Intelligent invoice duplicate checking method and device based on machine learning, equipment and storage medium

Abstract

The invention relates to an invoice duplicate checking technology, and discloses an intelligent invoice duplicate checking method and device based on machine learning, equipment and a storage medium. The method comprises the following steps: acquiring an invoice to be subjected to duplicate checking, and performing invoice personalized key field identification operation on the invoice to be subjected to duplicate checking by utilizing a pre-trained invoice key information identification model to obtain a standard key field set; and performing natural language invoice duplicate checking on the key field set by using a pre-constructed invoice database to obtain an invoice duplicate checking result. The invoice duplicate checking accuracy and efficiency can be improved.

Description

Technical Field

[0001] The present invention relates to the field of invoice duplication checking, and in particular to an intelligent invoice duplication checking method, device, equipment and computer-readable storage medium based on machine learning. Background Art

[0002] In the modern business environment, financing activities have become an important part of corporate operations and development. When applying for financing, the financing party usually needs to provide a large number of financial vouchers and invoices to prove the authenticity and legitimacy of its financial status. However, due to the large amount of invoice data and the wide variety of types, manual review of invoices is inefficient and prone to errors or omissions, making the review work in the financing process cumbersome and time-consuming. Therefore, how to efficiently and accurately check the duplication of invoices to ensure their uniqueness and authenticity has become an urgent problem to be solved.

[0003] Current invoice duplicate checking mainly relies on traditional database retrieval and simple string matching methods, which have many shortcomings when facing large-scale invoice data. For example, simple string matching is difficult to handle format differences and noise data that may exist in invoices (date formats may be different, such as "YYYY-MM-DD" or "DD-MM-YYYY"; invoice data field names may be inconsistent, such as "Invoice Number" or "Invoice No"), resulting in inaccurate duplicate checking results. Summary of the invention

[0004] The present invention provides a method, device, equipment and storage medium for intelligent invoice duplication checking based on machine learning, the main purpose of which is to improve the accuracy and efficiency of invoice duplication checking.

[0005] To achieve the above purpose, the present invention provides an intelligent invoice duplicate checking method based on machine learning, comprising: Obtain the invoice to be checked for duplicates, and use the pre-trained invoice key information recognition model to perform an invoice personalized key field recognition operation on the invoice to be checked for duplicates, and obtain a standardized key field set; The pre-built invoice database is used to perform a natural language invoice duplication check on the key field set to obtain an invoice duplication check result.

[0006] Optionally, before using the pre-trained invoice key information recognition model, the method further includes: Acquire a historical invoice sample set, and perform a data cleaning operation on the historical invoice sample set based on a preset cleaning rule to obtain a clean invoice sample set; Performing data enhancement operations based on random noise addition and format conversion on the clean invoice sample set to obtain an enhanced invoice sample set, and extracting a preset proportion of invoice samples from the enhanced invoice sample set to obtain a training set; Extracting a training sample from the training set in turn, performing field name recognition operation on the training sample using the pre-built invoice key information recognition model, performing synonym conversion on each recognized field name to obtain a standard field name, obtaining field values ​​of each standard field name, and performing standardization processing on each extracted field value to obtain a prediction field set; Using a cross entropy loss algorithm, calculating a loss value between the predicted field set and the real field set corresponding to the training sample; Minimize the loss value to obtain the network model parameters when the loss value is the smallest, and perform network reverse propagation on the network model parameters to obtain an updated invoice key information recognition model; Determining whether all training samples in the training set participate in the training process; When not all the training samples in the training set participate in the training process, returning to the above operation step of extracting one training sample from the training set in turn, and iteratively optimizing the updated invoice key information recognition model; When all the training samples in the training set participate in the training process, the training process is stopped to obtain the invoice key information recognition model corresponding to the training set; By using the K-fold cross-validation algorithm, the training set is obtained from the enhanced invoice sample set multiple times to obtain model parameters of the invoice key information recognition model corresponding to the multiple training sets, and the average operation of each model parameter is calculated to obtain the trained invoice key information recognition model.

