Bill reimbursement method and device, storage medium and computer equipment

Abstract

The invention discloses a bill reimbursement method and device, a storage medium and computer equipment, and mainly aims at solving the problems that existing bill reimbursement is not timely, the time cost is high and errors are prone to occurring, and the method comprises the steps that tasks to be reimbursed are divided into business travel type reimbursement tasks and budget application type reimbursement tasks based on application bill information; executing a first workflow corresponding to the business travel type reimbursement task, or executing a second workflow corresponding to the budget application type reimbursement task; acquiring bill picture information in the workflow; performing information extraction processing on the bill picture information to obtain a corresponding bill text; carrying out classification processing on the bill picture information to obtain a corresponding bill type; recommending a matched reimbursement business receipt to the user based on a preset matching rule, the application receipt information and the receipt type; and pre-filling the reimbursement business document, and sending the obtained pre-filled reimbursement business document to a corresponding user for confirmation to obtain a target reimbursement business document.

Description

Technical Field

[0001] The present invention relates to the technical field of bill information processing, and particularly to a method and device for bill reimbursement, a storage medium, and a computer device. Background Art

[0002] Currently, there are mainly two situations for bill reimbursement within an enterprise: one is business travel reservation bill reimbursement, and the other is reimbursement based on bills. For business travel reservation bill reimbursement, there is a clear business trip plan before the trip, and then it is reserved through a business travel reservation platform and reimbursed after the business trip ends. For reimbursement based on bills, the handler needs to manually fill in the content and attach relevant bill certificates. However, business travel reservation bill reimbursement is often restricted by the time when the reimbursement personnel submit reimbursement information, which often affects business travel settlement. Reimbursement based on bills is not only time-consuming and laborious but also easily affected by human errors, resulting in delays or failures in reimbursement applications. Therefore, the current two bill reimbursement technologies are increasingly unable to meet the requirements of efficient financial work. Summary of the Invention

[0003] In view of this, the present invention provides a method and device for bill reimbursement, a storage medium, and a computer device, mainly aiming to solve the problems of untimely bill reimbursement, high time cost, and easy errors in the existing bill reimbursement.

[0004] According to one aspect of the present invention, there is provided a method for bill reimbursement, including:

[0005] Based on the application document information submitted by the user, dividing the reimbursement task to be processed into a business travel reimbursement task and a budget application reimbursement task;

[0006] Executing a first workflow corresponding to the business travel reimbursement task; the first workflow includes obtaining first bill picture information; or, executing a second workflow corresponding to the budget application reimbursement task; the second workflow includes obtaining second bill picture information;

[0007] Performing information extraction processing on the first bill picture information or the second bill picture information to obtain corresponding bill text; and performing classification processing on the first bill picture information or the second bill picture information to obtain corresponding bill types;

[0008] Recommending a matching reimbursement business document to the user based on a preset matching rule, the application document information, and the bill type;

[0009] Pre-filling the reimbursement business document based on the bill text and the application document information to obtain a pre-filled reimbursement business document; and sending the pre-filled reimbursement business document to the corresponding user for confirmation processing to obtain a target reimbursement business document.

[0010] Further, the information extraction process for the first bill picture information or the second bill picture information to obtain the corresponding bill text includes:

[0011] Obtain historical bill picture information, and perform annotation processing on the key bill text in the historical bill picture information to obtain the annotated historical bill picture information;

[0012] Use the annotated historical bill picture information to perform model training on a natural language processing model to obtain a trained key information recognition model;

[0013] Use the key information recognition model to perform key information extraction on the first bill picture information or the second bill picture information to obtain the corresponding bill text.

[0014] Further, the classification process for the first bill picture information or the second bill picture information to obtain the corresponding bill type includes:

[0015] Obtain historical bill picture information, and perform annotation processing on the bill type in the historical bill picture information to obtain a bill classification training sample;

[0016] Use the bill classification training sample to perform training on an image classification model to obtain a trained bill type classification model;

[0017] Use the bill type classification model to perform classification on the first bill picture information or the second bill picture information to obtain the corresponding bill type.

[0018] Further, before recommending a matching reimbursement business document to the user based on the preset matching rule, the application document information, and the bill type, the method further includes:

[0019] Divide various application document information into travel-related documents or budget application documents;

[0020] Perform secondary division on the travel-related documents to obtain various travel-related documents to be matched; and perform secondary division on the budget application documents to obtain various budget application documents to be matched;

[0021] Configure corresponding reimbursement business documents and bill types for various travel-related documents to be matched to obtain the preset matching rule; or, configure corresponding reimbursement business documents and bill types for various budget application documents to be matched to obtain the preset matching rule.

[0022] Further, recommending a reimbursement business document that matches the user based on the preset matching rule, the application document information, and the bill type includes:

[0023] Obtaining the preset matching rule corresponding to the application document information;

[0024] Performing a query process on the preset matching rule based on the application document information and the bill type to obtain the matching reimbursement business document; and recommending the reimbursement business document to the relevant user.

[0025] Further, pre-filling the reimbursement business document based on the bill text and the application document information to obtain a pre-filled reimbursement business document includes:

[0026] Obtaining the application content in the application document information;

[0027] Comparing the application content with the relevant text in the bill text to obtain a relevant text comparison result;

[0028] Filling the application content and the bill text corresponding to the reasonable relevant text comparison result into the reimbursement business document to obtain the pre-filled reimbursement business document.

