Automatic process processing system and method for accounting data

Abstract

The invention relates to the field of accounting data processing, in particular to an accounting data automatic process processing system and method. The data acquisition module is used for acquiring original data in accounting business; the data preprocessing module is connected with the data acquisition module, and the data preprocessing module is used for preprocessing the original data to obtain standard data; the intelligent classification module is connected with the data preprocessing module, and the intelligent classification module is used for classifying the standard data based on a preset classification model to obtain at least two pieces of classification data; the flow processing module is connected with the intelligent classification module, and the flow processing module is used for carrying out flow processing on each classification data according to a preset processing flow so as to respectively obtain processing results; and the report generation module is connected with the process processing module, and the report generation module is used for generating a corresponding accounting report according to each processing result. According to the invention, through series connection of multiple modules, the working efficiency of accounting work is improved.

Description

Technical Field

[0001] The present invention relates to the field of accounting data processing, and in particular to an accounting data automated process processing system and method. Background Art

[0002] In the field of accounting data processing, as businesses continue to expand, accounting data has become large in volume, diverse in types, and complex in structure. Traditional accounting data processing methods mostly rely on a combination of manual operations and preset rules.

[0003] For example, during data classification, staff members need to manually categorize various types of accounting data into different categories based on their own experience and fixed classification standards. When new business models emerge, such as e-commerce platforms’ livestreaming sales revenue sharing and financial companies’ innovative financial product revenue accounting, manual classification of these data, which lack historical experience and have ambiguous data characteristics, is not only inefficient but also prone to classification errors, impacting subsequent financial analysis and decision-making. Furthermore, even when using preset rules for classification, the applicability of these rules is greatly reduced when faced with issues such as inconsistent data formats and missing information.

[0004] For example, in terms of data processing, most current accounting data processing systems operate in a fragmented fashion. After data collection, the lack of a unified data cleaning and preprocessing mechanism allows a large amount of redundant and erroneous data to flow directly into subsequent processing, resulting in slow classification and calculation modules and even erroneous results. During report generation, irregularities in initial data processing force staff to spend significant time on secondary verification and adjustments, making it difficult to automate the entire process from data collection to report output. This significantly impacts accounting efficiency and increases both labor and time costs for businesses.

[0005] In summary, in the prior art, both manual and computer work efficiency in accounting work is low, so the technical problem actually solved by the present invention is how to improve work efficiency in accounting work. Summary of the Invention

[0006] In order to overcome the technical defect of low work efficiency in accounting work in the above-mentioned prior art, the purpose of the present invention is to provide an accounting data automation process processing system and method, which improves the work efficiency in accounting work by connecting multiple modules in series.

[0007] The present invention discloses an accounting data automation process processing system, which includes a data acquisition module, a data pre-processing module, an intelligent classification module, a process processing module and a report generation module; wherein, The data collection module is used to collect raw data in accounting business; The data preprocessing module is connected to the data acquisition module, and the data preprocessing module is used to preprocess the original data to obtain standard data; The intelligent classification module is connected to the data preprocessing module, and is used to classify the standard data based on a preset classification model to obtain at least two classification data; The process processing module is connected to the intelligent classification module, and the process processing module is used to process each classification data according to the preset processing flow to obtain processing results respectively; The report generation module is connected to the process processing module, and the report generation module is used to generate corresponding accounting reports according to each processing result.

[0008] Preferably, the intelligent classification module includes a data feature extraction unit and a classification decision unit; wherein, The data feature extraction unit is used to extract feature information from standard data; The classification decision unit is used to make classification decisions based on the feature information extracted by the data feature extraction unit and a preset classification model to obtain classification data; The classification model is built using a deep learning algorithm, and the classification decision is calculated as follows: in, Indicated as data X belongs to category The probability of Represented as category The probability of data X appearing in , Represented as category The prior probability of occurrence, Expressed as the total number of categories, Represents the first categories, is the index variable, .

[0009] Preferably, the data feature extraction unit extracts feature information from the standard data based on a convolutional neural network; The convolutional layer of the convolutional neural network has a preset calculation formula, which is: in, Represented as the mth feature map of the lth layer, Represented as a set of input feature maps connected to the mth feature map, Expressed as The h-th feature map of the layer, It is represented as the convolution kernel of the lth layer connecting the hth input feature map and the mth input feature map. Denoted as the bias of the m-th feature map, ” represents a convolution operation.