[0007] Optionally, the pre-trained invoice key information recognition model is used to perform an invoice personalized key field recognition operation on the invoice to be checked for duplicates, and obtain a standardized key field set, including: Performing a keyword text recognition operation on the invoice to be checked for duplicates to obtain a text keyword set; According to a preset standard field name set, the text keywords in the text keyword set are converted into synonyms to obtain standard keywords, and the field values ​​corresponding to each of the standard keywords are extracted; Perform structural processing on each field value to obtain each standard field value; A standard key field set is constructed according to the correspondence between each of the standard keywords and each of the standard field values.

[0008] Optionally, the use of a pre-built invoice database to perform a natural language invoice duplicate check on the key field set to obtain an invoice duplicate check result includes: Quantizing the standard key field set to obtain a key field vector set; Using a pre-built single-layer long short-term memory network to perform vector aggregation on the key field vector set, to obtain a semantic fingerprint feature vector; Using a cosine similarity algorithm, calculate the similarity between the historical semantic fingerprint feature vector corresponding to each historical invoice in the pre-built invoice database and the semantic fingerprint feature vector; Determine whether the similarity is greater than a preset repetition threshold interval; When the similarity value is greater than the duplication threshold interval, the invoice to be checked for duplicates is determined to be a complete duplicate; When the similarity value is within the duplication threshold range, the invoice to be checked for duplicates is determined to be a partial duplicate; When the similarity value is less than the duplication threshold interval, it is determined that the invoice to be checked for duplicates is non-duplicate.

[0009] Optionally, after obtaining the invoice duplicate checking result, the method further includes: When the duplicate checking result is no duplication, the invoice to be checked for duplicates is determined to be a new invoice, and the new invoice is saved in the invoice database; When the duplicate checking result is partial duplication, the invoice to be checked for duplicate is marked as awaiting manual review; When the duplicate checking result is a complete duplicate, the invoice to be checked for duplicates is marked as a duplicate.

[0010] Optionally, after obtaining the invoice duplication check result, the method further includes: visually displaying the invoice duplication check result.

[0011] Optionally, the method further includes: regularly storing the invoice database on the blockchain.

[0012] In order to solve the above problems, the present invention also provides an intelligent invoice duplicate checking device based on machine learning, the device comprising: An intelligent field extraction module is used to obtain the invoices to be checked for duplicates, and use a pre-trained invoice key information recognition model to perform an invoice personalized key field recognition operation on the invoices to be checked for duplicates, so as to obtain a standardized key field set; The duplicate checking module is used to use the pre-built invoice database to perform invoice duplicate checking on the key field set to obtain the invoice duplicate checking result.

[0013] In order to solve the above problem, the present invention further provides an electronic device, the electronic device comprising: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the above-mentioned machine learning-based intelligent invoice duplication checking method.

[0014] In order to solve the above problems, the present invention also provides a computer-readable storage medium, in which at least one computer program is stored. The at least one computer program is executed by a processor in an electronic device to implement the above-mentioned intelligent invoice duplication checking method based on machine learning.

[0015] The embodiment of the present invention first identifies the key information in the invoice to be checked for duplicates through an invoice key information identification model, wherein the invoice key information identification model can convert each field name into a standardized synonym, and can also convert the field value into a standardized structure, ensuring that the invoice duplicate check will not fail due to factors such as the expression method and data structure; in addition, the present invention also constructs a semantic fingerprint feature vector of each historical invoice to be checked for duplicates and the invoice to be checked for duplicates through a natural language invoice duplicate check method, thereby performing a duplicate check analysis through similarity calculation to avoid duplicate check failures due to incorrect invoice information import. Therefore, the intelligent invoice duplicate check method, device, equipment and storage medium based on machine learning provided by the embodiment of the present invention can improve the accuracy and efficiency of invoice duplicate check. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of a process flow of an intelligent invoice duplicate checking method based on machine learning provided by an embodiment of the present invention; Figure 2 A functional module diagram of an intelligent invoice duplicate checking device based on machine learning provided in one embodiment of the present invention; Figure 3 A schematic diagram of the structure of an electronic device for implementing the intelligent invoice duplication checking method based on machine learning provided in one embodiment of the present invention.

