TWM575887U - Intelligent accounting system - Google Patents

Abstract

An intelligent accounting system, the intelligent accounting system communication is connected to a client electronic device, the client electronic device stores the voucher image data, and the intelligent accounting system includes a voucher memory module and a voucher management module. The voucher memory module is configured to store voucher image data transmitted by the client electronic device, and the voucher management module captures voucher image data in the voucher memory module, and the voucher management module includes an image processing unit and a text recognition unit. And semantic analysis unit. The image processing unit performs image processing on the voucher image data, and the text recognition unit converts the voucher image data into voucher text data, and the semantic analysis unit receives the voucher text data transmitted from the text recognition unit, and analyzes the voucher text data to The voucher text data is classified into accounting items to generate a billing data.

Description

Intelligent accounting system

The novel provides an intelligent accounting system, in particular to an intelligent accounting system with automated processing of accounting data.

Today's SMEs usually do not set up a separate accounting department, so every time at the end of the month, the invoices on the stack and the details of the accounts on the account are definitely a headache for SME managers. In addition, the accounting process is cumbersome and can't be mistaken. Most SME managers have limited accountant resources, so they appoint accountkeepers outside the company to help manage their accounts. However, it is a lot of time to record and classify various documents (such as invoices, electronic invoices, import orders, receipts, accounting documents, etc.), resulting in ineffective timeliness. And manual processing and classification are also prone to errors or omissions in accounting. Therefore, how to more effectively and accurately manage the accounting of the enterprise and reduce the mistake of human error is the value of the general knowledge in the field.

The purpose of the novel is to provide an intelligent accounting system that can efficiently manage the intelligent accounting of the enterprise and reduce the errors of human error. The present invention provides an intelligent accounting system, which is connected to a client electronic device, the client electronic device stores at least one voucher image data, and the voucher image data corresponds to an accounting document. The intelligent accounting system includes a voucher memory module and a voucher management module. The voucher memory module is configured to store the voucher image data transmitted by the client electronic device, and the voucher management module retrieves the voucher image data in the voucher memory module. The voucher management module comprises: an image processing unit, a character recognition unit, a word library, a word vector corpus, a semantic analysis unit, and a correction automatic unit, wherein the thesaurus includes a department vocabulary, a product noun library, and a transaction object vocabulary. The text recognition unit receives the voucher image data and converts the voucher image data into voucher text data. The vocabulary in the thesaurus records the correspondence between various product names and accounting items. The word vector corpus includes a plurality of word vectors, which correspond to various product names classified into the thesaurus. The semantic analysis unit receives the voucher text data transmitted from the text recognition unit, the voucher text data includes a product name and a transaction object data, and the semantic analysis unit performs the following steps: (a1) in the product name and the transaction object word library word The library searches for the product name in the voucher text data, and determines whether the product name included in the voucher text data exists in the vocabulary. If yes, the following step (a2) is performed; if not, the following steps are performed ( A3). (a2) Find the accounting item corresponding to the product name. (a3) finding a word vector closest to the product name in the word vector corpus, aligning the word vector into the thesaurus, and finding an accounting item having the closest vector distance to the word vector, and Add the product name to the thesaurus. (a4) generating a billing data according to step (a2) or step (a3), the billing data including an accounting item corresponding to the product name and the product name included in the voucher text data. The above described features and advantages will be more apparent from the following description.