[0010] Preferably, the process processing module includes at least two sub-processing units, each sub-processing unit corresponding to at least one classification data; The sub-processing unit processes the corresponding classified data based on preset processing rules, which include data calculation and data association analysis; When the sub-processing unit performs data calculation, it calculates based on the following formula: Among them, R represents the calculation result, Represented as the kth classification data item, It is represented as the weight corresponding to the k-th classification data item, and p is represented as the total number of classification data items; When performing data association analysis, the sub-processing unit mines data association rules based on the Apriori algorithm.

[0011] Preferably, the report generation module includes a report database and a data filling unit; wherein There are several accounting report templates preset in the report database. The data filling unit selects the corresponding accounting report template according to the processing results obtained by the process processing module, and fills the processing results into the selected accounting report template to generate an accounting report.

[0012] Preferably, it also includes a storage module connected to the data acquisition module, the data pre-processing module, the intelligent classification module, the process processing module and the report generation module respectively; The storage module is used to establish a storage subunit whenever the data collection module collects original data in accounting business. The storage subunit stores the original data, standard data, classified data, processing results and accounting reports in each accounting business in real time.

[0013] Preferably, it further comprises a verification module connected to the data acquisition module, the data pre-processing module, the intelligent classification module, the process processing module, the report generation module and the storage module respectively; The verification module is used to verify the consistency of the data stored and the data in the original module each time the storage subunit stores data. If the data stored is inconsistent with the data in the original module, the storage subunit stores the data again. The verification module is also used to reversely obtain the inverse original data when the storage sub-unit stores the processing results, and compare the inverse original data with the original data to determine the consistency. When the inverse original data is inconsistent with the original data, the storage module re-stores the data of the original module step by step. When the inverse original data is consistent with the original data, the storage sub-unit stores the judgment result of the verification module.

[0014] A second object of the present invention is to provide a processing method based on the above-mentioned accounting data automated process processing system, comprising the following steps: Collect various raw data in accounting business; Preprocess the raw data to obtain standard data; classifying the standard data based on the classification model to obtain at least two classification data; Perform process processing on each classified data to obtain processing results respectively; Generate corresponding accounting reports based on the processing results.

[0015] After adopting the above technical scheme, compared with the existing technology, the beneficial effect of the present invention is that it improves the work efficiency during accounting work by connecting multiple modules in series; through the intelligent classification module, the preset classification model is used to automatically classify standard data without human intervention, and it can quickly adapt to the original data of accounting business, with high classification accuracy, and can effectively process data with diverse formats and incomplete information, thereby improving the accuracy of classification; by constructing a full-process automation module for data, each module is connected and data flows automatically; the data preprocessing module filters invalid data, and the process processing module automatically executes corresponding rules according to the classified data, and the report is automatically generated, and the response improves processing efficiency and reduces labor costs; automated processing reduces human intervention, effectively avoids some unnecessary errors, and improves the accuracy and reliability of accounting business. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a flow chart of an automated accounting data processing system according to the present invention; Figure 2 The figure is a schematic diagram of the steps of an automated accounting data processing method according to the present invention. DETAILED DESCRIPTION

[0017] The advantages of the present invention are further described below with reference to the accompanying drawings and specific embodiments.

[0018] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.

[0019] The terms used in this disclosure are for the purpose of describing specific embodiments only and are not intended to limit the disclosure. As used in this disclosure and the appended claims, the singular forms "a," "an," "the," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.

[0020] It should be understood that although the terms first, second, third, etc. may be used in this disclosure to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from each other. For example, without departing from the scope of this disclosure, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining."

[0021] In the description of the present invention, it should be understood that the terms "longitudinal", "transverse", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention.

[0022] In the description of the present invention, unless otherwise specified and limited, it should be noted that the terms "installed", "connected" and "connected" should be understood in a broad sense. For example, it can be a mechanical connection or an electrical connection, or it can be the internal communication between two components. It can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to the specific circumstances.

[0023] In the following description, the suffixes such as "module", "component" or "unit" used to represent elements are only used to facilitate the description of the present invention and have no specific meaning. Therefore, "module" and "component" can be used interchangeably.