[0017] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0018] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0019] The embodiment of the present application provides a method for checking duplicate invoices based on machine learning. In the embodiment of the present application, the execution subject of the method for checking duplicate invoices based on machine learning includes but is not limited to at least one of the electronic devices such as a server, a terminal, etc. that can be configured to execute the method provided in the embodiment of the present application. In other words, the method for checking duplicate invoices based on machine learning can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be an independent server, or it can be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0020] Reference Figure 1 FIG. 1 is a flow chart of an intelligent invoice duplicate checking method based on machine learning provided by an embodiment of the present invention. In this embodiment, the intelligent invoice duplicate checking method based on machine learning includes: S1. Obtain the invoice to be checked for duplicates, and use the pre-trained invoice key information recognition model to perform an invoice personalized key field recognition operation on the invoice to be checked for duplicates to obtain a standardized key field set.

[0021] The invoice to be checked for duplicates in the embodiment of the present invention refers to the invoice that the financing party wants to save to the financing party system after obtaining the invoice given by the product party and passing the query and authentication through the external third-party invoice notarization website. Before the invoice is saved in the financing party system, a duplicate check is required.

[0022] Furthermore, the invoice key information recognition model is a neural network model that combines image recognition and text recognition technology, and is specifically used to extract invoice information and convert it into a standard format.

[0023] In detail, in the embodiment of the present invention, the pre-trained invoice key information recognition model is used to perform an invoice personalized key field recognition operation on the invoice to be checked for duplicates, and a standardized key field set is obtained, including: Performing a keyword text recognition operation on the invoice to be checked for duplicates to obtain a text keyword set; According to a preset standard field name set, the text keywords in the text keyword set are converted into synonyms to obtain standard keywords, and the field values ​​corresponding to each of the standard keywords are extracted; Perform structural processing on each field value to obtain each standard field value; A standard key field set is constructed according to the correspondence between each of the standard keywords and each of the standard field values.

[0024] In the embodiment of the present invention, since the financing party may have multiple product parties and multiple financial systems, the invoices in each system may be presented in different forms, such as field names such as "Invoice Number" or "Invoice No", or date formats such as "YYYY-MM-DD" or "DD-MM-YYYY". However, they are the same information despite different expressions and structures. Therefore, it is necessary to set specifications.

[0025] In detail, in an embodiment of the present invention, the standard field name set includes the standard nouns of each field name and the standard structure of each type of field, for example, the financial data format is 1,000,000.00￥; the date data format is YYYY-MM-DD; the invoice number is named Invoice Number; ...

[0026] The embodiment of the present invention implements conversion of field names and field values ​​according to the standard field name set to obtain a standard key field set.

[0027] Furthermore, in an embodiment of the present invention, before using the pre-trained invoice key information recognition model, the method further includes: Acquire a historical invoice sample set, and perform a data cleaning operation on the historical invoice sample set based on a preset cleaning rule to obtain a clean invoice sample set; Performing data enhancement operations based on random noise addition and format conversion on the clean invoice sample set to obtain an enhanced invoice sample set, and extracting a preset proportion of invoice samples from the enhanced invoice sample set to obtain a training set; Extracting a training sample from the training set in turn, performing field name recognition operation on the training sample using the pre-built invoice key information recognition model, performing synonym conversion on each recognized field name to obtain a standard field name, obtaining field values ​​of each standard field name, and performing standardization processing on each extracted field value to obtain a prediction field set; Using a cross entropy loss algorithm, calculating a loss value between the predicted field set and the real field set corresponding to the training sample; Minimize the loss value to obtain the network model parameters when the loss value is the smallest, and perform network reverse propagation on the network model parameters to obtain an updated invoice key information recognition model; Determining whether all training samples in the training set participate in the training process; When not all the training samples in the training set participate in the training process, returning to the above operation step of extracting one training sample from the training set in turn, and iteratively optimizing the updated invoice key information recognition model; When all the training samples in the training set participate in the training process, the training process is stopped to obtain the invoice key information recognition model corresponding to the training set; By using the K-fold cross-validation algorithm, the training set is obtained from the enhanced invoice sample set multiple times to obtain model parameters of the invoice key information recognition model corresponding to the multiple training sets, and the average operation of each model parameter is calculated to obtain the trained invoice key information recognition model.

