Conversational bookkeeping method and electronic equipment
Through the dialogue-based accounting method, multi-modal interaction and accounting assistant functions are used to solve the problems of low interaction efficiency and lack of a systematic accounting framework in the existing technology, and an efficient and convenient accounting experience is achieved.
Patent Information
- Application Number
- CN202510173531.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-18
- Publication Date
- 2025-06-06
AI Technical Summary
The existing electronic accounting applications have low interaction efficiency, lack a systematic accounting framework, limited automated accounting methods, and insufficient functions to support users' immediate suggestions and data discovery during the accounting process.
It provides a dialogue-based accounting method, which simplifies user operations through multi-modal interaction (text, voice, images, etc.), supports custom accounting lists, uses accounting assistant functions and search enhancement generation technology to generate accounting entries that meet the requirements of double accounting.
It improves the efficiency of user input, simplifies the accounting process, helps users establish a systematic double accounting thinking framework, and improves accounting efficiency and user experience.
Smart Images

Figure CN120104016A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of computers, and in particular to a conversational accounting method, an accounting device, an electronic device, a client-server system, a computer-readable storage medium, and a computer program product. Background Art
[0002] With the miniaturization and intelligence of electronic devices, such as the development of smart phones, more and more people use electronic accounting applications to record daily income and expenditure and provide relevant basis for their financial planning. The accounting applications commonly seen on electronic devices in the market currently usually provide an interactive method based on a graphical user interface (GUI), mainly completing accounting tasks by users clicking icons, selecting categories, entering amounts, etc. However, these electronic accounting methods have the following problems:
[0003] First, low interaction efficiency: users need to manually perform multiple steps to complete a simple accounting task;
[0004] Second, there is no systematic framework for bookkeeping: In order to lower the threshold, common bookkeeping software focuses on the running account of income and expenditure, but in order to be relatively complete, it also includes many categories, such as transfers, balances, debits and credits, etc. Although this reduces the difficulty for ordinary users to start bookkeeping, because there is no double-entry bookkeeping framework, it is difficult for ordinary users to figure out the relationship between these categories. Even if they keep books for a long time, it is difficult to establish a complete bookkeeping thinking framework.
[0005] Third, the limitations of existing automated accounting methods: Some applications can automatically identify transaction information, or import transaction accounts, automatically classify and summarize, but they do not take into account that the classification is determined by both objective factors and user subjective factors. This seems fast, but these automated methods are not based on a suitable chart of accounts. In fact, it is difficult for users to control the granularity of accounting. Accounting that is too detailed often makes users feel that they have not gained much after spending a lot of time, and then give up accounting.
[0006] Fourth, insufficient functions: lack of immediate suggestions and help for problems encountered by users in the accounting process, lack of mining of accumulated data to help users continuously improve their accounting capabilities.
[0007] Existing tools are either similar to corporate accounting software that are too professional for ordinary users to use, or, as mentioned above, are not systematic enough to meet the gradually in-depth needs of users, making it difficult for users to persist in accounting for a long time and achieve effective financial management. Summary of the invention
[0008] The present application provides a conversational bookkeeping method, bookkeeping device, electronic device, client-server system, computer-readable storage medium and computer program product, which are intended to help users quickly record transactions through chat. Users can define the chart of accounts independently to control the granularity of bookkeeping. Through the bookkeeping assistant function, users can get timely feedback on problems encountered during the bookkeeping process, so that a systematic double-entry bookkeeping thinking framework can be gradually established.
[0009] Furthermore, as users use it more deeply, they can continuously increase their knowledge in accounting and finance and better manage their own finances. It aims to improve the accounting efficiency and user experience of individuals, families and small financial scenarios through the use of innovative human-computer interaction technology and artificial intelligence technology.
[0010] The present invention provides an accounting method based on a conversational user interface, aiming to solve the above problems in the prior art. By supporting multimodal interaction, the accounting method of the present invention simplifies user operations, especially improves the efficiency of user input, and improves interaction efficiency and user experience.
[0011] In order to achieve the above-mentioned purpose, the embodiment of the present application adopts the following technical solution:
[0012] In the first aspect, a conversational accounting method is provided, which is applied to an electronic device, and the method includes: the electronic device displays a conversational user interface including a display area and an input area, and the conversational user interface supports multimodal interaction; receives transaction information input by a user through the input area of the conversational user interface; the electronic device generates a corresponding transaction text by analyzing the transaction information, and automatically generates accounting entries based on the transaction text; stores the transaction text and the corresponding generated accounting entries in an accounting book, and displays the execution results to the user in the display area; and supports the modification and deletion of accounting entries, as well as the setting of the chart of accounts, the accounting assistant, and the accounting report query function. This accounting method is implemented through a flexible conversational user interface, so that users can quickly input and manage financial data, meet the user's requirements for the immediacy of accounting, and improve the user's accounting efficiency.