Please refer to FIG. 1. FIG. 1 illustrates the relationship between the intelligent accounting system and the client electronic device of the embodiment. The client electronic device 10 includes a lens 11 and a client operating program 12, and the client operating program 12 is capable of communicating with the intelligent accounting system 100 for data exchange. Please refer to FIG. 2 . FIG. 2 is a schematic diagram of photographing the accounting document 81 by using the lens 11 of the client electronic device. The image of the accounting document 81 is stored as the voucher image data 13 by taking a picture of the accounting document 81 by the lens 11 . In this embodiment, the client electronic device 10 is a smart phone, but the invention is not limited thereto. The client electronic device 10 can also be a desktop computer or a notebook computer, and the accounting document 81 can be used by the scanner. And converted into voucher image data 13. In this creation, accounting vouchers refer to written evidence of economic activities and clear economic responsibilities. Accounting vouchers are an important basis for registering books and accounting supervision, such as: unified invoices, taxable vouchers, electronic invoices, and various receipts. Referring to FIG. 1 , the smart accounting system 100 includes a voucher memory module 110 and a voucher management module 120 . The smart accounting system 100 is connected to a client electronic device 10, so the smart accounting system 100 can store the voucher image data 13 sent by the client electronic device 10 in the voucher memory module 110. The voucher management module 120 captures the voucher image data in the voucher memory module 110. The voucher management module 120 includes a voucher classification unit 122, a voucher feature pattern classification library 122a, an image processing unit 124, and a text recognition unit 126. A account processing unit 127, a semantic analysis unit 128, a term library 128a, a word vector corpus 128b, and a correction machine learning unit 129. The vocabulary 128a includes a department vocabulary, a product noun library, and a transaction target vocabulary, and the product noun library records the correspondence between various product names and accounting items, and the transaction object vocabulary in the vocabulary 128a also records the transaction object (ie, The name of the customer and the supplier, and the departmental vocabulary in the vocabulary 128a records the company's department name. Moreover, the word vector corpus 128b includes a plurality of word vectors that correspond to various product names in the thesaurus 128a. Word vector (also known as word embedding or representation) is a technique widely used in recent years to use a vector to represent each vector representation. Please refer to FIG. 1 and FIG. 3 at the same time. FIG. 3 is a flowchart of a method for identifying accounting documents according to the embodiment. First, in step S110, the smart accounting system 100 reads the voucher image data 13, which is transmitted from the client electronic device 10, for example, in the present embodiment. Thereafter, in step S120, the image processing unit 124 preprocesses the voucher image data 13 to improve the sharpness of the voucher image data 13. In this embodiment, the pre-processing refers to image denoising or binarization processing on the voucher image data 13 to improve the sharpness of the voucher image data 13. In addition, the image processing unit 124 also measures the appearance of the voucher image data 13 to determine the size of the accounting document 81. Moreover, the image processing unit 124 can also perform feature point detection on the voucher image data 13 to find a plurality of feature points (not shown). Then, in step S130, the voucher classification unit 122 classifies the voucher image data 13 by the voucher feature pattern classification library 122a. Since the voucher feature style classification library 122a stores a plurality of voucher styles, the voucher classification unit 122 can classify the voucher image data 13, for example, classifying the voucher image data 13 into water fee, electricity fee, gas fee by the size of the accounting document 81. Wait. In this embodiment, the voucher classification unit 122 classifies the voucher image data 13 by Convolutional Neural Networks, and can improve the classification efficiency of the voucher image data 13. Again, please refer to FIG. 4 at the same time, and FIG. 4 is a schematic diagram of accounting documents. In step S132, the document image data 13 is segmented into a plurality of fields 131a-131f by the feature points detected by the image processing unit 124. Then, step S134 is implemented to label each field of the voucher image data 13, for example, the field 131a is marked as the open receipt unit 131a, the field 131b is marked as the transaction object, the field 131c is marked as the product name, and the field 131d is marked. For the amount, field 131e is marked as the opening time, and field 131f is marked as the receipt number. The reason why the labels 131a to 131f are tagged is that the semantic analysis unit 128 does not make an error when analyzing the data, and will be described in more detail below. Then, in step S140, the character recognition unit 126 converts the image processed document image data 13 into a voucher text data 13'. In this embodiment, the character recognition unit 126 converts the text image in the voucher image data 13 into real text data by using the OCR character recognition technology, that is, the voucher text data 13', that is, the text can be written by the text editing software. Edited information. The voucher text material 13' includes at least one product name, that is, a product name. In addition, since the image processing unit 124 has already tagged the fields of the voucher image data 13, the character recognition unit 126 can also associate the voucher text data 13' with the tags of the respective fields. For example, the text "○○○寿司名店有限公司" corresponds to the label "transaction object" of the field 131b, and the text of "NT$63,000" corresponds to the label "amount" of the field 131d. In addition, as mentioned in the previous paragraph, the reason for labeling each of the fields 131a to 131f is that the semantic analysis unit 128 does not make an error when analyzing the data. As can be seen from FIG. 4, since the field 131b corresponding to the "transaction object" may have a name similar to the product name (sushi in the present embodiment), if the field 131b is tagged in advance, the semantic analysis unit 128 You can't make a mistake and classify the voucher into an incorrect accounting item. Next, in step S150, the semantic analysis unit 128 searches in the vocabulary 128a with the product name in the voucher text data 13' (for example, "Billing and