[0028] In the embodiment of the present invention, the preset data cleaning rules include processing missing values, outliers, etc. to ensure data quality.

[0029] Then, the embodiment of the present invention performs data enhancement on the clean invoice sample set by adding random noise and format conversion. The random noise addition refers to adding some randomly generated noise data to the training samples, which can help the model learn to ignore unimportant changes and focus on capturing the essential features of the data; while the format conversion is to improve the model's ability to recognize data with different structures but the same information.

[0030] Furthermore, the preset ratio is 0.7, that is, 70% of the samples are randomly extracted as the training set.

[0031] During the training process, the model is trained by the cross entropy loss algorithm in the embodiment of the present invention, and the model training progress is controlled by the K-fold cross validation algorithm. Among them, the K-fold cross validation is a statistical method used to evaluate the performance of machine learning models on independent data sets, which is mainly used for model tuning and preventing overfitting; cross entropy loss is a common method for evaluating the difference between model output and true label, especially widely used in classification problems, which will not be elaborated here.

[0032] S2. Using a pre-built invoice database, perform a natural language invoice duplication check on the key field set to obtain an invoice duplication check result.

[0033] Among them, the invoice database refers to a database that stores the local invoices of the financing party. The invoice database can be regularly uploaded to the blockchain according to preset embedding points, such as business process nodes, preset time frequencies, etc., to achieve data backup and improve data security.

[0034] In detail, in the embodiment of the present invention, the use of the pre-built invoice database to perform natural language invoice duplication check on the key field set to obtain the invoice duplication check result includes: Quantizing the standard key field set to obtain a key field vector set; Using a pre-built single-layer long short-term memory network to perform vector aggregation on the key field vector set, to obtain a semantic fingerprint feature vector; Using a cosine similarity algorithm, calculate the similarity between the historical semantic fingerprint feature vector corresponding to each historical invoice in the pre-built invoice database and the semantic fingerprint feature vector; Determine whether the similarity is greater than a preset repetition threshold interval; When the similarity value is greater than the duplication threshold interval, the invoice to be checked for duplicates is determined to be a complete duplicate; When the similarity value is within the duplication threshold range, the invoice to be checked for duplicates is determined to be a partial duplicate; When the similarity value is less than the duplication threshold interval, it is determined that the invoice to be checked for duplicates is non-duplicate.

[0035] In the embodiment of the present invention, a natural language invoice duplication checking method is used to construct a semantic fingerprint feature vector of each historical invoice to be checked for duplicates and the invoice to be checked for duplicates, so as to perform duplicate checking analysis through similarity calculation to avoid duplicate checking failure due to incorrect invoice information import.

[0036] Specifically, in the embodiment of the present invention, Word2Vec is used to perform vector aggregation on the key field vector set to obtain a semantic fingerprint feature vector; then a single-layer LSTM is used, through its internal state (cell state) and gating mechanism, to effectively process the long-distance dependency in the input sequence, and implement the weighted average of each vector or other algorithms to obtain the semantic fingerprint feature vector, and finally the cosine similarity is used to calculate the similarity between the invoice to be checked for duplicates and the historical invoices.

[0037] In the embodiment of the present invention, a repetition threshold interval (0.8, 0.95) is set. When the similarity is greater than this interval, it can be determined as a complete repetition. When it is within this interval, it is determined as a partial repetition. When it is less than this interval, it is determined as no repetition.