[0029] Further, after sending the pre-filled reimbursement business document to the corresponding user for confirmation processing to obtain a target reimbursement business document, the method further includes:

[0030] Performing a compliance check process on the target reimbursement business document based on a preset compliance check rule and marking the pre-filled content that does not meet the compliance check rule;

[0031] Performing an anomaly analysis process on the target reimbursement business document based on a preset calculation model and marking the anomalous pre-filled content.

[0032] According to another aspect of the present invention, there is provided a bill reimbursement device, including:

[0033] A document division module, configured to divide the reimbursement task to be processed into a business trip reimbursement task and a budget application reimbursement task based on the application document information submitted by the user;

[0034] A workflow execution module, configured to execute a first workflow corresponding to the business trip reimbursement task; the first workflow includes obtaining first bill picture information; or, executing a second workflow corresponding to the budget application reimbursement task; the second workflow includes obtaining second bill picture information;

[0035] An extraction and classification module for performing information extraction processing on the first bill picture information or the second bill picture information to obtain corresponding bill texts; and performing classification processing on the first bill picture information or the second bill picture information to obtain corresponding bill types;

[0036] A recommendation module for recommending matching reimbursement business documents to the user based on preset matching rules, the application document information, and the bill type;

[0037] A pre-filling and confirmation module for pre-filling the reimbursement business document based on the bill text and the application document information to obtain a pre-filled reimbursement business document; and sending the pre-filled reimbursement business document to the corresponding user for confirmation processing to obtain a target reimbursement business document.

[0038] Further, the extraction and classification module includes a text extraction unit, and the text extraction unit is used for:

[0039] Obtaining historical bill picture information, and performing annotation processing on key bill texts in the historical bill picture information to obtain annotated historical bill picture information;

[0040] Using the annotated historical bill picture information to perform model training processing on a natural language processing model to obtain a trained key information recognition model;

[0041] Using the key information recognition model to perform key information extraction processing on the first bill picture information or the second bill picture information to obtain the corresponding bill text.

[0042] Further, the extraction and classification module further includes a bill classification unit, and the bill classification unit is used for:

[0043] Obtaining historical bill picture information, and performing annotation processing on the bill type in the historical bill picture information to obtain a bill classification training sample;

[0044] Using the bill classification training sample to perform training processing on an image classification model to obtain a trained bill type classification model;

[0045] Using the bill type classification model to perform classification processing on the first bill picture information or the second bill picture information to obtain the corresponding bill type.

[0046] Further, the device further includes a matching rule setting module, and the matching rule setting module is used for:

[0047] Dividing various application document information into travel-related documents or budget application documents;

[0048] Perform secondary classification on the business trip vouchers to obtain various business trip vouchers to be matched; and perform secondary classification on the budget application vouchers to obtain various budget application vouchers to be matched.

[0049] Configure corresponding reimbursement business vouchers and voucher types for each of the business trip vouchers to be matched to obtain the preset matching rules; or, configure corresponding reimbursement business vouchers and voucher types for each of the budget application vouchers to be matched to obtain the preset matching rules.

[0050] Further, the recommendation module is further configured to:

[0051] Obtain the preset matching rules corresponding to the application voucher information.

[0052] Perform query processing in the preset matching rules based on the application voucher information and the voucher type to obtain the matching reimbursement business vouchers; and recommend the reimbursement business vouchers to relevant users.

[0053] Further, the pre-filling and confirmation module includes an information pre-filling unit, and the information pre-filling unit is configured to:

[0054] Obtain the application content in the application voucher information.

[0055] Compare the application content with the relevant text in the voucher text to obtain a relevant text comparison result.

[0056] Fill the application content and the voucher text corresponding to the reasonable relevant text comparison result into the reimbursement business voucher to obtain the pre-filled reimbursement business voucher.

[0057] Further, the device further includes an audit module, and the audit module is configured to:

[0058] Perform compliance check processing on the target reimbursement business voucher based on preset compliance check rules, and mark the pre-filled content that does not meet the compliance check rules.

[0059] Perform abnormal analysis processing on the target reimbursement business voucher based on a preset calculation model, and mark the abnormal pre-filled content.

[0060] According to another aspect of the present invention, there is provided a storage medium storing at least one executable instruction, and the executable instruction causes a processor to perform operations corresponding to the above-mentioned method for voucher reimbursement.

[0061] According to another aspect of the present invention, there is provided a computer device, including a processor, a memory, a communication interface and a communication bus, and the processor, the memory and the communication interface complete communication with each other through the communication bus;

[0062] The memory is used to store at least one executable instruction, and the executable instruction causes the processor to execute the operations corresponding to the method of bill reimbursement as described above.

[0063] By means of the above technical solutions, the technical solutions provided by the embodiments of the present invention have at least the following advantages:

[0064] The present invention provides a method and device for bill reimbursement, a storage medium, and a computer device. Compared with the prior art, the present invention divides the tasks to be reimbursed into business trip reimbursement tasks and budget application reimbursement tasks based on the application document information submitted by the user; then executes the first workflow corresponding to the business trip reimbursement tasks; or executes the second workflow corresponding to the budget application reimbursement tasks; separates the reimbursement scenarios with different requirements for processing, making the method of the present invention more widely applicable. The bill text and bill type are also obtained through information extraction processing and classification processing respectively, and the reimbursement business documents that match are recommended to the user based on the preset matching rules, application document information and bill type; and the reimbursement business documents are pre-filled based on the bill text and application document information to obtain pre-filled reimbursement business documents, realizing the automatic matching and automatic filling of reimbursement business documents, not only improving the efficiency and timeliness of bill reimbursement, but also reducing the error rate of bill reimbursement, saving time cost and labor cost for bill reimbursement.

