Financial bill reimbursement accounting archive system and method based on trust creation environment

Abstract

The present invention discloses a financial bill reimbursement accounting archive system and method based on a trust-based innovation environment, including: collecting historical financial bill reimbursement accounting archives for data preprocessing to form a sample data set; building a recognition model based on a natural language processing algorithm and a deep learning algorithm, recognizing the sample data set to obtain text information and bill images; storing the text information and the bill images in a table in an archive management system that applies blockchain technology to ensure that the data cannot be tampered with; combining a data analysis method, analyzing the stored text information and the bill images, and outputting the analysis results. The present invention improves the accuracy, efficiency and security of financial bill management by comprehensively utilizing natural language processing technology, deep learning algorithms and blockchain technology.

Description

Technical Field

[0001] The present invention relates to the technical field of financial management and information processing, and in particular to a financial bill reimbursement accounting archive system and method based on a trusted computing environment. Background Art

[0002] With the rapid development of information technology, especially the progress of artificial intelligence, natural language processing, deep learning and blockchain technology, the application of these technologies in financial bill management has begun to show its great potential and value; natural language processing technology makes it possible to extract useful information from large amounts of text data, while deep learning technology greatly improves the accuracy of image recognition, which is especially important when processing unstructured financial bill data; in addition, blockchain technology provides new guarantees for data security with its unique immutability. However, although these technologies have each brought improvements to financial bill management, their comprehensive application and optimization still face many challenges.

[0003] Existing technologies still have some shortcomings when dealing with financial bill management, including: although single technologies (such as NLP or deep learning) have been applied to bill recognition and classification, they usually operate independently and lack a comprehensive framework to fully utilize the combined effects of these technologies; although blockchain technology provides security for data, how to effectively integrate it with the financial bill management process and how to deal with the large amount of data brought about by it remains a problem; existing methods generally lack sufficient accuracy and efficiency in predicting reimbursement trends and identifying abnormal reimbursement activities, which limits their application value in financial decision support. Summary of the invention

[0004] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0005] In view of the above existing problems, the present invention is proposed.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] Collect historical financial bill reimbursement accounting files for data preprocessing to form a sample data set;

[0008] Building a recognition model based on a natural language processing algorithm and a deep learning algorithm to recognize the sample data set and obtain text information and bill images;

[0009] The text information and the bill image are stored in a table in a file management system using blockchain technology to ensure that the data cannot be tampered with;

[0010] Combined with the data analysis method, the stored text information and the bill image are analyzed and the analysis results are output.

[0011] As a preferred solution of the financial bill reimbursement accounting file method based on the trusted innovation environment described in the present invention, the data types of the historical financial bill reimbursement accounting files include text data, numerical data and image data.

[0012] As a preferred solution of the financial bill reimbursement accounting file method based on the trust creation environment described in the present invention, the text data at least includes the bill description, reimbursement type and supplier name;

[0013] The numerical data at least includes the bill amount, date and tax amount;

[0014] The image data at least includes a scanned or captured document image.

[0015] As a preferred solution of the financial bill reimbursement accounting archive method based on the trust creation environment described in the present invention, the data preprocessing includes data cleaning, feature extraction and data conversion; wherein:

[0016] The data cleaning: removing duplicate values, filling missing values, and correcting erroneous values;

[0017] The feature extraction: extracting keywords, phrases, and images from the historical financial bill reimbursement accounting files through natural language processing algorithms and image processing algorithms to obtain bill features;

[0018] The data conversion: converting the bill features into vector form and then standardizing them to form a high-quality data set suitable for subsequent model calculations.

[0019] As a preferred solution of the financial bill reimbursement accounting archive method based on the trust creation environment described in the present invention, a recognition model is constructed based on a natural language processing algorithm and a deep learning algorithm, including:

[0020]

[0021] Among them, R(T, I) is the comprehensive output of the recognition model, T is the text data, I is the image data, σ(T n ) is a natural language processing function used to process the nth element in the text data, φ(I j ) represents the processing of the jth image data point by the deep learning image recognition algorithm, ω j is the weight of the j-th image data point, Si (T) is the text analysis function, ψ(I k ) is the image feature extraction function, φ is the image recognition processing of the convolutional neural network, L is the total amount of text data, K is the total amount of image data, σ is the nonlinear processing parameter of the sigmoid function for text data, γ i and δ k To adjust the coefficient of feature influence, α and β are model parameters for adjusting text data and image data.