[0024] In this embodiment, an accounting data automation process processing system is provided, including a data acquisition module, a data preprocessing module, an intelligent classification module, a process processing module and a report generation module; wherein the data acquisition module is used to collect original data in accounting business; the data preprocessing module is connected to the data acquisition module, and the data preprocessing module is used to preprocess the original data to obtain standard data; the intelligent classification module is connected to the data preprocessing module, and the intelligent classification module is used to classify the standard data based on a preset classification model to obtain at least two classified data; the process processing module is connected to the intelligent classification module, and the process processing module is used to process each classified data according to a preset processing process to obtain processing results respectively; the report generation module is connected to the process processing module, and the report generation module is used to generate corresponding accounting reports according to each processing result.

[0025] See Figure 1 As shown, in this embodiment, an accounting data automation process processing system will be described in detail. The system mainly includes a data acquisition module, a data preprocessing module, an intelligent classification module, a process processing module and a report generation module.

[0026] Data Collection Module: This module is used to collect raw data related to accounting operations during the course of accounting. Collection methods include, but are not limited to, file import, database integration, web crawler, and API interface collection. Collected raw data includes, but is not limited to, financial accounting data and business operations data. Financial accounting data includes, but is not limited to, revenue, costs, and asset management data. Business operations data includes, but is not limited to, procurement, sales, and manufacturing data.

[0027] File import and collection supports data extraction from various electronic file formats, such as Excel, CSV, and PDF. Enterprises can use the system's file import function to directly transfer existing accounting data files, such as sales reports and supplier invoices, to the data collection module. The data collection module automatically identifies the data structure and content of the imported accounting data files based on pre-set file parsing rules, extracting key information and collecting it as raw data. Database integration and collection integrates with various existing internal databases, such as enterprise resource databases and customer relationship databases. The data collection module uses standard data interface protocols, such as ODBC (Open Database Connectivity) and JDBC (Java Database Connectivity), to achieve real-time data interaction with databases and obtain raw data. Web crawler collection utilizes web crawler technology to collect data from public financial data sources or the enterprise's own online business platforms. By setting specific crawling rules and target URLs, the web collection module accesses web pages, parses the HTML page structure, and extracts raw data related to accounting operations. API interface collection, through data interaction with third-party service platforms or partners' systems through API (application programming), thereby collecting raw data in accounting business.

[0028] Data preprocessing module: used to preprocess the original data to obtain standard data. The preprocessing methods of the data preprocessing module include but are not limited to data cleaning, format conversion and data standardization, and data integration to obtain standard data.

[0029] Data cleaning includes but is not limited to removing duplicate data, processing missing data, and correcting erroneous data. Removing duplicate data: The original data may contain duplicate or erroneous data. The pre-processing module will scan the data according to the preset duplication judgment rules. If duplicate original data is scanned, only the most recently entered original data will be retained. Processing missing data: For missing fields in the original data, such as supplier contact information, fill them through historical statistical features. It should be noted that if there is any content related to accounting and finance, it will be supplemented manually. Correcting erroneous data: The original data may have problems such as entry errors during the collection process, such as decimal point misplacement, etc. By setting data verification rules, the entered original data can be verified and corrected.

[0030] Format conversion includes but is not limited to unifying data formats and data type conversions. Unifying data formats: Since the raw data comes from different data sources, there are differences in format. The data preprocessing module will convert these raw data in different formats into a set standard format to facilitate processing by subsequent modules. Data type conversion: Convert data types according to processing requirements. For example, convert string type date data to date type to facilitate date calculation and sorting; convert character type amount data to numeric type to facilitate mathematical operations. It should also be noted that during the conversion process, the data preprocessing module will check the type compatibility of the raw data to ensure accuracy and completeness.

[0031] Data standardization includes, but is not limited to, normalization and encoding. Normalization: Normalizes raw numerical data to ensure that data of different magnitudes have the same weight and comparability in subsequent calculations and analyses. Encoding: For categorical data, the data preprocessing module converts it into numerical data using encoding.

[0032] Data integration includes, but is not limited to, linked data merging and data splitting. Linked data merging: Raw data may be scattered across different data sets. The data preprocessing module will merge the relevant raw data based on the relationships between the received raw data. Data splitting: For fields containing excessive information, the data preprocessing module will split them as needed.