[0038] In the embodiment of the present invention, when the duplicate checking result is no duplication, the invoice to be checked for duplicates is determined to be a new invoice, and the new invoice is saved in the invoice database; When the duplicate checking result is partial duplication, the invoice to be checked for duplicate is marked as awaiting manual review; When the duplicate checking result is a complete duplicate, the invoice to be checked for duplicates is marked as a duplicate.

[0039] In an embodiment of the present invention, the above-mentioned invoice duplicate check results can be visualized to construct an audit window, which may include sub-windows such as invoice summary, unjudged, invoice in doubt, and AI passed, so as to enable rapid inspection by relevant staff.

[0040] The embodiment of the present invention first identifies the key information in the invoice to be checked for duplicates through an invoice key information identification model, wherein the invoice key information identification model can convert each field name into a standardized synonym, and can also convert the field value into a standardized structure, ensuring that the invoice duplicate check will not fail due to factors such as the expression method and data structure; in addition, the present invention also constructs the semantic fingerprint feature vector of each historical invoice to be checked for duplicates and the invoice to be checked for duplicates through a natural language invoice duplicate check method, thereby performing a duplicate check analysis through similarity calculation to avoid the failure of the duplicate check due to an error in the import of invoice information. Therefore, the intelligent invoice duplicate check method based on machine learning provided by the embodiment of the present invention can improve the accuracy and efficiency of invoice duplicate check.

[0041] like Figure 2 , which is a functional module diagram of an intelligent invoice duplicate checking device based on machine learning provided in one embodiment of the present invention.

[0042] The intelligent invoice duplicate checking device 100 based on machine learning of the present invention can be installed in an electronic device. According to the functions implemented, the intelligent invoice duplicate checking device 100 based on machine learning can include an intelligent field extraction module 101 and a duplicate checking module 102. The module of the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0043] In this embodiment, the functions of each module / unit are as follows: The intelligent field extraction module 101 is used to obtain the invoice to be checked for duplicates, and use the pre-trained invoice key information recognition model to perform an invoice personalized key field recognition operation on the invoice to be checked for duplicates, so as to obtain a standardized key field set; The duplicate checking module 102 is used to use the pre-built invoice database to perform invoice duplicate checking on the key field set to obtain the invoice duplicate checking result.

[0044] In detail, each module described in the intelligent invoice duplicate checking device 100 based on machine learning in the embodiment of the present application is used in the same manner as described above. Figures 1 to 3 The intelligent invoice duplication checking method based on machine learning has the same technical means as described in the text and can produce the same technical effects, so I will not go into details here.

[0045] like Figure 3 , is a schematic diagram of the structure of an electronic device 1 for implementing an intelligent invoice duplicate checking method based on machine learning provided by an embodiment of the present invention.

[0046] The electronic device 1 may include a processor 10, a memory 11, a communication bus 12 and a communication interface 13, and may also include a computer program stored in the memory 11 and executable on the processor 10, such as an intelligent invoice duplication checking program based on machine learning.

[0047] In some embodiments, the processor 10 may be composed of an integrated circuit, for example, a single packaged integrated circuit, or a plurality of packaged integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and combinations of various control chips. The processor 10 is the control core (ControlUnit) of the electronic device 1, and uses various interfaces and lines to connect various components of the entire electronic device, and executes various functions of the electronic device and processes data by running or executing programs or modules stored in the memory 11 (for example, executing a machine learning-based intelligent invoice duplicate checking program, etc.), and calling data stored in the memory 11.

[0048] The memory 11 includes at least one type of readable storage medium, and the readable storage medium includes a flash memory, a mobile hard disk, a multimedia card, a card-type memory (for example, SD or DX memory, etc.), a magnetic memory, a disk, an optical disk, etc. In some embodiments, the memory 11 may be an internal storage unit of an electronic device, such as a mobile hard disk of the electronic device. In other embodiments, the memory 11 may also be an external storage device of an electronic device, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device. Further, the memory 11 may also include both an internal storage unit of the electronic device and an external storage device. The memory 11 can not only be used to store application software and various types of data installed in the electronic device, such as the code of an intelligent invoice duplicate checking program based on machine learning, but also can be used to temporarily store data that has been output or is to be output.