Financial Statement Production" shown in the field 131c in FIG. 4), and in the step. In S160, it is determined whether or not the article name corresponding to the voucher file 13' exists in the vocabulary 128a. If yes, the following step S170 is performed, and if not, the following step S180 is performed. In step S170, the semantic analysis unit 128 finds the accounting item corresponding to the product name in the voucher file 13', for example, "Billing and Financial Statement Production" corresponds to the accounting item "Labor Fee" (No. 2071). It should be noted that when the semantic analysis unit 128 finds the accounting item corresponding to the product name in the voucher text data 13', in addition to referring to the product noun library, it also refers to the departmental thesaurus and the transaction object lexicon; The department reporting the accounting document 8 and the transaction object on the accounting document 8 also affect the classification of the accounting item. In step S180, the semantic analysis unit 128 finds a word vector closest to the product name in the voucher text material 13' in the word vector corpus 128b, compares the word vector in the thesaurus 128a, and finds out The vector distance of the word vector is the accounting item corresponding to the product name corresponding to the closest word vector. Thereafter, in step S190, the semantic analysis unit 128 generates a ledger data 23 (shown in Fig. 5), and the ledger data 23 is an accounting item corresponding to the product name in the voucher text data 13'. Please also refer to Figure 5, which is a breakdown of the details of the accounting data. In other embodiments, after the plurality of billing materials 23 are collated by the account processing unit 127, a further billing information such as a voucher 2331, a day bill 2332, a report 2333, and a business tax 2334 is generated. Therefore, compared with the use of labor to deal with the internal accounting of the enterprise, the novel intelligent accounting system 100 can more efficiently calculate the business tax required to be paid, greatly reducing the time cost. After the step S190 is completed, the step S195 is executed, and the correcting machine learning unit 129 receives the account information 23 sent from the semantic analysis unit 128, and an audit data page (not shown) can be generated on the screen of the terminal control unit 20. The operator of the terminal control unit 20 checks the account information 23. If the correcting machine learning unit 129 determines that the correspondence between the product name and the accounting item in the document text 13' is incorrect, the item name and the accounting item in the accounting data 23 are corrected. Correspondence relationship, and the corresponding relationship between the product name and the revised product name and the accounting item is added to the thesaurus 128a. That is to say, if the corresponding product name is not found in the vocabulary 128a by using the word vector in step S180, and the classified accounting item can not be found, the auditing information page is used to correct the vocabulary 128a through machine learning. The product name and the word vector corresponding to the word vector corpus 128b. In this way, if the accounting documents containing the product name are processed again, the accounting items to which they belong can be immediately determined and classified. Therefore, the semantic analysis unit 128 can also increase the accuracy of the accounting item classification after the machine learning of the feedback as shown in step S195. In the present embodiment, the correcting machine learning unit 129 is communicatively coupled to the terminal control unit 20, and the operator of the terminal control unit 20 is, for example, a financial person in the organization, and the financial person checks the account by the terminal control unit 20. Is the information 23 correct? Here, the vocabulary 128a is further explained. Since the vocabulary 128a stores the correspondence between the plurality of product names and the respective accounting items, the semantic analysis unit 128 can sort the vouchers and texts 13' into accounting materials. For example, when the product name in the voucher file 13' includes the "pencil" keyword, if the vocabulary 128a records "pencil" and its corresponding accounting item, the semantic analysis unit 128 can find the accountant in the vocabulary 128a. The corresponding item name is "Stationary Supplies" and the item number is "5153". Alternatively, when the product name in the voucher text data 13' includes the keywords "water fee" or "electricity fee", if the vocabulary 128a records "water fee" or "electricity fee" and the corresponding accounting item, the semantic analysis unit 128 The corresponding name is found in the vocabulary 128a as "hydroelectric gas fee" and the item number is "5161". Further, the vocabulary 128a may also record the name of the transaction object, the unified number, or other related information, so that the accounting data 23 generated by the semantic analysis unit 128 may contain the data of the transaction object. Please refer to FIG. 1 and FIG. 6 simultaneously. In addition to allowing the client electronic device 10 to read general accounting documents, the client operating program 12 can also cause the client electronic device 10 to read an electronic invoice 83 to generate a voucher text. 13'. In detail, the electronic invoice 83 itself has a two-dimensional barcode (QR code) 830, and the basic information of the document text 13' is stored in the two-dimensional barcode 830, so when the lens 11 is aligned with the two-dimensional barcode 830 of the electronic invoice 83 At the same time, the client electronic device 10 can also read the voucher text material 13' from the two-dimensional bar code 830 and transfer it to the account processing unit 127 of the smart accounting system 100. In this way, the account processing unit 127 can perform subsequent processing (generating the voucher 2331, the journal 2332, the report 2333, and the sales tax 2334). In summary, the intelligent accounting system of the present invention can convert the image data of the accounting documents into text data, and automatically classify the accounting items, thereby making various accounting materials. In addition, the intelligent accounting system integrates reinforcement learning to improve its recognition rate. Although the present invention has been disclosed in the above preferred embodiments, it is not intended to limit the present invention, and any one of ordinary skill in the art can make some modifications and refinements without departing from the spirit and scope of the present invention. Therefore, the scope of protection of this new type is subject to the definition of the scope of the patent application.