[0065] The above description is only an overview of the technical solutions of the present invention. In order to be able to understand the technical means of the present invention more clearly, it can be implemented according to the content of the description. And in order to make the above and other purposes, features and advantages of the present invention more obvious and understandable, the following specifically illustrates the specific embodiments of the present invention. Description of the Drawings

[0066] By reading the detailed description of the preferred embodiments below, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention. And throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0067] Figure 1 It shows a schematic flowchart of a method for bill reimbursement provided by an embodiment of the present invention;

[0068] Figure 2 It shows a schematic flowchart of another method for bill reimbursement provided by an embodiment of the present invention;

[0069] Figure 3 It shows a schematic flowchart of another method for bill reimbursement provided by an embodiment of the present invention;

[0070] Figure 4 It shows a schematic flowchart of yet another method for bill reimbursement provided by an embodiment of the present invention;

[0071] Figure 5 It shows a schematic structural diagram of a device for bill reimbursement provided by an embodiment of the present invention;

[0072] Figure 6 It shows a schematic structural diagram of a computer device provided by an embodiment of the present invention. Detailed implementation manners

[0073] Hereinafter, exemplary embodiments of the present disclosure will be described in more detail with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art.

[0074] An embodiment of the present invention provides a method for bill reimbursement. As Figure 1 shown, the method includes:

[0075] 101. Based on the application document information submitted by the user, divide the reimbursement tasks to be processed into business trip reimbursement tasks and budget application reimbursement tasks;

[0076] In an embodiment of the present invention, before performing bill reimbursement, the user needs to submit corresponding application documents according to specific reimbursement requirements. For example, business trip application forms, vacation application forms, training fee budget forms, expert consultation fee budget forms, etc. are not specifically limited in the embodiments of the present invention. Based on the application document information submitted by the user, the current execution end divides the reimbursement tasks to be processed into business trip reimbursement tasks and budget application reimbursement tasks. For example, the reimbursement tasks corresponding to application document information such as business trip application forms and vacation application forms are divided into business trip reimbursement tasks; the reimbursement tasks corresponding to application document information such as training fee budget forms and expert consultation fee budget forms are divided into budget application reimbursement tasks, etc., which are not specifically limited in the embodiments of the present invention.

[0077] 102. Execute a first workflow corresponding to the business trip reimbursement task; the first workflow includes obtaining first bill picture information; or, execute a second workflow corresponding to the budget application reimbursement task; the second workflow includes obtaining second bill picture information;

[0078] In an embodiment of the present invention, the current execution end executes a first workflow corresponding to a business travel reimbursement task, where the first workflow represents a preset workflow corresponding to the business travel reimbursement task and includes obtaining first bill picture information. For example, relevant reservation information of the corresponding user on various business travel reservation platforms is obtained through an interface, including flight reservation relevant information, train ticket reservation relevant information, bus ticket reservation relevant information, hotel reservation relevant information, etc., which are not specifically limited in the embodiments of the present invention. In addition, after the corresponding user has fulfilled the reserved content, first bill picture information corresponding to each reservation relevant information is obtained, including airline itinerary receipts, train tickets, passenger bus tickets, passenger ship tickets, fixed-amount invoices, special VAT invoices, ordinary VAT invoices, electronic special VAT invoices, electronic ordinary VAT invoices, etc., which are not specifically limited in the embodiments of the present invention. Or,

[0079] In an embodiment of the present invention, the current execution end executes a second workflow corresponding to a budget application reimbursement task, where the second workflow represents a preset workflow corresponding to the budget application reimbursement task and includes obtaining second bill picture information. For example, special VAT invoices, ordinary VAT invoices, electronic special VAT invoices, electronic ordinary VAT invoices, etc. related to the budget application uploaded by the corresponding user are directly obtained, which are not specifically limited in the embodiments of the present invention.

[0080] 103. Perform information extraction processing on the first bill picture information or the second bill picture information to obtain corresponding bill text; and perform classification processing on the first bill picture information or the second bill picture information to obtain corresponding bill types;

[0081] In an embodiment of the present invention, the current execution end needs to perform information extraction processing on the first bill picture information or the second bill picture information obtained in step 102 to obtain corresponding bill text. Among them, information extraction processing means using optical character recognition technology (OCR) to analyze and identify the text information on the bill picture information according to preset rules. For example, when a user uploads a round-trip train ticket, the system can read the bill text through OCR technology, which may include the departure place, departure time, destination, return time, etc., which are not specifically limited in the embodiments of the present invention.

[0082] In addition, the current execution end also needs to perform classification processing based on the first bill image information or the second bill image information obtained in step 102 to obtain the corresponding bill type. Among them, the classification processing represents the process of classifying according to the inherent feature information of the bill image. For example, the first bill image information or the second bill image information is divided into airline itinerary receipts, train tickets, passenger bus tickets, passenger ship tickets, special VAT invoices, ordinary VAT invoices, electronic special VAT invoices, electronic ordinary VAT invoices, digital electronic ordinary invoices, digital electronic special invoices, fixed-amount invoices, toll invoices, etc. The embodiments of the present invention do not make specific limitations.

[0083] 104. Recommend matching reimbursement business documents to the user based on the preset matching rules, the application document information, and the bill type;

[0084] In the embodiments of the present invention, the current execution end recommends matching reimbursement business documents to the user based on the preset matching rules, the application document information submitted by the user in step 101, and the bill type determined in step 103. Among them, the preset matching rules record the corresponding relationships between each reimbursement business document, the application document information submitted by the user, and the bill type. The matching reimbursement business document can be directly determined by means of searching, which improves the efficiency and automation level of the user's bill reimbursement.