[0013] In a possible implementation, the input area of the conversational user interface supports input types including text, voice, images, and other electronic accounting documents, which can be either single-type input or a combination of the above types, and has multi-modal input parsing capabilities. Among them, images include but are not limited to invoices, receipts, and orders; other electronic accounting documents include but are not limited to electronic invoices, bank statements, QR codes, and barcodes.
[0014] In a possible implementation, the display area of the conversational user interface supports display content including, but not limited to, single-type display or multi-type combined display of text, voice, image, and control. The display area or the entire conversational user interface can be switched to web page display according to user input to load or embed web page content, and support interactive operations of web page content.
[0015] Among them, the display area of the dialogue user interface or the entire dialogue user interface can be based on user input, and such user input includes but is not limited to input information in the input area of the dialogue user interface, user touching the controls, and user clicking, scrolling and filling out forms on web page content.
[0016] In one possible implementation, when the electronic device receives transaction information input by the user from the input area of the conversational user interface, whether the transaction information is a single type or a multi-modal combination input, it will be converted into transaction text through parsing. Among them, when the transaction information is text, it is directly used as the transaction text; when the transaction information is voice information, the voice content is analyzed by automatic speech recognition (ASR) technology, and the extracted original text is processed, including but not limited to normalization, removal of irrelevant characters and spelling correction, etc., to obtain the preprocessed transaction text; when the transaction information is image information, the image content is analyzed by optical character recognition (OCR) technology, and the extracted original text is processed, including but not limited to removing invalid characters and advertisements and other information irrelevant to the transaction, to obtain the preprocessed transaction text; when the transaction information is other electronic accounting vouchers, the parsing method suitable for the type of electronic accounting vouchers is adopted, including but not limited to file parsing, QR code parsing and barcode parsing, and the extracted original text is processed, including but not limited to removing noise in the text, such as extra spaces, special characters and format errors, to obtain the preprocessed transaction text; when the transaction information is a combination of multiple single types mentioned above, after being parsed and preprocessed separately, they are directly spliced and merged to generate the transaction text.
[0017] In a possible implementation, the electronic device generates accounting entries using the retrieval enhancement generation technology. The process includes: performing similarity retrieval through a vector database, retrieving information related to the input description from a knowledge base based on the transaction text generated above; inputting the retrieval results, the transaction text, the chart of accounts, and the generation requirements for generating accounting entries that meet the double-entry bookkeeping requirements into the generation model as context to generate accounting entries that meet the user's needs.
[0018] The knowledge base includes the accounting book, the chart of accounts, accounting-related knowledge, question-and-answer records, and a vector database. The vector database stores embedded vectors to support efficient similarity retrieval. The embedded vectors are calculated by a pre-trained embedding model based on the accounting book, the chart of accounts, the accounting-related knowledge, and the question-and-answer record data. The chart of accounts includes all pre-defined accounting accounts and can be reset according to user needs. The accounting-related knowledge includes but is not limited to accounting principles and financial statement preparation texts. The question-and-answer record consists of questions asked by users and help messages from the accounting assistant.
[0019] In one possible implementation, each generated accounting entry is assigned a unique identifier (ID) for easy retrieval and operation. Considering the uniqueness, quantity growth and length optimization of the identifier, Nano ID (a lightweight unique identifier generation scheme) can be used as a unique identifier generation scheme to ensure efficient management and query of accounting entries. The electronic device stores the aforementioned transaction text and its corresponding generated accounting entry with Nano ID added in the accounting book, and displays the accounting entry in the display area.
[0020] In one possible implementation, the electronic device receives input information from an input area of a dialogue user interface. When the input information is text for modifying an existing accounting entry, the original accounting entry saved in the accounting book is modified, and the feedback result of the execution is displayed in the display area. When the input information is text for deleting an existing accounting entry, the original accounting entry saved in the accounting book and its corresponding transaction text are deleted, and the feedback result of the execution is displayed in the display area. When the input information is text or voice for setting a chart of accounts, a page for setting a chart of accounts is opened, and specific accounting accounts in the chart of accounts are added, modified and / or deleted through interaction with the chart of accounts page. When the input information is text or voice for starting an accounting assistant, the accounting assistant page is opened, and the user obtains a help message through a dialogue with the accounting assistant page. When the input information is text or voice for querying financial statements, corresponding financial statements are generated according to the principle of double-entry bookkeeping, and a financial statement page is opened, and the user obtains corresponding financial statement information through interaction with the financial statement query page.
[0021] In one possible implementation, the electronic device receives touch input from a control on a dialogue user interface, and when the input is a touch of a control for setting a chart of accounts, a page for setting a chart of accounts is opened, and specific accounting accounts in the chart of accounts are added, modified and / or deleted through interaction with the chart of accounts page; when the input is a touch of a control for starting an accounting assistant, the accounting assistant page is opened, and the user obtains a help message by interacting with the accounting assistant page; when the input is a touch of a control for querying financial statements, corresponding financial statements are generated according to the principle of double-entry bookkeeping, and a financial statement page is opened, and the user obtains corresponding financial statement information through interaction with the financial statement query page.