[0033] Intelligent classification module: used to classify the standard data based on a preset classification model to obtain at least two classification data, and the classification is performed based on the feature information of the standard data.

[0034] The process-based processing module is used to process categorized data based on a pre-defined process flow to obtain processing results. Specifically, the process-based processing module is equipped with multiple sub-processing units, each dedicated to processing a specific type of categorized data, to improve the targeted and professional nature of categorized data processing. Each sub-processing unit is equipped with processing rules to process the corresponding categorized data and obtain processing results.

[0035] Report generation module: used to generate corresponding accounting reports based on each obtained processing result.

[0036] Furthermore, the intelligent classification module includes a data feature extraction unit and a classification decision unit; wherein the data feature extraction unit is used to extract feature information from the standard data; the classification decision unit is used to make classification decisions based on a preset classification model according to the feature information extracted by the data feature extraction unit to obtain classified data; the classification model is constructed using a deep learning algorithm, and the calculation formula for the classification decision is: .in, Indicated as data X belongs to category The probability of Represented as category The probability of data X appearing in , Represented as category The prior probability of occurrence, Expressed as the total number of categories, Represents the first categories, is the index variable, .

[0037] In this embodiment, the intelligent classification module will be described in detail. The intelligent classification module includes a data feature extraction unit and a classification decision unit. The data feature extraction unit is used to extract feature information from standard data and convert it into a numerical vector that can be processed by the algorithm. The classification decision unit is used to make classification decisions based on the feature information extracted by the data feature extraction unit and a preset classification model to obtain classified data. The classification decision calculation formula is: .in, Indicated as data X belongs to category The probability of Represented as category The probability of data X appearing in , Represented as category The prior probability of occurrence, Expressed as the total number of categories, Represents the first categories, is the index variable, .

[0038] For example, in a company's accounting department, employee insurance documents are classified into "travel expenses" ( ), "Business entertainment expenses" ( ), "office expenses" ( ) three categories, that is, n = 3. There is a reimbursement document data X to be classified. The document shows that the reimbursement amount is 5,000 yuan, the reimbursement date is October 15, 2024, the reimbursement person is traveling to Shanghai, and the reimbursement reason is "business negotiation with Shanghai customers."

[0039] Prior probability is the probability of occurrence of various items based on historical data statistics. According to statistics, in the company's past reimbursement data, the proportion of "travel expenses" appearing is 40%, that is, ; The proportion of "business entertainment expenses" is 30%, that is, ; The proportion of "office expenses" is 30%, that is, .

[0040] Calculate likelihood probability , the likelihood probability represents the probability of the known category The probability of data X appearing under the condition of . "Travel expenses" category: In the historical "travel expenses" data, the data of traveling to Shanghai and reimbursement amount of about 5,000 yuan, and the reimbursement reason is related to business negotiation accounts for 10%. Therefore, "Business entertainment expenses" category: In the historical "business entertainment expenses" data, the data of going to Shanghai, reimbursing an amount of 5,000 and the reason for business negotiation accounted for 60%, so "Office Expenses" category: In the historical "Office Expenses" data, the probability of this type of data appearing is 5%, so .

[0041] Calculate the posterior probability : Travel expense category: “Business entertainment expenses” category: "Office Expenses" Category: According to the above calculation, the posterior probabilities of the three categories are obtained, where Therefore, this reimbursement document is classified as "business entertainment expenses".

[0042] Furthermore, the data feature extraction unit extracts feature information from the standard data based on a convolutional neural network. A calculation formula is preset in the convolutional layer of the convolutional neural network, and the calculation formula in the convolutional layer is: .in, Represented as the mth feature map of the lth layer, Represented as a set of input feature maps connected to the mth feature map, Expressed as The h-th feature map of the layer, It is represented as the convolution kernel of the lth layer connecting the hth input feature map and the mth input feature map. Denoted as the bias of the m-th feature map, ” represents a convolution operation.

[0043] In this embodiment, the data feature extraction unit will be described in detail. The data feature extraction unit extracts feature information from standard data based on a convolutional neural network. The convolutional neural network includes an input layer, a convolution layer, a pooling layer, an activation function layer, a fully connected layer, and an input layer. It should be noted that the convolutional neural network as a whole is a mature technology in the existing art, so it will not be explained in detail in this embodiment. Only the convolution layer of the designed scheme will be described.