[0022] As a preferred solution of the financial bill reimbursement accounting file method based on the trust creation environment described in the present invention, combined with the data analysis method, the stored text information and the bill image are analyzed, including:

[0023]

[0024] Among them, X is the reimbursement data matrix, Y is the reimbursement activity label vector, Z is the external data vector, N is the total number of reimbursement records, M is the total number of features, and W is the total number of features. a is the feature weight, λ is the adjustment coefficient, g(t, Z) is the external data analysis function, t is the time, Y a Indicates whether the a-th reimbursement record is marked as abnormal, X ab represents the value of the a-th reimbursement record on the b-th feature, Z k is the value of the kth external data;

[0025] When Y a When it is 1, it is marked as abnormal;

[0026] When Y a When it is 0, it is marked as normal.

[0027] As a preferred solution of the financial bill reimbursement accounting file method based on the trust creation environment described in the present invention, the value range of F(X, Y, Z) is (0, +∞), where:

[0028] When the output value is close to 0, it means that the possibility of abnormal reimbursement activity is relatively small;

[0029] When the output value is close to +∞, it indicates that the possibility of abnormal reimbursement activity is relatively high.

[0030] As a preferred solution of the financial bill reimbursement accounting archive system based on the trust creation environment described in the present invention, it includes:

[0031] one or more processors;

[0032] A memory stores operable instructions, which, when executed by the one or more processors, cause the one or more processors to perform operations, including the process of the aforementioned financial bill reimbursement accounting file method based on the trusted innovation environment.

[0033] Beneficial effects of the present invention:

[0034] 1. The present invention collects and pre-processes historical financial bill reimbursement accounting files to form a sample data set, applies natural language processing technology and deep learning algorithms to establish a recognition model, and performs efficient text information recognition and bill image recognition on the sample data set. The recognized and classified information is stored in a table in the file management system, and the blockchain technology is used to ensure that the data cannot be tampered with. While ensuring data security, it also improves the transparency of data processing;

[0035] 2. Applying data analysis technology to conduct in-depth analysis of stored data can not only effectively predict the trend of financial reimbursement, but also promptly identify and handle abnormal reimbursement activities, provide decision-making support for enterprises, optimize financial management processes, and make predictions and identification more accurate and efficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

[0037] Figure 1 It is a flow chart of the financial bill reimbursement accounting file method based on the trusted innovation environment shown in the present invention. DETAILED DESCRIPTION

[0038] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0039] Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without making any creative work should fall within the scope of protection of the present invention.

[0040] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0041] Example 1

[0042] According to an embodiment of the present invention, Figure 1 The flowchart shown in the figure is a method for financial bill reimbursement accounting archive based on the trust creation environment, including:

[0043] S1: Collect historical financial bill reimbursement accounting files for data preprocessing to form a sample data set.

[0044] Among them, what needs to be explained in this step is:

[0045] The data types of historical financial bill reimbursement accounting archives include text data, numerical data and image data;

[0046] Text data includes, but is not limited to, bill description, reimbursement type, and vendor name;

[0047] Numerical data includes, but is not limited to, bill amount, date, and tax amount;

[0048] Image data includes, but is not limited to, scanned or captured images of documents.

[0049] As an example, data acquisition of historical financial bill reimbursement accounting files can be achieved by exporting bill and reimbursement-related data records from the company's internal financial software and ERP (Enterprise Resource Planning) system.

[0050] As an example, data acquisition of historical financial bill reimbursement accounting files can be done through an API interface or bank statements to collect payment and collection records directly related to the financial bills.

[0051] As an example, data acquisition of historical financial bill reimbursement accounting archives can be obtained by collecting electronic bill data, which usually exists in digital format and is convenient for automated processing.

[0052] Furthermore, data preprocessing includes data cleaning, feature extraction and data conversion; among which:

[0053] Data cleaning: remove duplicate values, fill missing values, and correct erroneous values;

[0054] Feature extraction: extract keywords, phrases, and images from historical financial bill reimbursement accounting files through natural language processing algorithms and image processing algorithms to obtain bill features;

[0055] Data conversion: Convert the invoice features into vector form and then standardize them to form a high-quality data set suitable for subsequent model calculations.

[0056] As an example, the text of financial documents is recognized through natural language processing algorithms, and each piece of text information is encoded into a vector representation to facilitate subsequent calculation and processing.

[0057] As an example, image features in financial documents can be extracted through image processing algorithms and output as vector representations through a convolutional neural network model to facilitate subsequent computational processing.

[0058] The feature extraction, standardization, and vector conversion in the aforementioned data preprocessing can all be achieved through means or methods in the prior art, and will not be described in detail in this example.