[0049] The communication bus 12 may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is configured to realize connection and communication between the memory 11 and at least one processor 10, etc.

[0050] The communication interface 13 is used for communication between the above-mentioned electronic device 1 and other devices, including a network interface and a user interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device and other electronic devices. The user interface may be a display (Display), an input unit (such as a keyboard (Keyboard)), and optionally, the user interface may also be a standard wired interface, a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an OLED (Organic Light-Emitting Diode, organic light-emitting diode) touch device, etc. Among them, the display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the electronic device and to display a visual user interface.

[0051] Figure 3 Only an electronic device with components is shown, and those skilled in the art will understand that Figure 3 The structure shown does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0052] For example, although not shown, the electronic device 1 may also include a power source (such as a battery) for supplying power to various components. Preferably, the power source may be logically connected to the at least one processor 10 through a power management device, so that the power management device can realize functions such as charging management, discharging management, and power consumption management. The power source may also include one or more DC or AC power sources, recharging devices, power failure detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The electronic device 1 may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.

[0053] It should be understood that the embodiment is for illustration only and the scope of the patent application is not limited to this structure.

[0054] The intelligent invoice duplicate checking program based on machine learning stored in the memory 11 of the electronic device 1 is a combination of multiple instructions. When running in the processor 10, it can achieve: Obtain the invoice to be checked for duplicates, and use the pre-trained invoice key information recognition model to perform an invoice personalized key field recognition operation on the invoice to be checked for duplicates, and obtain a standardized key field set; The pre-built invoice database is used to perform a natural language invoice duplication check on the key field set to obtain an invoice duplication check result.

[0055] Specifically, the specific implementation method of the processor 10 for the above instructions can refer to the description of the relevant steps in the corresponding embodiment of the accompanying drawings, which will not be repeated here.

[0056] Furthermore, if the module / unit integrated in the electronic device 1 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. The computer-readable storage medium can be volatile or non-volatile. For example, the computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).

[0057] The present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor of an electronic device, the computer program can implement: Obtain the invoice to be checked for duplicates, and use the pre-trained invoice key information recognition model to perform an invoice personalized key field recognition operation on the invoice to be checked for duplicates, and obtain a standardized key field set; The pre-built invoice database is used to perform a natural language invoice duplication check on the key field set to obtain an invoice duplication check result.

[0058] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0059] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0060] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0061] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0062] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is limited by the appended claims rather than the above description, so it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any attached figure mark in the claims should not be regarded as limiting the claims involved.

[0063] The blockchain referred to in this invention is a new application mode of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, encryption algorithm, etc. Blockchain is essentially a decentralized database, a string of data blocks generated by cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify the validity of its information (anti-counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, platform product service layer, and application service layer.

[0064] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0065] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the system claim can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

[0066] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. An intelligent invoice duplicate checking method based on machine learning, characterized in that: The method comprises: Obtain the invoice to be checked for duplicates, and use the pre-trained invoice key information recognition model to perform an invoice personalized key field recognition operation on the invoice to be checked for duplicates, and obtain a standardized key field set; The pre-built invoice database is used to perform a natural language invoice duplication check on the key field set to obtain an invoice duplication check result.