[0044] The calculation formula is preset in the convolutional layer of the convolutional neural network. The calculation formula in the convolutional layer is: .in, Represented as the mth feature map of the lth layer, Represented as a set of input feature maps connected to the mth feature map, Expressed as The h-th feature map of the layer, It is represented as the convolution kernel of the lth layer connecting the hth input feature map and the mth input feature map. Denoted as the bias of the m-th feature map, ” represents a convolution operation.

[0045] For example, let's take a company's sales invoice data processing as an example to explain the calculation formula in detail. Assume that the company receives a large number of sales invoices every month and needs to extract features such as invoice number, customer name, sales amount, sales date, etc. from the invoice data to facilitate subsequent classification. Now select a sales invoice data and convert it into a matrix form that can be processed by the computer as input data

[0046] Assume that the convolutional neural network is currently in the lth layer and needs to calculate the mth feature map The set of input feature maps connected to the mth feature map is , assuming Contains 3 input feature maps (corresponding to the invoice number, sales amount, and sales date in the invoice). The convolution kernel of the lth layer connecting the hth input feature map and the mth output feature map is To simplify the explanation, we set It is a 1×3 matrix, which corresponds to the weight coefficients of extracting invoice number, sales amount, and sales date features, such as = , indicating that the sales amount has a higher weight in the privilege claim. The bias of the mth feature map is set up Input features In the code, the data corresponding to the invoice number, sales amount, and sales date are respectively , , .

[0047] Calculate the value of the mth feature map according to the formula:

[0048] In this example, through the calculation of the convolution layer, multiple key information in the sales invoice is weighted and summed through the convolution kernel and biased to obtain a comprehensive feature value. It integrates the features of invoice number, sales amount, sales date and other information, which are then used by the intelligent classification module to make classification decisions. For example, the feature values ​​obtained after the convolution layer is calculated for different types of sales invoices are Presenting different distribution characteristics, the classification decision unit can judge whether the invoice belongs to different categories such as normal invoice, large-value sales invoice, etc. based on the writing feature value, thereby realizing the effective classification and processing of sales invoice data.

[0049] Furthermore, the process processing module includes at least two sub-processing units, each of which corresponds to at least one classified data; the sub-processing unit processes the corresponding classified data based on preset processing rules, and the preset processing rules include data calculation and data association analysis; when performing data calculation, the sub-processing unit calculates based on the following calculation formula: . Among them, R represents the calculation result, Represented as the kth classification data item, It is represented as the weight corresponding to the k-th classification data item, and p is represented as the total number of classification data items; when the sub-processing unit performs data association analysis, it mines data association rules based on the Apriori algorithm.

[0050] In this embodiment, the process processing module will be described in detail. The process processing module includes at least two sub-processing units, each of which corresponds to a type of classification data. The sub-processing unit processes the classification data of the corresponding category according to the preset processing rules. The preset processing rules include data calculation and data association analysis. When performing data calculation, the preset processing rules are based on the calculation formula. Perform data calculation. R represents the calculation result. Represented as the kth classification data item, The weight corresponding to the kth classification data item is represented by p, and the total number of classification data items is represented by p. When performing association analysis, the sub-processing unit mines data association rules based on the Apriori algorithm. The specific process is as follows: the minimum support threshold min-sup and the minimum confidence threshold min-conf are set, and all frequent items that meet the minimum support are found by scanning the transactions in the data set; for each frequent item, its confidence is calculated. If the confidence is greater than or equal to the minimum confidence threshold min-conf, an association rule is generated. The confidence calculation formula is: .in, represents the association rule from the number of data items X to the number of data items Y, Indicates the support of the simultaneous occurrence of data item sets X and Y, Expresses the support of the occurrence of data item X. For example, in corporate accounting data processing, data item X may represent "purchase of office supplies with an amount greater than 5,000" and data item set Y may represent "requires dual approval by the department head and the financial director." Using the Apriori algorithm, all purchase data transactions are scanned. When frequent items X and Y that exceed the minimum support threshold min-sup (e.g., the probability of such occurrence in past purchase data exceeds 10%) are found, the data item is considered as a support. ,When calculating its confidence, if it is greater than the minimum confidence threshold min-conf (such as 80%), then generate the association rule "When purchasing office supplies and the amount is greater than 5,000 yuan, dual approval by the person in charge and the financial director is required."