[0059] S2: Build a recognition model based on natural language processing algorithm and deep learning algorithm, recognize the sample data set, and obtain text information and bill images. Among them, what needs to be explained in this step is:

[0060]

[0061] Among them, R(T, I) is the comprehensive output of the recognition model, T is the text data, I is the image data, σ(T n ) is a natural language processing function used to process the nth element in the text data, φ(I j ) represents the processing of the jth image data point by the deep learning image recognition algorithm, ω j is the weight of the j-th image data point, S i (T) is the text analysis function, ψ(I k ) is the image feature extraction function, φ is the image recognition processing of the convolutional neural network, L is the total amount of text data, K is the total amount of image data, σ is the nonlinear processing parameter of the sigmoid function for text data, γ i and δ k To adjust the coefficient of feature influence, α and β are model parameters for adjusting text data and image data.

[0062] Preferably, this step greatly improves the accuracy and automation level of data identification by automatically extracting key information from a large amount of unstructured data, reduces the need for human intervention, and improves the ability to process large-scale data sets.

[0063] S3: Store text information and bill images in a table format in an archive management system that uses blockchain technology to ensure that the data cannot be tampered with.

[0064] In an optional embodiment, the stored ticket data is protected by utilizing the immutability of the blockchain. Whenever a ticket is identified and stored, the relevant information is created as a block and added to the blockchain. Each block contains a certain amount (e.g., 100) of ticket data. The hash value of the block depends on the hash of the previous block, forming a chain structure. Any tampering with the data will cause the hash value to mismatch, which can be easily discovered.

[0065] After the data is successfully written into the blockchain, a unique hash value is generated for each data block, and this hash value is linked to the hash value of the previous block to ensure the integrity and immutability of the data chain, which can be expressed by the following formula:

[0066]

[0067] Among them, Γ is the number of data entries contained in the current block, H block is the block hash value, indicating the immutability of the block in the blockchain, D q is the encrypted form of the qth data entry, H q-1 is the hash value of the previous block, hash(·) is the hash function used to calculate the hash value of data, τ is the attenuation factor in the weighting function, which is used to adjust the weighted value of data and enhance data security, and x represents an attribute of the data, such as a timestamp or serial number.

[0068] It should be noted that changes in the hash value reflect any slight changes in the data. Therefore, when the block hash value matches the expected value given by the system, it means that the data has not been tampered with.

[0069] Preferably, this step ensures the authenticity and integrity of the data by securely storing the identified and classified information in a tabular form in the archive management system and utilizing the immutability of blockchain technology to protect the data, thereby providing reliable technical support for financial management and auditing.

[0070] It should also be noted that the information of calculation, identification and classification of the aforementioned expected values ​​is stored in a table form, which can be achieved by means or methods in the prior art, and will not be described in detail in this example.

[0071] S4: Combine the data analysis method to analyze the stored text information and bill image, and output the analysis results. It should be noted that the mathematical expression formula of this step is as follows:

[0072]

[0073] Among them, X is the reimbursement data matrix, Y is the reimbursement activity label vector, Z is the external data vector, N is the total number of reimbursement records, M is the total number of features, and W is the total number of features. ais the feature weight, λ is the adjustment coefficient, g(t, Z) is the external data analysis function, t is the time, Y a Indicates whether the a-th reimbursement record is marked as abnormal, X ab represents the value of the a-th reimbursement record on the b-th feature, Z k is the value of the kth external data.

[0074] As an example, when Y a When it is 1, it is marked as abnormal.

[0075] As an example, when Y a When it is 0, it is marked as normal.

[0076] As an example, the external data vector is obtained by data conversion from external data, and the external data includes but is not limited to economic indexes, industry reimbursement standards, and holiday information.

[0077] As an example, the reimbursement data matrix includes but is not limited to employee information, time information and reimbursement data, wherein employee information includes but is not limited to the employee's department, position, and historical reimbursement records; time information includes but is not limited to the date and time of the reimbursement and the time required for each stage of the reimbursement process; reimbursement data includes but is not limited to the company's monthly / annual reimbursement limit and the upper limit of a single reimbursement amount.

[0078] It should also be noted that the value range of F(X, Y, Z) is (0, +∞), and its analysis result is the result corresponding to the output value of the formula, where:

[0079] When the output value is close to 0, it means that the possibility of abnormal reimbursement activity is relatively small;

[0080] When the output value is close to +∞, it indicates that the possibility of abnormal reimbursement activity is relatively high.

[0081] Preferably, through the analysis in this step, actions can be taken based on data-driven insights to optimize financial management processes, improve the efficiency of resource allocation, and effectively prevent and reduce financial risks.

[0082] It should be noted that other aspects disclosed in the embodiments of the present invention also propose a financial bill reimbursement accounting archive system based on a trusted innovation environment, including: one or more processors and a memory.

[0083] The memory is used to store operable instructions, which, when executed by the one or more processors, cause the one or more processors to perform operations, including the process of the financial bill reimbursement accounting file method based on the trust creation environment of the aforementioned embodiment, especially Figure 1 The process of the method shown.