2. The intelligent invoice duplicate checking method based on machine learning as claimed in claim 1, characterized in that: Before using the pre-trained invoice key information recognition model, the method further includes: Acquire a historical invoice sample set, and perform a data cleaning operation on the historical invoice sample set based on a preset cleaning rule to obtain a clean invoice sample set; Performing data enhancement operations based on random noise addition and format conversion on the clean invoice sample set to obtain an enhanced invoice sample set, and extracting a preset proportion of invoice samples from the enhanced invoice sample set to obtain a training set; Extracting a training sample from the training set in turn, performing field name recognition operation on the training sample using the pre-built invoice key information recognition model, performing synonym conversion on each recognized field name to obtain a standard field name, obtaining field values ​​of each standard field name, and performing standardization processing on each extracted field value to obtain a prediction field set; Using a cross entropy loss algorithm, calculating a loss value between the predicted field set and the real field set corresponding to the training sample; Minimize the loss value to obtain the network model parameters when the loss value is the smallest, and perform network reverse propagation on the network model parameters to obtain an updated invoice key information recognition model; Determining whether all training samples in the training set participate in the training process; When not all the training samples in the training set participate in the training process, returning to the above operation step of extracting one training sample from the training set in turn, and iteratively optimizing the updated invoice key information recognition model; When all the training samples in the training set participate in the training process, the training process is stopped to obtain the invoice key information recognition model corresponding to the training set; By using the K-fold cross-validation algorithm, the training set is obtained from the enhanced invoice sample set multiple times to obtain model parameters of the invoice key information recognition model corresponding to the multiple training sets, and the average operation of each model parameter is calculated to obtain the trained invoice key information recognition model.

3. The intelligent invoice duplicate checking method based on machine learning as claimed in claim 1, characterized in that: The pre-trained invoice key information recognition model is used to perform an invoice personalized key field recognition operation on the invoice to be checked for duplicates, and a standardized key field set is obtained, including: Performing a keyword text recognition operation on the invoice to be checked for duplicates to obtain a text keyword set; According to a preset standard field name set, the text keywords in the text keyword set are converted into synonyms to obtain standard keywords, and the field values ​​corresponding to each of the standard keywords are extracted; Perform structural processing on each field value to obtain each standard field value; A standard key field set is constructed according to the correspondence between each of the standard keywords and each of the standard field values.

4. The intelligent invoice duplicate checking method based on machine learning as claimed in claim 1, characterized in that: The method of using the pre-built invoice database to perform natural language invoice duplication check on the key field set to obtain the invoice duplication check result includes: Quantizing the standard key field set to obtain a key field vector set; Using a pre-built single-layer long short-term memory network to perform vector aggregation on the key field vector set, to obtain a semantic fingerprint feature vector; Using a cosine similarity algorithm, calculate the similarity between the historical semantic fingerprint feature vector corresponding to each historical invoice in the pre-built invoice database and the semantic fingerprint feature vector; Determine whether the similarity is greater than a preset repetition threshold interval; When the similarity value is greater than the duplication threshold interval, the invoice to be checked for duplicates is determined to be a complete duplicate; When the similarity value is within the duplication threshold range, the invoice to be checked for duplicates is determined to be a partial duplicate; When the similarity value is less than the duplication threshold interval, it is determined that the invoice to be checked for duplicates is non-duplicate.

5. The intelligent invoice duplicate checking method based on machine learning as claimed in claim 4, characterized in that: After obtaining the invoice duplicate checking result, the method further includes: When the duplicate checking result is no duplication, the invoice to be checked for duplicates is determined to be a new invoice, and the new invoice is saved in the invoice database; When the duplicate checking result is partial duplication, the invoice to be checked for duplicate is marked as awaiting manual review; When the duplicate checking result is a complete duplicate, the invoice to be checked for duplicates is marked as a duplicate.

6. The intelligent invoice duplicate checking method based on machine learning as claimed in claim 5, characterized in that: After obtaining the invoice duplicate checking result, the method further includes: visually displaying the invoice duplicate checking result.

7. The intelligent invoice duplicate checking method based on machine learning as claimed in claim 1, characterized in that: The method also includes: regularly storing the invoices in the invoice database on the blockchain.

8. An intelligent invoice duplicate checking device based on machine learning, characterized in that: The device comprises: An intelligent field extraction module is used to obtain the invoices to be checked for duplicates, and use a pre-trained invoice key information recognition model to perform an invoice personalized key field recognition operation on the invoices to be checked for duplicates, so as to obtain a standardized key field set; The duplicate checking module is used to use the pre-built invoice database to perform invoice duplicate checking on the key field set to obtain the invoice duplicate checking result.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the machine learning-based intelligent invoice duplication checking method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the intelligent invoice duplication checking method based on machine learning is implemented as described in any one of claims 1 to 7.

Images