[0085] 105. Pre-fill the reimbursement business document based on the bill text and the application document information to obtain a pre-filled reimbursement business document; and send the pre-filled reimbursement business document to the corresponding user for confirmation processing to obtain the target reimbursement business document.

[0086] In the embodiments of the present invention, the current execution end pre-fills the matching reimbursement business document recommended to the user in step 104 based on the bill text extracted in step 103 and the application document information submitted by the user in step 101 to obtain a pre-filled reimbursement business document. Among them, annotation information can be added to the bill text after extraction, such as adding annotations related to the departure place, departure time, destination, return time, etc. When specifically pre-filling, it can be pre-filled based on the content of the annotation. Among them, the application document information submitted by the user in step 101 can be annotated at the submission stage, and when specifically pre-filling, it is also pre-filled based on the content of the annotation.

[0087] In an embodiment of the present invention, after the current execution end obtains the pre-filled reimbursement business document, it is also necessary to send the pre-filled reimbursement business document to the corresponding user. After receiving the pre-filled reimbursement business document, the user can review whether the document type is correct, review the pre-filled content, etc., which are not specifically limited in the embodiments of the present invention. After the user reviews and believes that the pre-filled reimbursement business document conforms to the actual reimbursement situation, a confirmation message is sent to the current execution end. After receiving the confirmation message from the user, the current execution end determines the pre-filled reimbursement business document as the target reimbursement business document.

[0088] It should be noted that when the user believes that the pre-filled reimbursement business document does not conform to the actual reimbursement situation, the user can also perform manual modification processing on the pre-filled reimbursement business document. After receiving the manual modification information from the user, the current execution end specially displays the manual modification information. For example, the content manually modified by the user is displayed in different colored fonts, the content manually modified by the user is displayed in a shaded form, the content manually modified by the user is displayed in a framed form, etc., which are not specifically limited in the embodiments of the present invention. In the embodiments of the present invention, the content manually modified by the user is displayed and processed in a special display manner, which is convenient for later financial review processing and later improvement processing of the bill reimbursement method, which are not specifically limited in the embodiments of the present invention.

[0089] Furthermore, as a refinement and extension of the specific implementation manner of the above embodiment, in order to accurately extract the key information in the bill image information, another bill reimbursement method is provided, such as Figure 2 shown, the step of performing information extraction processing on the first bill image information or the second bill image information to obtain the corresponding bill text includes:

[0090] 201. Obtain historical bill image information, and perform annotation processing on the key bill text in the historical bill image information to obtain the annotated historical bill image information;

[0091] In an embodiment of the present invention, the current execution end obtains historical bill picture information. For example, it obtains historical bill pictures such as airline tickets, train tickets, passenger bus tickets, passenger ship tickets, special VAT invoices, ordinary VAT invoices, electronic special VAT invoices, electronic ordinary VAT invoices, digital electronic ordinary invoices, digital electronic special invoices, fixed-amount invoices, and toll invoices. The embodiments of the present invention do not make specific limitations. And it performs annotation processing on the key bill text in the historical bill picture information to obtain the annotated historical bill picture information. Among them, the key information includes the occurrence address, occurrence date, etc. The embodiments of the present invention do not make specific limitations. Specifically, when performing annotation processing on the key bill text in the historical bill picture information, it can extract the bill text and bill type based on a pre-trained model of a Chinese corpus that combines a convolutional recurrent neural network (CRNN) and a loss function (CTC) for processing sequence annotation tasks. The embodiments of the present invention do not make specific limitations.

[0092] 202. Use the annotated historical bill picture information to perform model training processing on the natural language processing model to obtain a trained key information recognition model;

[0093] In an embodiment of the present invention, the current execution end uses the annotated historical bill picture information in step 201 to perform model training processing on the natural language processing (NLP) model to obtain a trained key information recognition model for extracting key bill information. In an embodiment of the present invention, the BERT (Bidirectional Encoder Representations from Transformers) series models in the natural language processing (NLP) model are preferably used, such as the RoBerta (A Robustly Optimized BERT Pretraining Approach) model, etc. The embodiments of the present invention do not make specific limitations.

[0094] 203. Use the key information recognition model to perform key information extraction processing on the first bill picture information or the second bill picture information to obtain the corresponding bill text.

[0095] In an embodiment of the present invention, the current execution end uses the trained key information recognition model in step 202 to perform key information extraction processing on the first bill picture information or the second bill picture information to obtain the corresponding bill text.

[0096] It should be noted that since the historical bill picture information is annotated in step 201, the type of key information can be identified during key information extraction. For example, the departure station, arrival station, travel time, etc. in the train ticket can be identified, and corresponding tags are added to the identified key information to facilitate subsequent pre-filling of the corresponding reimbursement business form.

[0097] Further, as a refinement and extension of the specific implementation manner of the above embodiment, in order to accurately classify various bill picture information, another method for bill reimbursement is provided. For example, Figure 3 As shown in the figure, the steps of classifying the first bill picture information or the second bill picture information to obtain the corresponding bill type include:

[0098] 301. Obtain historical bill picture information, and perform annotation processing on the bill types in the historical bill picture information to obtain bill classification training samples;

[0099] In the embodiment of the present invention, the current execution end obtains historical bill picture information. For example, historical bill pictures such as air tickets, train tickets, passenger bus tickets, passenger ship tickets, special VAT invoices, ordinary VAT invoices, electronic special VAT invoices, electronic ordinary VAT invoices, digital electronic ordinary invoices, digital electronic special invoices, fixed-amount invoices, and toll invoices are obtained. The embodiment of the present invention does not make specific limitations. And perform bill type annotation processing on the obtained historical bill picture information to obtain bill classification training samples. For example, label all air tickets with "air ticket", label all train tickets with "train ticket", etc., or use "type 1", "type 2", etc. to label the above two types of bills. The embodiment of the present invention does not make specific limitations.