[0084] The financial bill reimbursement accounting archive system based on the trust-based innovation environment also includes a computer-readable medium storing software, which includes instructions that can be executed by one or more computers, and these instructions enable one or more computers to perform operations through such execution. These operations include the process of the financial bill reimbursement accounting archive method based on the trust-based innovation environment of the aforementioned embodiment, especially Figure 1 The process of the method shown.

[0085] It should be appreciated that embodiments of the present invention can be implemented or carried out by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer-readable memory.

[0086] The method may be implemented in a computer program using standard programming techniques - including a non-transitory computer-readable storage medium configured with a computer program, where the storage medium so configured causes a computer to operate in a specific and predefined manner.

[0087] Each program may be implemented in a high-level procedural or object-oriented programming language to communicate with a computer system.

[0088] However, the program can be implemented in assembly or machine language, if desired.

[0089] In any case, the language may be a compiled or interpreted language.

[0090] Furthermore, the program can be run on an application specific integrated circuit programmed for this purpose.

[0091] The processes described herein (or variations and / or combinations thereof) may be performed under the control of one or more computer systems configured with executable instructions and may be implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications) that executes collectively on one or more processors, by hardware, or a combination thereof.

[0092] The computer program includes a plurality of instructions executable by one or more processors.

[0093] Further, the method may be implemented in any type of computing platform operably connected to a suitable computer, including but not limited to a personal computer, a minicomputer, a mainframe, a workstation, a network or distributed computing environment, a separate or integrated computer platform, or in communication with a charged particle tool or other imaging device, etc.

[0094] It should be noted that the above embodiments are only used to illustrate the technical solutions 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 solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A financial bill reimbursement accounting file method based on a trust-based environment, characterized in that: include: Collect historical financial bill reimbursement accounting files for data preprocessing to form a sample data set; Building a recognition model based on a natural language processing algorithm and a deep learning algorithm to recognize the sample data set and obtain text information and bill images; Build recognition models based on natural language processing algorithms and deep learning algorithms, including: in, is the comprehensive output of the recognition model, T is the text data, I is the image data, is a natural language processing function used to process the nth element in text data. represents the processing of the j-th image data point by the deep learning image recognition algorithm, is the weight of the j-th image data point, is a text analysis function, is the image feature extraction function, is the image recognition processing of convolutional neural network, L is the total amount of text data, K is the total amount of image data, is the nonlinear processing parameter of the sigmoid function used for text data, and To adjust the coefficient of feature influence, and To adjust the model parameters of text data and image data; The text information and the bill image are stored in a table in a file management system using blockchain technology to ensure that the data cannot be tampered with; Combined with the data analysis method, the stored text information and the bill image are analyzed and the analysis results are output: Among them, X is the reimbursement data matrix, Y is the reimbursement activity label vector, Z is the external data vector, N is the total number of reimbursement records, and M is the total number of features. is the feature weight, is the adjustment factor, is the external data analysis function, t is the time, Indicates whether the a-th reimbursement record is marked as abnormal. represents the value of the a-th reimbursement record on the b-th feature, is the value of the kth external data; when When it is 1, it is marked as abnormal; when When it is 0, it is marked as normal; Said The value range is (0, + ),in: When the output value is close to 0, it means that the possibility of abnormal reimbursement activity is relatively small; When the output value is close to + , it indicates that the possibility of abnormal reimbursement activity is relatively high.

2. According to the method for financial bill reimbursement accounting archives based on the trust creation environment of claim 1, it is characterized in that: The data types of the historical financial bill reimbursement accounting files include text data, numerical data and image data.

3. According to claim 2, the financial bill reimbursement accounting file method based on the trust creation environment is characterized in that: The text data at least includes a bill description, a reimbursement type, and a supplier name; The numerical data at least includes the bill amount, date and tax amount; The image data at least includes a scanned or captured document image.

4. According to claim 1, the financial bill reimbursement accounting file method based on the trust creation environment is characterized in that: The data preprocessing includes data cleaning, feature extraction and data conversion; wherein: The data cleaning: removing duplicate values, filling missing values, and correcting erroneous values; The feature extraction: extracting keywords, phrases, and images from the historical financial bill reimbursement accounting files through natural language processing algorithms and image processing algorithms to obtain bill features; The data conversion: converting the bill features into vector form and then standardizing them to form a high-quality data set suitable for subsequent model calculations.

5. A financial bill reimbursement accounting archive system based on a trust-based environment, characterized in that: include: one or more processors; A memory storing operable instructions, wherein the instructions, when executed by the one or more processors, cause the one or more processors to perform operations, wherein the operations include the process of the financial bill reimbursement accounting archive method based on the trusted innovation environment as described in any one of claims 1 to 4.

Images