[0051] Furthermore, the report generation module includes a report database and a data filling unit; wherein the report database has several accounting report templates preset, the data filling unit selects the corresponding accounting report template according to the processing results obtained by the process processing module, and fills the processing results into the selected accounting report template to generate an accounting report.

[0052] In this embodiment, the report generation module is described in detail. The report generation module includes a report database and a data filling unit. The report database stores a variety of preset accounting report templates, including but not limited to standard templates that comply with generally accepted accounting principles, such as the balance sheet, income statement, and cash flow statement. It also includes enterprise-defined special analysis report templates, including but not limited to department details and project profit analysis tables. The data filling unit is used to fill the processing results obtained by the process processing module into the corresponding accounting report template to generate a complete accounting report.

[0053] It should be noted that when the data filling unit fills the corresponding accounting report template according to the processing results, how to select the corresponding accounting report template is described in detail, including but not limited to judging by time attributes, business attributes, and data usage.

[0054] The time attribute is determined by reading the timestamp or accounting period identifier in the processing result. For example, if the time range of the processing result is October 1, 2024, to October 31, 2024, the monthly report template is selected to fill in and generate the annual accounting report. If the time range is January 1, 2024, to December 31, 2024, the annual report template is selected to fill in and generate the annual accounting report.

[0055] The time attribute is judged as follows: when the intelligent classification module obtains the classified data, a category label is entered for each classified data, such as "sales revenue data", "cost accounting data", and "asset change data". When the process processing module processes each classified data in a process-based manner, the processing result obtained carries a category label, and the data filling unit selects the corresponding accounting report template according to the content of the category label. For example, when the category label in a processing result is "accounts receivable aging analysis", the data filling unit selects the accounts receivable special analysis template for filling.

[0056] The data usage determination involves using a process-based processing module to determine the usage of each categorized data entry (e.g., "tax declaration," "internal management," or "external audit") based on the category label entered, and then selecting the corresponding accounting report template. For example, if the usage of a categorized data entry is determined to be "tax declaration," the "Annual Enterprise Income Tax Return" template will be selected for filling.

[0057] It should be noted that users can customize and edit the generated accounting reports and preset accounting report templates.

[0058] Furthermore, it also includes a storage module connected to the data acquisition module, data preprocessing module, intelligent classification module, process processing module and report generation module respectively; the storage module is used to establish a storage sub-unit whenever the data acquisition module collects original data in accounting business, and the storage sub-unit stores the original data, standard data, classified data, processing results and accounting reports in each accounting business in real time.

[0059] In this embodiment, the system will be described in detail again. The system also includes a storage module, which is connected to the data acquisition module, the data preprocessing module, the intelligent classification module, the process processing module and the report generation module. The storage module is used to establish a storage subunit whenever the data acquisition collects the original data in the accounting business, that is, the storage module stores each type of data in the accounting business separately each time the system performs an accounting business, and integrates and stores the various types of data in multiple accounting businesses, so that when an error is found, it is convenient to directly call the corresponding storage subunit for viewing and data comparison. It should be noted that each storage subunit stores all the data in the accounting business, such as the original data collected by the data acquisition module, the standard data of the data preprocessing module, the classified data of the intelligent classification module, the processing results of the process processing module and the accounting report of the report generation module.

[0060] Furthermore, it also includes a verification module connected to the data acquisition module, data preprocessing module, intelligent classification module, process processing module, report generation module and storage module respectively; the verification module is used to verify the consistency of the data at the time of storage with the data in the original module each time the storage subunit stores. When the data at the time of storage is inconsistent with the data in the original module, the storage subunit re-stores the data; the verification module is also used to reversely infer when the storage subunit stores the processing result to obtain the inverse original data, and compare the inverse original data with the original data to determine the consistency. When the inverse original data is inconsistent with the original data, the storage module re-stores the data of the original module step by step. When the inverse original data is consistent with the original data, the storage subunit stores the judgment result of the verification module.

[0061] In this embodiment, the system will be described in detail again. The system also includes a verification module, which is connected to the data acquisition module, data preprocessing module, intelligent classification module, process processing module, report generation module, and storage module. The verification module is used to verify the data output by the verification module, which is connected to the data acquisition module, data preprocessing module, intelligent classification module, process processing module, and report generation module. Specifically, each time the storage subunit stores data, the verification module verifies the consistency of the stored data. If the data stored by the storage subunit is inconsistent with the data in the original module, the storage subunit will re-store the data.