[0022] In a possible implementation, after the accounting assistant function is started, the user's input in the input area is converted into a question text. The accounting assistant uses the question text as input, performs similarity search through the vector database, retrieves information related to the input description from the knowledge base, and inputs the search results, the question text, the chart of accounts, and the generation requirements for generating answers as context into the generation model to generate help messages. The help messages of the accounting assistant include but are not limited to: questions raised by users about the accounting method of specific transactions, questions raised by users about adding new accounting accounts or modifying existing accounting accounts, questions raised by users about setting or modifying the chart of accounts, such as merging duplicate or similar accounting accounts, deleting long-unused accounting accounts, and suggesting new accounting accounts based on user transaction habits and emerging consumption patterns.
[0023] In one possible implementation, the help message includes three buttons: Regenerate, Like, and Dislike. The user can click the Regenerate button to generate a new help message, and the user can click the Like and Dislike buttons to evaluate the quality of the generated help message. The question text, the help message, and the evaluation of the help message will all be stored in the question and answer record.
[0024] In a possible implementation, the aforementioned knowledge base can perform an update operation at a predetermined time interval, such as one month, or under specific trigger conditions, such as when 1,000 new accounting entries are added to the accounting book, based on the update of the accounting books, chart of accounts, accounting-related knowledge, and question and answer records. The update operation includes recalculating the embedding vector of the above data and storing it in the vector database to improve the accuracy of the accounting entry generation and the relevant suggestions provided by the accounting assistant. Among them, when embedding the question and answer records, high-quality question and answer record entries are screened out as embedded texts based on the user's evaluation of each record through the control.
[0025] In one possible implementation, the financial report page design is optimized for small-screen electronic devices and large-screen electronic devices: for small-screen electronic devices, the page displays simplified core financial information, including but not limited to expenditure statistics, income statistics and balance sheets, to ensure efficient information presentation within limited screen space; for large-screen electronic devices, the page displays more comprehensive financial information, including not only the basic content on small-screen electronic devices, but also provides a more detailed display of financial data, while supporting but not limited to the functions of querying, modifying, and deleting accounting entries, as well as statistical analysis of the usage of accounting subjects, thereby providing users with a more intuitive, flexible and in-depth financial management experience on large-screen devices.
[0026] In a second aspect, the present application provides an accounting device, the accounting device comprising:
[0027] The input-output module is used to display the conversational user interface of the expenditure multimodal interaction, receive the user's input and display the execution results.
[0028] The processing module is used for processing including: extracting and processing the content of input information, generating, modifying and deleting accounting entries, setting up the chart of accounts, running the accounting assistant, and generating financial statements and providing inquiries.
[0029] The storage module is used to store data including temporarily stored data and data that needs to be stored for a long time, wherein the temporarily stored data includes the intermediate data required for the operation of the accounting device; the long-term stored data includes accounting books, charts of accounts, accounting-related knowledge, question and answer records, and embedded vectors, which are used to support retrieval enhancement generation technology.
[0030] Furthermore, the input-output module includes two submodules, one is an input module and the other is a display module. The input module is used for user input, supports touch control of controls, supports single type or multiple single type combination input, the single type includes text, voice, image and other electronic accounting vouchers, and has multimodal input parsing capabilities. Among them, images include but are not limited to invoices, receipts and orders, and other electronic accounting vouchers include but are not limited to electronic invoices, bank statements, QR codes and barcodes; the display module is used to display the results of the execution, and the supported content includes single type display or multiple types of combined display of text, voice, image and interactive components. The display area or the entire dialogue user interface can be switched to a web page display according to user input to load or embed web page content, and support interactive operations of web page content.
[0031] Furthermore, the processing module includes an extraction module, an instruction generation module, an intelligent model module and an execution module.
[0032] The extraction module is used to perform multimodal analysis on the user input information and extract text from it. Among them, the multimodal analysis includes: when the input information is text, it is directly used as the input text; when the input information is voice information, the voice content is analyzed by automatic voice recognition technology, and the extracted original text is processed, including but not limited to normalization, removal of irrelevant characters and spelling correction, etc., to obtain the preprocessed input text; when the input information is image information, the image content is analyzed by optical character recognition technology, and the extracted original text is processed, including but not limited to removing invalid characters and information such as advertisements that are irrelevant to the transaction, to obtain the preprocessed input text; when the input information is other electronic vouchers for accounting, the parsing method suitable for the type of electronic vouchers for accounting is adopted, including but not limited to file parsing, QR code parsing and barcode parsing, and the extracted original text is processed, including but not limited to removing noise in the text, such as extra spaces, special characters and format errors, to obtain the preprocessed input text; when the input information is a combination of multiple single types mentioned above, after being parsed and preprocessed separately, they are directly spliced and merged to generate a unified input text.