[0100] 302. Use the bill classification training samples to train an image classification model to obtain a trained bill type classification model;

[0101] In the embodiment of the present invention, the current execution end uses the bill classification training samples obtained after the annotation processing in step 301 to train an image classification model to obtain a trained bill type classification model. Among them, the goal of the image classification model is to assign the bill classification training samples to predefined category labels according to different image features in the bill classification training samples. In the embodiment of the present invention, the Vit model is preferably used for training processing to obtain a trained bill type classification model, which can realize the classification of bill types.

[0102] 303. Use the bill type classification model to classify the first bill picture information or the second bill picture information to obtain the corresponding bill type.

[0103] In an embodiment of the present invention, the current execution end classifies the first bill picture information or the second bill picture information by using the bill type classification model trained in step 302 to obtain the corresponding bill type. Specifically, the first bill picture information or the second bill picture information is used as the input of the bill type classification model. After being recognized by the bill type classification model, the corresponding category label is output, and then the bill type corresponding to the first bill picture information or the second bill picture information is determined based on the output category label.

[0104] Further, as a refinement and extension of the specific implementation manner of the above embodiment, in order to quickly match the reimbursement business documents required by the finance department, another bill reimbursement method is provided. For example, Figure 4 As shown, before the step of recommending the matching reimbursement business documents to the user based on the preset matching rule, the application document information, and the bill type, the method further includes:

[0105] 401. Classify various pieces of the application document information into travel-related documents or budget application documents;

[0106] In an embodiment of the present invention, the current execution end classifies various pieces of application document information into travel-related documents or budget application documents. For example, the application document information such as business trip application forms, vacation application forms, and communication transportation record forms is classified into travel-related documents; the application document information such as training fee budget forms, expert consultation fee budget forms, bid evaluation meeting fee budget forms, and group fee budget forms is classified into budget application documents. The embodiment of the present invention does not make specific limitations.

[0107] 402. Perform secondary classification on the travel-related documents to obtain various to-be-matched travel-related documents; and perform secondary classification on the budget application documents to obtain various to-be-matched budget application documents;

[0108] In an embodiment of the present invention, the current execution end performs secondary classification on the travel-related documents completed in step 401 to obtain various to-be-matched travel-related documents. Among them, the to-be-matched travel-related documents can be set according to the company's business. For example, the business trip application form and the communication transportation record form are classified into the same to-be-matched travel-related document; the vacation application form and the compensatory leave form are classified into the same to-be-matched travel-related document, etc. The embodiment of the present invention does not make specific limitations. Similarly, the current execution end also performs secondary classification on the budget application documents completed in step 401 to obtain various to-be-matched budget application documents. The embodiment of the present invention does not make specific limitations.

[0109] 403. Configure the corresponding reimbursement business documents and bill types for various to-be-matched travel-related documents respectively to obtain the preset matching rule; or, configure the corresponding reimbursement business documents and bill types for various to-be-matched budget application documents respectively to obtain the preset matching rule.

[0110] In an embodiment of the present invention, the current execution end respectively configures corresponding reimbursement business documents and invoice types for various to-be-matched business trip-related documents to obtain a preset matching rule. For example, for the to-be-matched business trip-related documents corresponding to a business trip application form and a communication transportation record form, a business trip expense reimbursement form can be configured, and at the same time, invoice types such as airline itinerary receipts, train tickets, passenger bus tickets, passenger ship tickets, fixed-amount invoices, special VAT invoices, ordinary VAT invoices, electronic special VAT invoices, and electronic ordinary VAT invoices are configured. The embodiments of the present invention do not make specific limitations.

[0111] In an embodiment of the present invention, the current execution end also respectively configures corresponding reimbursement business documents and invoice types for various to-be-matched budget application documents to obtain the preset matching rule. For example, for the to-be-matched business trip-related documents corresponding to a training fee budget form and an expert consultation fee budget form, a general reimbursement form can be configured, and at the same time, invoice types such as special VAT invoices, ordinary VAT invoices, electronic special VAT invoices, and electronic ordinary VAT invoices are configured. The embodiments of the present invention do not make specific limitations.

[0112] The step of recommending a matching reimbursement business document to the user based on the preset matching rule, the application document information, and the invoice type includes:

[0113] 404. Obtain the preset matching rule corresponding to the application document information;

[0114] 405. Perform a query process in the preset matching rule based on the application document information and the invoice type to obtain the matching reimbursement business document; and recommend the reimbursement business document to the relevant user.

[0115] In an embodiment of the present invention, the current execution end obtains the preset matching rule corresponding to the application document information submitted by the user in step 101, and then performs a query process in the preset matching rule based on the application document information and the classified invoice type. For example, if the user submits a business trip application form and obtains invoice picture information through an interface, after classification processing, the corresponding invoice types of the obtained invoice picture information are airline itinerary receipts and ordinary VAT invoices. By searching in the preset matching rule, it is determined that the business trip expense reimbursement form is the matching reimbursement business document, and the business trip expense reimbursement form is recommended to the relevant user. The embodiments of the present invention do not make specific limitations.