[0062] More specifically, for example, when the data acquisition module outputs the original data, both the storage subunit and the verification module will receive the output original data. The storage subunit will store it, and the verification module will compare the original data stored in the storage subunit with the original data actually received by the verification module from the data acquisition unit to determine the consistency. When the original data obtained from the storage subunit is inconsistent with the original data received from the data acquisition unit, the storage subunit will re-store the original data collected by the data acquisition unit until the original data obtained by the verification module from the storage subunit is consistent with the original data received from the data acquisition unit, and then the next step will be carried out.

[0063] For another example, when the original data obtained by the verification module from the storage subunit is consistent with the original data received from the data acquisition unit, after the data preprocessing unit preprocesses the original data to obtain standard data, both the verification module and the storage subunit will receive this standard data, and the storage subunit will store the standard data. The verification module will obtain the stored standard data from the storage subunit and compare it with the standard data received from the data preprocessing module to determine the consistency. When the standard data obtained from the storage subunit is inconsistent with the standard data received from the data preprocessing module, the storage subunit will re-store the standard data of the data preprocessing module until the standard data obtained by the verification module from the storage subunit is consistent with the original data received from the data preprocessing module, and then proceed to the next step.

[0064] It should be noted that the storage subunit stores the outputs of the intelligent classification module, the process processing module, and the report generation module in the same manner as above, and will not be described in detail here.

[0065] It should be noted that this embodiment verifies the output of each module when it is stored through the verification module, so that when the stored data is subsequently checked, it is guaranteed that each data is accurate data at the time of output.

[0066] In this embodiment, the verification module is also used to reversely obtain the inverse original data based on the processing result when the storage subunit stores the processing result output by the process processing module, and compare the inverse original data with the original data collected by the data acquisition module to determine the consistency. When the inverse original data is inconsistent with the original data, the storage module re-stores the output of the data acquisition module, the data preprocessing module, the intelligent classification module, and the process processing module until the inverse original data is consistent with the original data. The storage subunit will store the processing result. In this embodiment, the consistency of the data in the entire processing process is judged again by reversely deducing to the original data, and the accuracy of the data stored in the storage subunit is again guaranteed. It should also be noted that when the verification module performs reverse deduction, reverse calculation is performed according to the processing method in the module provided in the aforementioned embodiment, thereby completing the reverse deduction.

[0067] This embodiment also provides a processing method based on the accounting data automation process processing system provided in the above embodiment, including the following steps: collecting various types of original data in accounting business; pre-processing the original data to obtain standard data; classifying the standard data based on the classification model to obtain at least two classified data; performing process processing on each classified data to obtain processing results respectively; and generating corresponding accounting reports based on the processing results.

[0068] See Figure 2 As shown, this embodiment will describe the processing method in detail, which specifically includes the following steps: Step S100: Collect various raw data from accounting operations. Collection methods include, but are not limited to, file import, database integration, web crawler, and API interface collection. Collected raw data includes, but is not limited to, financial accounting data and business operations data. Financial accounting data includes, but is not limited to, revenue, costs, and asset responsibilities. Business operations data includes, but is not limited to, procurement, sales, and manufacturing. Step S200: Obtaining standard data based on the raw data obtained in step S100. In this step, the raw data obtained in step S100 is preprocessed to obtain standard data. Preprocessing methods include, but are not limited to, data cleaning, format conversion, data standardization, and data integration. Step S300: Based on the standard data obtained in step S200, the standard data is classified based on the classification model to obtain at least two classified data, and the classification is performed based on the feature information of the standard data. Step S400: Based on the categorized data obtained in step S300, process each categorized data item to obtain a processing result. Specifically, multiple sub-processing units are provided, each of which is responsible for processing a specific type of categorized data, thereby improving the pertinence and professionalism of the categorized data processing. Each sub-processing unit is configured with processing rules to process the corresponding categorized data item and obtain a processing result. Step S500: Generate an accounting report based on the processing results obtained in step S400.