[0033] The instruction generation module has two types of inputs: one is the input text output from the extraction module, and the other is the touch information from the input module. The input text is semantically parsed through natural language processing technology. When it is determined that the input text includes a new transaction, the input text is used as the input of the intelligent model module, and the accounting entries that meet the requirements are generated by the intelligent model. The electronic device will generate an instruction to add the accounting entry and include the accounting entry in the instruction to add the accounting entry; when the input text includes a modified existing transaction, an instruction to modify the accounting entry is generated; when the input text includes deleting a specific existing transaction, an instruction to delete the accounting entry is generated; when the input text includes a request to set the chart of accounts, or when the touch information comes from touching the control for setting the chart of accounts, an instruction to set the chart of accounts is generated; when the input text includes a request to start an accounting assistant, or when the touch information comes from touching the control for starting the accounting assistant, an instruction to start the accounting assistant is generated; when the input text includes a request to query financial statements, or when the touch information comes from touching the control for querying financial statements, an instruction to query financial statements is generated.
[0034] The intelligent model module is used to generate the help messages required for the accounting entries and bookkeeping assistant functions, and utilizes the search enhancement generation, performs similarity search through the vector database, and retrieves information related to the input description from the knowledge base based on the input text. On the one hand, the search results, the input text, the chart of accounts, and the generation requirements for generating accounting entries that meet the double-entry bookkeeping requirements can be combined as context input to the generation model to generate accounting entries that meet user requirements; on the other hand, the search results, the input text, the chart of accounts, and the generation requirements for generating answers can be combined as context input to the generation model to generate help messages. The knowledge base includes the accounting book, the chart of accounts, accounting-related knowledge, question and answer records, and a vector database. The vector database stores embedded vectors to support efficient similarity search. The embedded vector is calculated by a pre-trained embedding model based on the accounting book, the chart of accounts, the accounting-related knowledge, and the question and answer record data.
[0035] The help messages include, but are not limited to, answers to questions raised by users regarding accounting methods for specific transactions, answers to questions raised by users regarding adding new accounting subjects or modifying existing accounting subjects, answers to questions raised by users regarding setting up or modifying the chart of accounts, such as merging duplicate or similar accounting subjects, deleting accounting subjects that have not been used for a long time, and suggesting new accounting subjects based on user transaction habits and emerging consumption patterns.
[0036] The execution module is used to specifically execute the instructions generated by the instruction generation module. When the instruction is an instruction to add an accounting entry, the accounting entry is stored in the accounting book stored in the storage module, and the accounting entry is displayed in the display area through the display module; when the instruction is an instruction to modify an accounting entry, the modified accounting entry is updated to the accounting book stored in the storage module, and a message of the modification completion is displayed in the display area through the display module; when the instruction is an instruction to delete an accounting entry, the original accounting entry to be deleted and its corresponding transaction text are deleted from the accounting book stored in the storage module, and a message of the deletion completion is displayed in the display area through the display module; when the instruction is an instruction to delete an accounting entry, the original accounting entry to be deleted and its corresponding transaction text are deleted from the accounting book stored in the storage module, and a message of the deletion completion is displayed in the display area through the display module; When the command is to set the chart of accounts, the chart of accounts page will be displayed in the display area through the display module, and the user will set the chart of accounts by interacting with the setting page, and the set information will be saved in the storage module; when the command is to start the accounting assistant, the display module will display the corresponding accounting assistant page in the display area, and the user will obtain the help message by interacting with the accounting assistant page; when the command is to query the financial statements, the display module will pop up the financial statement page in the display area, and the user can set the query conditions by interacting with the page to obtain the corresponding financial statements.
[0037] In a third aspect, the present application provides an electronic device comprising a touch screen, one or more processors and one or more memories, wherein a computer program is stored in the memory, and the computer program comprises computer instructions. When the computer program in the memory is executed by the processor, the electronic device is able to implement any possible design method of the first aspect mentioned above.
[0038] In a fourth aspect, the present application provides a client-server system, which includes an electronic device and a remote server, wherein the electronic device and the remote server respectively include one or more processors and one or more memories, wherein the memories store a computer program, and the computer program includes computer instructions. When the computer program is executed on the electronic device and the remote server, the electronic device and the remote server respectively implement at least a part of the possible design method of the first aspect above, and collaboratively implement any possible design method of the first aspect above.
[0039] In a fifth aspect, the present application provides a computer-readable storage medium, comprising a computer program, which, when executed on an electronic device, enables the electronic device to implement any possible design method of the first aspect above; or, when the computer program is collaboratively executed on an electronic device and a remote server, enables the electronic device and the remote server to respectively implement at least a part of the possible design method of the first aspect above, and collaboratively complete any possible design method of the first aspect above.