[0116] Further, as a refinement and extension of the specific implementation manner of the above embodiment, in order to further improve the automation degree of invoice reimbursement, another invoice reimbursement method is provided. The step of pre-filling the reimbursement business document based on the invoice text and the application document information to obtain a pre-filled reimbursement business document includes:

[0117] Obtain the application content in the said application document information;

[0118] Compare and process the said application content with the relevant text in the bill text to obtain the relevant text comparison result;

[0119] Fill the application content and the corresponding bill text for which the relevant text comparison result is reasonable into the said reimbursement business document to obtain the pre-filled reimbursement business document.

[0120] In the embodiment of the present invention, the current execution end obtains the application content in the application document information submitted by the user. Taking the business trip application information as an example, information such as the destination and time is obtained, which is not specifically limited in the embodiment of the present invention. Then, the obtained application content is compared and processed with the relevant text in the bill text. For example, compare the destination in the business trip application information with the destination obtained from the air itinerary; compare the business trip time in the business trip application information with the departure time obtained from the air itinerary; if the comparison result of the destination is within a reasonable range and the comparison result of the departure time is also within a reasonable range, then the comparison result of the relevant text is determined to be reasonable, and the application content and the bill text are filled into the determined reimbursement business document correspondingly to obtain the pre-filled reimbursement business document, which is not specifically limited in the embodiment of the present invention. If any one of the comparison results of the destination and the departure time is not within a reasonable range, then the comparison result of the relevant text is determined to be unreasonable, and it is necessary to feedback the unreasonable relevant information to the user so that the user can correct the bill or the application document information based on the unreasonable relevant information.

[0121] Further, as a refinement and extension of the specific implementation manner of the above embodiment, in order to improve the audit efficiency of the financial department, another method for bill reimbursement is provided. After the step of sending the pre-filled reimbursement business document to the corresponding user for confirmation processing to obtain the target reimbursement business document, the method further includes:

[0122] Conduct a compliance check process on the target reimbursement business document based on a preset compliance check rule, and mark the pre-filled content that does not meet the compliance check rule;

[0123] Conduct an anomaly analysis process on the target reimbursement business document based on a preset calculation model, and mark the anomaly pre-filled content.

[0124] In the embodiments of the present invention, after the current execution end obtains the target reimbursement business document, it also obtains a preset compliance check rule, and performs compliance check processing on the target reimbursement business document based on the preset compliance check rule. Among them, the compliance check processing may include auditing the business logic and process compliance of the document, auditing the compliance of the total reimbursement amount range, auditing the compliance of the range of individual item amounts, etc., and the embodiments of the present invention do not make specific limitations. The current execution end also needs to mark the pre-filled content that does not meet the preset compliance check rule, so as to facilitate the finance department to quickly find the places that do not meet the compliance requirements during the audit, and improve the audit efficiency of the finance department.

[0125] Similarly, after the current execution end obtains the target reimbursement business document, it also obtains a preset calculation model, and performs anomaly analysis processing on the target reimbursement business document based on the preset calculation model. Among them, the anomaly analysis processing may include calculation anomalies of bill amounts, calculation anomalies of tax rate applications, etc., and the embodiments of the present invention do not make specific limitations. The current execution end also needs to mark the pre-filled content with anomalies, so as to facilitate the finance department to quickly find the anomalies during the audit, and improve the audit efficiency of the finance department.

[0126] It should be noted that in the embodiments of the present invention, the current execution end can also set up a background management module. The background management module mainly provides powerful data analysis functions, supports multi-dimensional data statistics and trend analysis. The system can automatically generate various financial statements and analysis reports to help financial personnel comprehensively understand the financial status of the enterprise and provide data support for decision-making.

[0127] The embodiments of the present invention provide a method for bill reimbursement. Compared with the prior art, the present invention divides the reimbursement tasks to be processed into business trip reimbursement tasks and budget application reimbursement tasks based on the application document information submitted by the user; then executes the first workflow corresponding to the business trip reimbursement tasks; or executes the second workflow corresponding to the budget application reimbursement tasks; separates the reimbursement scenarios with different requirements for processing, making the method of the present invention more widely applicable. It also obtains the bill text and bill type respectively through information extraction processing and classification processing, and recommends matching reimbursement business documents to the user based on the preset matching rules, application document information and bill type; also pre-fills the reimbursement business documents based on the bill text and application document information to obtain pre-filled reimbursement business documents, realizing the automatic matching and automatic filling of reimbursement business documents, not only improving the efficiency and timeliness of bill reimbursement, but also reducing the error rate of bill reimbursement, saving time cost and labor cost for bill reimbursement.

[0128] As an implementation of the method shown above Figure 1 The embodiments of the present invention provide a bill reimbursement device, as shown in Figure 5 shown, the device includes:

[0129] The document division module 51 is used to divide the reimbursement tasks to be processed into business trip reimbursement tasks and budget application reimbursement tasks based on the application document information submitted by the user.

[0130] The workflow execution module 52 is used to execute a first workflow corresponding to the business trip reimbursement task; the first workflow includes obtaining first bill picture information; alternatively, execute a second workflow corresponding to the budget application reimbursement task; the second workflow includes obtaining second bill picture information.

[0131] The extraction and classification module 53 is used to perform information extraction processing on the first bill picture information or the second bill picture information to obtain corresponding bill text; and perform classification processing on the first bill picture information or the second bill picture information to obtain corresponding bill types.

[0132] The recommendation module 54 is used to recommend matching reimbursement business documents to the user based on preset matching rules, the application document information, and the bill types.

[0133] The pre-filling and confirmation module 55 is used to pre-fill the reimbursement business document based on the bill text and the application document information to obtain a pre-filled reimbursement business document; and send the pre-filled reimbursement business document to the corresponding user for confirmation processing to obtain a target reimbursement business document.