[0069] It should be noted that the embodiments of the present invention have better practicability and do not impose any form of limitation on the present invention. Any technician familiar with the field may use the technical content disclosed above to change or modify it into an equivalent effective embodiment. However, any modification or equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. An accounting data automation process processing system, characterized in that: It includes data acquisition module, data preprocessing module, intelligent classification module, process processing module and report generation module; among them, The data acquisition module is used to collect raw data in accounting business; The data preprocessing module is connected to the data acquisition module, and the data preprocessing module is used to preprocess the raw data to obtain standard data; The intelligent classification module is connected to the data preprocessing module, and is used to classify the standard data based on a preset classification model to obtain at least two classification data; The process processing module is connected to the intelligent classification module, and is used to process each of the classified data in accordance with a preset processing flow to obtain processing results respectively; The report generation module is connected to the process processing module, and the report generation module is used to generate corresponding accounting reports according to each processing result.

2. The accounting data automation process processing system according to claim 1, characterized in that: The intelligent classification module includes a data feature extraction unit and a classification decision unit; wherein, The data feature extraction unit is used to extract feature information from the standard data; The classification decision unit is configured to make a classification decision based on the preset classification model according to the feature information extracted by the data feature extraction unit to obtain the classification data; The classification model is constructed using a deep learning algorithm, and the calculation formula for the classification decision is: in, Indicated as data X belongs to category The probability of Represented as category The probability of data X appearing in , Represented as category The prior probability of occurrence, Expressed as the total number of categories, Represents the first categories, is the index variable, .

3. The accounting data automation process processing system according to claim 2, characterized in that: The data feature extraction unit extracts feature information from the standard data based on a convolutional neural network; The convolutional layer of the convolutional neural network is preset with a calculation formula, which is: in, Represented as the mth feature map of the lth layer, Represented as a set of input feature maps connected to the mth feature map, Expressed as The h-th feature map of the layer, It is represented as the convolution kernel of the lth layer connecting the hth input feature map and the mth input feature map. Expressed as the bias of the mth feature map, " ” represents a convolution operation.

4. The accounting data automation process processing system according to claim 1, characterized in that: The process-based processing module includes at least two sub-processing units, each of which corresponds to at least one classified data; The sub-processing unit processes the corresponding classified data based on preset processing rules, wherein the preset processing rules include data calculation and data association analysis; When performing data calculation, the sub-processing unit calculates based on the following formula: Among them, R represents the calculation result, Represented as the kth classification data item, It is represented as the weight corresponding to the k-th classification data item, and p is represented as the total number of classification data items; When performing data association analysis, the sub-processing unit mines data association rules based on the Apriori algorithm.

5. The accounting data automation process processing system according to claim 1, characterized in that: The report generation module includes a report database and a data filling unit; Several accounting report templates are preset in the report database. The data filling unit selects the corresponding accounting report template according to the processing result obtained by the process processing module, and fills the processing result into the selected accounting report template to generate the accounting report.

6. The accounting data automation process processing system according to claim 5, characterized in that: It also includes a storage module connected to the data acquisition module, the data pre-processing module, the intelligent classification module, the process processing module and the report generation module respectively; The storage module is used to establish a storage subunit whenever the data acquisition module collects the original data in the accounting business. The storage subunit stores the original data, the standard data, the classified data, the processing results and the accounting reports in each accounting business in real time.

7. The accounting data automation process processing system according to claim 6, characterized in that: It also includes a verification module connected to the data acquisition module, the data pre-processing module, the intelligent classification module, the process processing module, the report generation module and the storage module respectively; The verification module is used to verify the consistency of the data stored and the data in the original module each time the storage subunit stores data. When the data stored is inconsistent with the data in the original module, the storage subunit stores the data again. The verification module is also used to reversely obtain the inverse original data when the storage sub-unit stores the processing result, and compare the inverse original data with the original data to determine the consistency. When the inverse original data is inconsistent with the original data, the storage module re-stores the data of the original module step by step. When the inverse original data is consistent with the original data, the storage sub-unit stores the judgment result of the verification module.

8. A processing method based on the accounting data automation process processing system according to any one of claims 1 to 7, characterized in that: The following steps are involved: Collect various raw data in accounting business; Preprocessing the raw data to obtain standard data; Classify the standard data based on a classification model to obtain at least two classification data; Performing process processing on each of the classified data to obtain processing results respectively; Generate corresponding accounting reports based on the processing results.

Images