[0040] In a sixth aspect, the present application provides a computer program product, comprising a computer program, which, when executed on an electronic device, enables the electronic device to implement any possible design method of the first aspect above; or, when the computer program is collaboratively executed on an electronic device and a remote server, enables the electronic device and the remote server to respectively implement at least a part of the possible design method of the first aspect above, and collaboratively complete any possible design method of the first aspect above. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] Figure 1 A schematic diagram of a common accounting scenario on a mobile phone provided in an embodiment of the present application;
[0042] Figure 2 A structural intention of an interconnected scene and an electronic device provided in an embodiment of the present application;
[0043] Figure 3 A software structure block diagram of an electronic device provided in an embodiment of the present application;
[0044] Figure 4A first group of interface schematic diagrams provided for an embodiment of the present application;
[0045] Figure 5 A second set of interface schematic diagrams and order schematic diagrams provided for an embodiment of the present application;
[0046] Figure 6 A third group of interface schematic diagrams provided for an embodiment of the present application;
[0047] Figure 7 A fourth group of interface schematic diagrams provided for an embodiment of the present application;
[0048] Figure 8 A fifth group of interface schematic diagrams provided for an embodiment of the present application;
[0049] Fig. 9 The sixth group of interface schematic diagrams provided in the embodiment of the present application;
[0050] Fig.10 A flowchart of an accounting method provided in an embodiment of the present application;
[0051] Fig.11 A schematic diagram of the structure of an accounting device provided in an embodiment of the present application; DETAILED DESCRIPTION
[0052] In order to more clearly explain the problems to be solved by this embodiment, the technical solutions and their advantages, the technical solutions in the embodiments of this application will be described in detail below in conjunction with the accompanying drawings and specific implementation methods. In the description of the embodiments of this application, the terms used below are only for the purpose of describing specific embodiments and are not intended to limit this application.
[0053] The company names, trademarks, and product names (such as Jingdong, Canon, EOS 600D, DeepSeek, WeChat, China Merchants Bank, etc.) mentioned in this embodiment are only for illustration and example purposes, and are intended to help understand the embodiments of the present invention. The reference to these names does not constitute an endorsement, recommendation, or association of the relevant brands, nor does it mean that the present invention is limited to use with these specific brands or products. The present invention can be implemented independently of any third-party brands, products, or services mentioned above. The ownership of any third-party trademarks or brand names belongs to their respective owners. With the popularization of computers, people have gradually transitioned from traditional manual bookkeeping to using computers to keep accounts, which has brought many benefits and significantly improved the efficiency and accuracy of bookkeeping. Compared with manual bookkeeping, computers can quickly process large amounts of data, reducing the risk of errors in manual calculations. In addition, computer bookkeeping also supports the classification, organization, and retrieval of data, making it convenient for users to access historical records at any time, generate various financial statements, and help users better analyze and manage their financial status.
[0054] However, traditional desktop computers also have many shortcomings for bookkeeping. For example, when a transaction occurs, people cannot write it down immediately and need to wait until they return to the computer to do so, which may lead to omissions or memory errors. With the rise and rapid popularization of mobile Internet and smartphones, more and more applications have appeared on smartphones, and mobile bookkeeping software has also emerged. This type of bookkeeping software has well solved the shortcomings of traditional desktop computer bookkeeping. From a technical point of view, users can record transactions at any time.
[0055] For example, taking a mobile phone as an example, Figure 1 It shows a scenario in which a user uses mobile accounting software to keep accounts.
[0056] The phone can display when it is turned on Figure 1 The interface 1 shown in the figure may include "Weather", "Calendar", "Mail", "Gallery", "Clock", "Music", "WeChat", "China Merchants Bank"
[0057] Icons for apps like Accounting, Camera, Contacts, Settings, and Phone.
[0058] The mobile phone can respond to the user's touch operation on the "accounting" application icon 101 in the interface 1, enter the accounting application, and display the following Figure 1 The accounting interface 2 shown can display the income, expenditure and balance of this month, the transaction record of the day, etc. The accounting interface 2 can include an accounting input button 201.
[0059] The mobile phone can respond to the user's touch operation on the accounting input button 201 and display the following Figure 1 The accounting interface 3 shown in the figure uses a card organization method. Different cards can be selected to represent different accounting categories, such as expenditure, income, transfer, balance, debit, etc. If the user wants to record an expenditure, he can select the expenditure card in the input interface 3, enter the payment amount, modify and select the specific category, account and time, and add notes to complete the expenditure accounting. Among them, the selection of category can be done by clicking the category control 301 in the interface 3 to pop up the selection page at the bottom as shown in the figure. Figure 1 As shown in interface 4. If it is an income, you can select the income card in input interface 4, enter the income amount, modify and confirm the specific category, account and time, and add notes to complete the income accounting.
[0060] Understandably, Figure 1 The applications and application names in the main interface 1 of the mobile phone shown are examples of the embodiments of the present application, and do not constitute any limitation on the applications and application names in the electronic device. Figure 1The accounting interfaces 1, 2, 3 and 4 shown are also examples in the embodiments of the present application, and the displayed content does not constitute any limitation on the content included in the actual accounting interface.