10‧‧‧Client electronic device

11‧‧‧ lens

12‧‧‧Client Operating Program

81‧‧‧Accounting documents

100‧‧‧Intelligent Accounting System

110‧‧‧Voucher Memory Module

13‧‧‧Voucher image data

13’‧‧‧Voucher text

120‧‧‧Voucher Management Module

122‧‧‧Voucher classification unit

122a‧‧‧Voucher Feature Style Classification Library

124‧‧‧Image Processing Unit

126‧‧‧Text recognition unit

127‧‧ ‧ account processing unit

128a‧‧ vocabulary

128b‧‧‧word vector corpus

129‧‧‧Correct machine learning unit

23‧‧‧Accounting Information

S110~S195‧‧‧Step flow chart

131a~131f‧‧‧ fields

20‧‧‧ Terminal Control Unit

2331‧‧‧Subpoena

2332‧‧‧ journal

2333‧‧‧Report

2334‧‧‧Business tax

83‧‧‧Electronic invoice

830‧‧‧2D barcode ­­

FIG. 1 is a diagram showing the relationship between the smart accounting system and the client electronic device of the embodiment. FIG. 2 is a schematic diagram of photographing accounting documents with a lens of a client electronic device. FIG. 3 is a flow chart showing a method for identifying an accounting document according to the embodiment. Figure 4 is a schematic diagram of an accounting document. Figure 5 is a breakdown of the breakdown of the accounting data. FIG. 6 is a schematic diagram of photographing an electronic invoice with a lens of a client electronic device.

Claims (6)

An intelligent accounting system, the communication is connected to a client electronic device, the client electronic device stores at least one voucher image data, and the voucher image data corresponds to an accounting document, the smart accounting system comprises: a voucher memory a module for storing the voucher image data transmitted by the client electronic device; a voucher management module for capturing the voucher image data in the voucher memory module, the voucher management module comprising: an image Processing unit, preprocessing the voucher image data to improve the clarity of the voucher image data, and measuring the appearance of the voucher image data to determine the size of the accounting document; a text recognition unit receiving the voucher image data Converting the voucher image data into a voucher text data; a term library, recording the correspondence between various product names and accounting items; a word vector corpus, including a plurality of word vectors, the word vectors corresponding to various product names in the thesaurus And a semantic analysis unit that receives from The voucher text data sent by the word recognition unit, the voucher text data includes a product name and a transaction object data, and the semantic analysis unit performs the following steps: (a1) in the thesaurus, the product name in the voucher text data Performing a search and determining whether the product name included in the voucher text data exists in the thesaurus, if yes, performing the following step (a2), if not, performing the following steps (a3); (a2) finding the product name Corresponding accounting item; (a3) find the word vector closest to the product name in the word vector corpus, compare the word vector to the thesaurus, and find the closest vector distance to the word vector a product name and an accounting item corresponding to the product name, and adding the product name to the vocabulary; and (a4) generating a accounting information according to step (a2) or step (a3), the accounting information including the vouchers The accounting item included in the product name and the product name. The intelligent accounting system according to claim 1, wherein the step of preprocessing the voucher image data comprises: performing image denoising on the voucher image data. The intelligent accounting accounting system of claim 1, wherein the step of preprocessing the voucher image data comprises: performing grayscale binarization processing and correction on the voucher image data. The intelligent accounting system as described in claim 1 further includes: a voucher feature style classification library, storing a plurality of voucher styles; a voucher classification unit connected to the voucher feature style classification library, the voucher classification unit The voucher image data is classified according to the voucher feature pattern classification library. For example, in the intelligent accounting system described in claim 4, the voucher classification unit further performs the following steps: (b1) performing feature point detection on the voucher image data by the image processing unit to find out a plurality of feature points; (b2) dividing the voucher image data into a plurality of fields by the detected feature points; and (b3) tagging each field of the voucher image data; wherein the text The identification unit corresponds the voucher text data to the tags of each field to further subdivide the voucher text data. The intelligent accounting system as claimed in claim 1, further comprising a correction machine learning unit, the correction machine learning unit receiving the account information transmitted from the semantic analysis unit, and checking the account information If the correcting machine learning unit determines that the correspondence between the product name and the accounting item included in the voucher text data is incorrect, correcting the correspondence between the product name and the accounting item in the accounting data, and the product name and the product name and accounting The corresponding relationship of the project is added to the thesaurus.