[0134] Further, the extraction and classification module 53 includes a text extraction unit, and the text extraction unit is used to:

[0135] Obtain historical bill picture information, and perform annotation processing on the key bill text in the historical bill picture information to obtain the annotated historical bill picture information.

[0136] Use the annotated historical bill picture information to perform model training processing on a natural language processing model to obtain a trained key information recognition model.

[0137] Use the key information recognition model to perform key information extraction processing on the first bill picture information or the second bill picture information to obtain the corresponding bill text.

[0138] Further, the extraction and classification module 53 further includes a bill classification unit, and the bill classification unit is used to:

[0139] Obtain historical bill picture information, and perform annotation processing on the bill types in the historical bill picture information to obtain bill classification training samples.

[0140] Using the bill classification training samples to train an image classification model, a trained bill type classification model is obtained;

[0141] Using the bill type classification model to classify the first bill picture information or the second bill picture information, the corresponding bill type is obtained.

[0142] Further, the device further includes a matching rule setting module, and the matching rule setting module is used for:

[0143] Dividing various application document information into travel - related documents or budget application documents;

[0144] Performing secondary division on the travel - related documents to obtain multiple to - be - matched travel - related documents; and performing secondary division on the budget application documents to obtain multiple to - be - matched budget application documents;

[0145] Configuring corresponding reimbursement business documents and bill types for various to - be - matched travel - related documents respectively to obtain the preset matching rules; or, configuring corresponding reimbursement business documents and bill types for various to - be - matched budget application documents respectively to obtain the preset matching rules.

[0146] Further, the recommendation module 54 is further used for:

[0147] Obtaining the preset matching rules corresponding to the application document information;

[0148] Performing query processing based on the application document information and the bill type in the preset matching rules to obtain the matching reimbursement business documents; and recommending the reimbursement business documents to relevant users.

[0149] Further, the pre - filling and confirmation module 55 includes an information pre - filling unit, and the information pre - filling unit is used for:

[0150] Obtaining the application content in the application document information;

[0151] Comparing the application content with relevant texts in the bill text to obtain a relevant text comparison result;

[0152] Filling the application content and the bill text corresponding to the reasonable relevant text comparison result into the reimbursement business document to obtain the pre - filled reimbursement business document.

[0153] Further, the device further includes an audit module, and the audit module is used for:

[0154] Perform compliance check processing on the target reimbursement business document based on preset compliance check rules, and mark the pre-filled content that does not meet the compliance check rules;

[0155] Perform anomaly analysis processing on the target reimbursement business document based on a preset calculation model, and mark the anomalous pre-filled content.

[0156] An embodiment of the present invention provides a device for bill reimbursement. Compared with the prior art, the present invention divides the reimbursement tasks to be processed into business trip reimbursement tasks and budget application reimbursement tasks based on the application document information submitted by the user; then executes the first workflow corresponding to the business trip reimbursement tasks; or, executes the second workflow corresponding to the budget application reimbursement tasks; separates the reimbursement scenarios with different requirements for processing, making the method of the present invention more widely applicable. Also, the bill text and bill type are obtained through information extraction processing and classification processing respectively, and the reimbursement business documents that match the user are recommended based on the preset matching rules, application document information, and bill type; also, the reimbursement business documents are pre-filled based on the bill text and application document information to obtain pre-filled reimbursement business documents, realizing the automatic matching and automatic filling of the reimbursement business documents, not only improving the efficiency and timeliness of bill reimbursement, but also reducing the error rate of bill reimbursement, saving time costs and labor costs for bill reimbursement.

[0157] According to an embodiment of the present invention, a storage medium is provided. The storage medium stores at least one executable instruction, and the computer executable instruction can execute the bill reimbursement method in any of the above method embodiments.

[0158] Figure 6 The structural schematic diagram of a computer device provided according to an embodiment of the present invention is shown. The specific implementation of the computer device is not limited in the specific embodiments of the present invention.

[0159] As Figure 6 shown, the computer device may include: a processor 602, a communication interface 604, a memory 606, and a communication bus 608.

[0160] Among them: The processor 602, the communication interface 604, and the memory 606 communicate with each other through the communication bus 608.

[0161] The communication interface 604 is used to communicate with network elements of other devices such as clients or other servers.

[0162] The processor 602 is used to execute the program 610, and specifically can execute the relevant steps of the above bill reimbursement method.

[0163] Specifically, the program 610 may include program code that includes computer operation instructions.

[0164] The processor 602 may be a central processing unit (CPU), or a specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention. One or more processors included in the computer device may be of the same type of processor, such as one or more CPUs; or may be of different types of processors, such as one or more CPUs and one or more ASICs.

[0165] The memory 606 is used to store the program 610. The memory 606 may include high-speed RAM memory and may also include non-volatile memory, such as at least one disk memory.

[0166] The program 610 may specifically be used to cause the processor 602 to perform the following operations:

[0167] Based on the application document information submitted by the user, divide the reimbursement tasks to be processed into business trip reimbursement tasks and budget application reimbursement tasks;

[0168] Execute a first workflow corresponding to the business trip reimbursement task; the first workflow includes obtaining first bill picture information; or execute a second workflow corresponding to the budget application reimbursement task; the second workflow includes obtaining second bill picture information;

[0169] Perform information extraction processing on the first bill picture information or the second bill picture information to obtain the corresponding bill text; and perform classification processing on the first bill picture information or the second bill picture information to obtain the corresponding bill type;

[0170] Recommend matching reimbursement business documents to the user based on a preset matching rule, the application document information, and the bill type;

[0171] Pre-fill the reimbursement business document based on the bill text and the application document information to obtain a pre-filled reimbursement business document; and send the pre-filled reimbursement business document to the corresponding user for confirmation processing to obtain a target reimbursement business document.