[0061] It can be seen from this that this type of accounting software can only record running accounts. Recording each account requires multiple operation steps and is inefficient. There is no systematic chart of accounts and the granularity of accounting cannot be controlled. It is not flexible enough and cannot support double-entry bookkeeping. It is difficult for users to gradually establish a systematic financial concept during use, which greatly dampens users' enthusiasm for long-term accounting.
[0062] In order to solve these problems, improve the efficiency and accuracy of bookkeeping, and enable users to master the real double-entry bookkeeping method in the bookkeeping process, improve the user's bookkeeping experience, and allow users to truly benefit from long-term bookkeeping, the present application embodiment provides a conversational bookkeeping method. The bookkeeping method provided in the present application embodiment is applied to electronic devices.
[0063] The accounting method provided in the embodiment of the present application can be used as an independent application program or as a part embedded in other application software. The method is applicable to scenarios where electronic devices are interconnected based on a communication network. The communication network can be a local area network, a wide area network transferred through a relay device, or a combination of a local area network and a wide area network. Figure 2 In the scenario shown in (1), a mobile phone, a laptop computer and a remote server can exchange data, such as voice, image, text, ledger and other data, as well as their processing results, through a communication network.
[0064] The accounting method provided in the embodiment of the present application can be used as an independent application. Specifically, the application developed based on the accounting method provided in the embodiment of the present application can be independently run on an electronic device, or run on a client-server system consisting of an electronic device and a remote server.
[0065] The accounting method provided in the embodiment of the present application can also be embedded as a part in other applications. Specifically, the application developed based on the accounting method provided in the embodiment of the present application can be used as a part of other independent applications and run together with the application. For example, the application is embedded in instant messaging applications such as WeChat, WhatsApp and LINE, so that users can complete accounting operations when using these instant messaging applications. When running these instant messaging applications on electronic devices, part of the accounting method runs on the electronic device together with the instant messaging application, and another part of the accounting method runs on a remote server to form a client-server system.
[0066] In some embodiments of the present application, the electronic device may also be a portable electronic device that also includes other functions such as a personal digital assistant and / or a music player function, such as a mobile phone, a tablet computer, a wearable device with wireless communication function (such as a smart watch), etc.
[0067] For example, Figure 2 (2) shows a schematic diagram of the structure of an electronic device.
[0068] The electronic device in the embodiment of the present application may be a mobile phone 200. The electronic device hereinafter is taken as an example of a mobile phone. Figure 2 (2) shows a schematic diagram of the structure of the mobile phone 200, which includes a processor 210, a storage module 223, a power module 260, a wireless communication module 250, a mobile communication module 240, an audio module 220, a camera module 222, a display screen 221, a sensor module 230 and other modules necessary for a smart phone.
[0069] It is to be understood that the structure illustrated in the embodiment of the present application does not constitute a specific limitation on the electronic device mobile phone 200. In other embodiments of the present application, the electronic device may include more or fewer components than shown in the figure, or combine some components, or split some components, or arrange the components differently. The components shown in the figure may be implemented in hardware, software, or a combination of software and hardware.
[0070] The processor may include one or more processing units, for example, the processor may include an application processor, a modem processor, a graphics processor, an image signal processor, a controller, a video codec, a digital signal processor, a baseband processor, and / or a neural network processor, etc. Among them, different processing units may be independent processors or integrated into one or more processors.
[0071] The processor can generate operation control signals based on instruction opcodes and timing signals to complete the control of instruction fetching and execution.
[0072] In the embodiment of the present application, the processor 210 can process the provided accounting method, obtain signals from the touch sensor 230A and / or the pressure sensor 230B, the audio module 220, and / or the camera module 222 on the display screen according to the specific situation to process the user's input information, and then perform processing instructions. The processor 210 executes these instructions, saves the results to the storage module, and displays the feedback information on the display screen 221. In this process, different processing units may be called according to the specific situation to achieve the best processing effect, and a third-party API may be called through a mobile communication module or a wireless communication module, and data on a remote server may be accessed.
[0073] The processor 210 may also be provided with a memory for storing instructions and data. In some embodiments, the memory in the processor 210 may be a high-speed cache memory. The memory may store instructions or data that have been used by the processor or are frequently used. If the processor 210 needs to use the instruction or data, it may be directly called from the memory. This avoids repeated access, reduces the waiting time of the processor, and thus improves the efficiency of the system.
[0074] In some embodiments, the processor 210 may include one or more interfaces. The interface may include an integrated circuit interface, an integrated circuit built-in audio interface, a pulse code modulation interface, a universal asynchronous receiver and transmitter interface, a mobile industry processor interface, a general input and output interface, and / or a universal serial bus interface. The processor may contact the sensor module 230, the audio module 220, the wireless communication module 250, the display screen 221, the camera module 222 and other modules through at least one of the above interfaces.
