with a cash remittance route mapping table, selecting a remittance route with the highest matching saturation and meeting preset business rules as the transaction remittance route, and using the transaction remittance route to complete the transaction settlement process, specifically including sending remittance route information of the transaction remittance route to the business end; if the business end does not confirm the transaction remittance route, receiving a remittance route adjustment instruction transmitted by the business end; re-matching the transaction remittance route based on the transaction-related business type and business restriction information in the remittance route adjustment instruction to obtain a new transaction remittance route, and sending the new transaction remittance route to the business end; and if the business end receives a remittance route confirmation instruction, using the transaction remittance route to complete the transaction settlement process; Before detecting a new financial transaction result and a transaction summary, the method further includes: upon detecting the new financial transaction result, retrieving the front-end transaction information table, the associated system parameter table, and the transaction details corresponding to each financial product in the financial transaction result; extracting the transaction information from the front-end transaction information table using a key element sorter based on a first-category transaction element set to obtain a first transaction information set, wherein the first-category transaction element set includes at least one of the following: counterparty characters, transaction type characters, transaction amount characters, and transaction date characters; extracting the transaction information from the associated system parameter table using a key element sorter based on a second-category transaction element set to obtain a second transaction information set, wherein the second-category transaction element set includes at least one of the following: account bank characters, intermediate bank characters, and account name characters; extracting the transaction information from the transaction details using a key element sorter based on a third-category transaction element set to obtain a third transaction information set, wherein the third-category transaction element set includes at least one of the following: counterparty abbreviation characters, currency characters, and transaction settlement category characters; and generating the transaction summary based on the first, second, and third transaction information sets.

2. The transaction route matching method according to claim 1, characterized in that: The step of extracting counterparty information, transaction financial institution information, and beneficiary information from the transaction summary using a natural language processing model to obtain a transaction element set includes: Using the natural language processing model to identify the transaction subject element entity in the transaction summary, obtain the transaction account and account name involved in the counterparty element, and obtain the counterparty information; The natural language processing model is used to identify the institutional element entity in the transaction summary, obtain the account bank and the intermediary bank of the associated financial institution element, and obtain the transaction financial institution information.

3. The transaction route matching method according to claim 2, characterized in that: The step of extracting counterparty information, transaction financial institution information, and beneficiary information from the transaction summary using a natural language processing model to obtain a transaction element set further includes: Detecting whether the counterparty information in the transaction element set includes the customer code SWIFTCODE of the account bank; If it is detected that the counterparty information does not contain the customer code SWIFTCODE of the account bank, generating beneficiary information based on the counterparty's short name information, and storing the beneficiary information in the cash remittance route mapping table; or When it is detected that the counterparty information contains the customer code SWIFTCODE of the account bank, the customer code of the account bank is used as the beneficiary information.

4. The transaction route matching method according to claim 1, characterized in that: The step of matching each transaction element in the transaction element set in a cash remittance route mapping table and selecting a remittance route with the highest matching saturation and conforming to preset business rules as the transaction remittance route also includes: When multiple candidate remittance routes are generated by matching each transaction element in the transaction element set with the cash remittance route mapping table, a matching saturation of each candidate remittance route is calculated based on the matched element fields in the cash remittance route mapping table; Sorting all the matching saturations to obtain a sorting result; The candidate remittance route with the highest matching saturation in the sorting results and meeting the preset business rules is selected as the transaction remittance route.

5. The transaction route matching method according to claim 1, characterized in that: After detecting whether the financial transaction result includes a transaction order, the method further includes: When detecting that the financial transaction result includes the transaction transaction record, extracting the transaction account information, transaction financial institution information, and transaction party information from the transaction transaction record; Generate a corresponding transaction remittance route and remittance route mapping based on the transaction account information, the transaction financial institution information, and the transaction parties' information, and use the transaction remittance route to authorize the transaction; When the transaction is authorized, the payment flow is initiated based on the transaction confirmation information, a settlement message is generated based on the payment flow, and the settlement message is sent to the counterparty to complete the transaction settlement process.

6. A transaction channel matching device based on natural language processing, characterized in that: include: a transaction order detection unit, configured to, upon detecting a new financial transaction result and a transaction summary, detect whether the financial transaction result includes a transaction order, and if no transaction order is detected, initiate a remittance-route self-matching mechanism; a transaction element extraction unit configured to, when the remittance-route self-matching mechanism is activated, extract counterparty information, transaction financial institution information, and beneficiary information from the transaction summary using a natural language processing model to obtain a set of transaction elements, wherein the natural language processing model is a pre-trained model for parsing summary text, and the transaction financial institution information includes at least: a transaction account bank and a transaction intermediary bank; a remittance route matching unit, configured to match each transaction element in the transaction element set with a cash remittance route mapping table, select a remittance route with the highest matching saturation and in compliance with preset business rules as the transaction remittance route, and use the transaction remittance route to complete the transaction settlement process; The route matching unit includes: a route sending module for sending route information of the transaction route to the service end; an adjustment instruction receiving module for receiving a route adjustment instruction transmitted by the service end if the service end has not confirmed the transaction route; a route rematching module for re-matching the transaction route based on the transaction-related business type and business restriction information in the route adjustment instruction to obtain a new transaction route, and sending the new transaction route to the service end; and a clearing module for completing the transaction clearing process using the transaction route upon receiving a route confirmation instruction from the service end. The transaction route matching device based on natural language processing also includes: an information table retrieval unit, which is used to retrieve the front-end transaction information table, the associated system parameter table and the transaction details corresponding to each financial product in the financial transaction result before detecting the new financial transaction result and the transaction summary; a first information extraction unit, which is used to extract the transaction information in the front-end transaction information table based on the first type of transaction element set by using a key element sorter to obtain a first transaction information set, wherein the first type of transaction element set includes at least one of the following: transaction counterparty characters, transaction type characters, transaction amount characters, and transaction date characters; a second information extraction unit, which is used to extract the transaction information in the front-end transaction information table based on the second type of transaction element set by using a key element sorter to obtain a first transaction information set. A key element set is used to extract the transaction information in the associated system parameter table using a key element sorter to obtain a second transaction information set, wherein the second transaction information set includes at least one of the following: account line characters, middle line characters, and account name characters; a third information extraction unit is used to extract the transaction information in the transaction details using a key element sorter based on the third transaction element set to obtain a third transaction information set, wherein the third transaction element set includes at least one of the following: counterparty abbreviation characters, currency characters, and transaction settlement category characters; a transaction summary generation unit is used to generate the transaction summary based on the first transaction information set, the second transaction information set, and the third transaction information set.

7. An electronic device, characterized in that: The invention comprises one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the transaction route matching method based on natural language processing as described in any one of claims 1 to 5.

8. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the transaction route matching method based on natural language processing according to any one of claims 1 to 5 are implemented.

Images