[0172] Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the present invention can be implemented by a general-purpose computing device. They can be concentrated on a single computing device or distributed on a network composed of multiple computing devices. Optionally, they can be implemented by program codes executable by the computing device. Thus, they can be stored in a storage device and executed by the computing device. And in some cases, the steps shown or described can be executed in a sequence different from here, or they can be separately fabricated into individual integrated circuit modules, or multiple modules or steps among them can be fabricated into a single integrated circuit module for implementation. In this way, the present invention is not limited to any specific combination of hardware and software.

[0173] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for bill reimbursement, characterized in that: include: Based on the application document information submitted by the user, the tasks to be reimbursed are divided into travel reimbursement tasks and budget application reimbursement tasks; Executing a first workflow corresponding to the travel reimbursement task; the first workflow includes obtaining first bill image information; Alternatively, executing a second workflow corresponding to the budget application reimbursement task; the second workflow includes obtaining second bill image information; Performing information extraction processing on the first bill image information or the second bill image information to obtain a corresponding bill text; and performing classification processing on the first bill image information or the second bill image information to obtain a corresponding bill type; Recommending matching reimbursement business documents to the user based on preset matching rules, the application document information and the bill type; The reimbursement business document is pre-filled based on the bill text and the application document information to obtain a pre-filled reimbursement business document; and the pre-filled reimbursement business document is sent to a corresponding user for confirmation processing to obtain a target reimbursement business document.

2. The method according to claim 1, characterized in that: The extracting process of the first bill image information or the second bill image information to obtain the corresponding bill text includes: Acquire historical bill image information, and annotate key bill text in the historical bill image information to obtain annotated historical bill image information; The annotated historical bill image information is used to perform model training processing on a natural language processing model to obtain a trained key information recognition model; The key information recognition model is used to perform key information extraction processing on the first bill image information or the second bill image information to obtain the corresponding bill text.

3. The method according to claim 1, characterized in that The classifying and processing the first bill image information or the second bill image information to obtain a corresponding bill type includes: Obtaining historical bill image information, and labeling the bill types in the historical bill image information to obtain bill classification training samples; The image classification model is trained using the bill classification training samples to obtain a trained bill type classification model; The first bill image information or the second bill image information is classified and processed using the bill type classification model to obtain the corresponding bill type.

4. The method according to claim 1, characterized in that Before recommending matching reimbursement business documents to the user based on the preset matching rules, the application document information and the bill type, the method further includes: Classify the various application document information into travel documents or budget application documents; The travel documents are divided twice to obtain a plurality of travel documents to be matched; and the budget application documents are divided twice to obtain a plurality of budget application documents to be matched; The corresponding reimbursement business documents and bill types are configured for each of the travel documents to be matched to obtain the preset matching rules; or, the corresponding reimbursement business documents and bill types are configured for each of the budget application documents to be matched to obtain the preset matching rules.

5. The method according to claim 4, characterized in that The recommending matching reimbursement business documents to the user based on the preset matching rules, the application document information and the bill type includes: Acquire the preset matching rule corresponding to the application document information; Based on the application document information and the bill type, query processing is performed in the preset matching rule to obtain the matching reimbursement business document; and the reimbursement business document is recommended to relevant users.

6. The method according to claim 1, characterized in that The pre-filling of the reimbursement business document based on the bill text and the application document information to obtain the pre-filled reimbursement business document includes: Obtaining application content from the application document information; Compare the application content with the relevant text in the bill text to obtain a relevant text comparison result; The application content and the bill text that are reasonable as a result of the relevant text comparison are filled into the reimbursement business document accordingly to obtain the pre-filled reimbursement business document.

7. The method according to any one of claims 1 to 6, characterized in that: After sending the pre-filled reimbursement business document to the corresponding user for confirmation processing and obtaining the target reimbursement business document, the method further includes: Performing compliance check processing on the target reimbursement business document based on preset compliance check rules, and marking pre-filled content that does not meet the compliance check rules; Based on a preset calculation model, the target reimbursement business document is analyzed and processed for abnormalities, and abnormal pre-filled content is marked.

8. A device for bill reimbursement, characterized in that: include: The document classification module is used to classify the reimbursement tasks into travel reimbursement tasks and budget application reimbursement tasks based on the application document information submitted by the user; A workflow execution module, configured to execute a first workflow corresponding to the travel reimbursement task; the first workflow includes obtaining first bill image information; Alternatively, executing a second workflow corresponding to the budget application reimbursement task; the second workflow includes obtaining second bill image information; An extraction and classification module, configured to perform information extraction processing on the first bill image information or the second bill image information to obtain a corresponding bill text; and perform classification processing on the first bill image information or the second bill image information to obtain a corresponding bill type; A recommendation module, used to recommend matching reimbursement business documents to the user based on preset matching rules, the application document information and the bill type; The pre-filling and confirmation module is used to pre-fill the reimbursement business document based on the bill text and the application document information to obtain a pre-filled reimbursement business document; and send the pre-filled reimbursement business document to the corresponding user for confirmation processing to obtain a target reimbursement business document.

9. A storage medium, wherein at least one executable instruction is stored in the storage medium, and the executable instruction executes an operation corresponding to the method for bill reimbursement according to any one of claims 1 to 7.

10. A computer device, comprising a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute an operation corresponding to the method for bill reimbursement according to any one of claims 1-7.

Images