[0075] The internal memory in the storage module 223 is used to store computer executable program codes, which include instructions. The internal memory may include a program storage area and a data storage area. The program storage area may store an operating system, application programs, such as a calendar program and an application program developed according to the present accounting method, and the data storage area may store various data created during the use of the mobile phone 200, such as captured images and processed data therebetween. The mobile phone 200 may also access data stored on a remote server through the wireless communication module 250 or the mobile communication module 240.
[0076] The wireless communication function of the mobile phone 200 can be realized through the antenna 1, the antenna 2, the mobile communication module 240, the wireless communication module 250, the modem processor and the baseband processor.
[0077] The mobile phone 200 can realize the display function through the display screen 221, the graphics processor in the processor 210, and the application processor.
[0078] The mobile phone 200 can realize the camera function through the camera module 222, the image signal processor, the display screen 221, the video codec, the graphics processor, the application processor, and the neural network processor in the processor 210.
[0079] The mobile phone 200 can implement audio functions such as voice input and playback through the speaker, receiver, microphone, headphone interface in the audio module 220 and one or more processing units in the processor 210.
[0080] In some embodiments, the software system of the mobile phone can adopt a layered architecture, an event-driven architecture, a micro-core architecture, a micro-service architecture or a cloud architecture. The present application embodiment takes the Android system of the layered architecture as an example to exemplify the software structure of the mobile phone 200.
[0081] Figure 3 An example of a software structure block diagram of an electronic device is provided.
[0082] The layered architecture divides the software into several layers, each with clear roles and division of labor. The layers communicate with each other through software interfaces. In some embodiments, the Android system is divided into four layers, from top to bottom: the application layer, the application framework layer, the Android runtime and system library, and the kernel layer.
[0083] An application can consist of a series of application packages.
[0084] like Figure 3 As shown, the applications include camera, SMS, map, China Merchants Bank, calendar, music, phone, WeChat, accounting application and other applications.
[0085] The application framework provides an application programming interface (API) and a programming framework for applications in the application layer. The application framework layer contains some predefined functions.
[0086] like Figure 3 As shown, the application framework may include a window manager, a content provider, a phone manager, a resource manager, a notification manager, a view system, and may also include an OCR module and an NLP module, etc.
[0087] The Android runtime includes the core library and the virtual machine. The Android runtime is responsible for the scheduling and management of the Android system.
[0088] The application layer and the application framework layer run in a virtual machine. The virtual machine executes the Java files of the application layer and the application framework layer as binary files. The virtual machine is used to perform functions such as object life cycle management, stack management, thread management, security and exception management, and garbage collection.
[0089] The system layer may include multiple functional modules, such as a surface manager, a media library, a 3D graphics processing library (such as OpenGL ES), a 2D graphics engine (such as SGL), etc.
[0090] The surface manager is used to manage the display subsystem and provide the fusion of 2D and 3D layers for multiple applications.
[0091] The media library supports playback and recording of a variety of commonly used audio and video formats, as well as static image files, etc. The media library can support a variety of video encoding formats, such as: MPEG4, H.264, MP3, AAC, AMR, JPG, PNG, etc. The kernel layer is a layer between hardware and software. The kernel layer includes at least a display driver, a camera driver, a WiFi driver, an audio driver, and a sensor driver. Among them, hardware can refer to various types of sensors, such as the touch sensor and pressure sensor involved in the embodiments of the present application.
[0092] The embodiment of the present application provides a conversational accounting method, which can be run on an electronic device as an independent application or as part of other applications, and the method includes: the electronic device displays a conversational user interface including a display area and an input area, and the conversational user interface supports multimodal interaction; receives transaction information input by the user through the conversational user interface input area; the electronic device generates a corresponding transaction text by parsing the transaction information, and generates accounting entries based on the transaction text; stores the transaction text and the corresponding generated accounting entries in the accounting book, and displays the execution results to the user in the display area; and supports the modification and deletion of accounting entries, as well as the setting of the chart of accounts, the accounting assistant and the accounting report query function. This accounting method realizes efficient and convenient conversational accounting through technologies such as automatic speech recognition, optical character recognition, natural language processing, and retrieval enhancement generation, improves the user's accounting efficiency, and meets personalized needs.
[0093] The following will combine the accompanying drawings and application scenarios, taking the life scenarios of accounting users as an example, to describe in detail the accounting method provided in the embodiments of the present application.
[0094] For example, if Figure 1 As shown, the mobile phone responds to the user's similar Figure 1 Click the accounting application icon 101 shown in the middle interface 1, and the mobile phone runs the accounting application, and the following is displayed: Figure 4 The default dialog user interface shown in interface 410 is the accounting dialog.
[0095] Here we first explain the conversational user interface. Figure 4The middle interface 410 shows a default dialogue user interface on a mobile phone. Among them, button 411 is an accounting assistant button. When the user clicks the button, the mobile phone will respond and run the accounting assistant function. The user can also open the accounting assistant page by entering "assistant" in the input area 415 and sending it. Topic 412 displays the current topic. As shown in the figure, the current interactive interface for adding, modifying and deleting accounting entries is "Accounting Dialogue". Topic 412 can also be displayed as "Accounting Assistant", "Accounting Chart" and "Financial Statements", corresponding to the mobile phone running the accounting assistant, setting accounting entries and querying financial statements respectively. Button 413 is a menu button, which includes two menu options, namely "Accounting Chart" and "Financial Statements". Click the corresponding menu option, and the mobile phone responds to the operation and runs the corresponding function. These two functions can also be started by entering "Accounting Chart" and "Report" in the input area 415 and sending them. Display area 414 can display the interaction between the user and the accounting application, including the "Accounting Dialogue" page and the "Accounting Assistant" page, and can also display the "Chart of Accounts" page and the "Financial Statement" page. Input area 415 is the input area of the dialogue user interface, which supports single type or multi-modal combination input. The types include text, voice, image and other accounting electronic vouchers, and have multi-modal input parsing capabilities. The images include but are not limited to invoices, receipts and orders, and the other accounting electronic vouchers include but are not limited to electronic invoices, bank statements, QR codes and barcodes. Input area 415 includes text input area 417, plus button 416, voice button 418, and send button 419. Among them, the function of text input area 417 is to input text information; plus button 416, the mobile phone responds to the click operation of the button, and can display a pop-up menu, including two menu options, namely "take a photo" and "attach a file". The mobile phone responds to the click operation of the "take a photo" option. The camera of the mobile phone will open. After the user takes a photo, the mobile phone will display the photo as follows. Figure 4 The input display area 421 shown in the middle interface 420 is a part of the input area, which pops up automatically when displaying images, voices and files; in response to clicking the "Attach File" option, the mobile phone can upload photos, images or other formats of files such as PDF, etc., and display them on the screen. Figure 4 The input display area 421 shown in the middle interface 420. Figure 4 The voice button 418 shown in the middle interface 410, the mobile phone responds to the pressing of the voice button 418, receives the user's voice through the audio module, and after the user releases the voice button 418, the recorded voice will be displayed as shown in the figure. Figure 4 The input display area 421 shown in the middle interface 420. When the user completes the input in the input area, click Figure 4After the send button 425 shown in the middle interface 420 is pressed, the mobile phone will further process the input information in response to the operation.
[0096] For the convenience of explanation and description, the following Table 1 shows the predefined “single male” chart of accounts by way of example.
[0097] Table 1
[0098]
[0099] The chart of accounts in the above table is classified according to three levels. Its specific classification and display format are only an example in the embodiment of this application and do not constitute any limitation to this accounting method. Among them, the secondary classification "China" in assets, liabilities and income refers to the currency of RMB, which is used to distinguish accounts in different currencies; the secondary classification "living" in the expenditure classification mainly puts together those categories related to life, especially residence, with relatively large amounts, and relatively stable and continuous expenses. The categories in "use" are also closely related to daily life, but the difference from "live" lies in the randomness of occurrence, or the relatively small amount.
[0100] The formulation of the chart of accounts is influenced by both objective and subjective factors, and there is no perfect solution. Large enterprises must comply with industry standards, laws and regulations, and financial reporting requirements. These objective constraints provide a basic framework for their account classification. In contrast, the classification of individuals and families is less restricted, but still needs to follow some basic logic, such as taxi fares should not be classified as clothing. More specific classifications can be flexibly defined based on personal needs and lifestyles. For example, clothing can be divided by purpose (formal, casual, sports) or type (tops, pants, shoes and hats). This flexibility enables the chart of accounts to better serve the accounting needs of individuals, families, or small businesses.
[0101] It should be noted that when the mobile phone runs the accounting application for the first time, it will generally notify the user to register. Only after the user account registration is completed can the execution continue. Figure 4 For example, users can register using their own mobile phone or email, or use other accounts such as WeChat and Alipay to register an account for the accounting application. After the user completes the registration of the accounting application, the accounting application will pop up a page to set the chart of accounts, reminding the user to set the chart of accounts first, such as Figure 4The theme on the interface 430 in the figure changes to "Chart of Accounts", and there is a close button 431 on the right side. Clicking this button can close the page of the chart of accounts and return to the accounting dialogue page. Four options are displayed on the page of the chart of accounts, "Single Male", "Single Female", "Small Family" and "Small Business". For example, the user clicks "Single Male", and the mobile phone responds to the operation, and the page switches to a specific category display. The category is displayed in the form of cards. The first page displayed is the "Income" page. The user can click different cards to display the corresponding page, such as Figure 4 The middle interface 440 shows the "Assets" page. It should be noted that when setting the chart of accounts, after selecting the "Single Male" option, the user can then click on different cards to browse. If modification is required, click on Figure 4 The modification button 442 on the middle interface 440, the mobile phone responds to this operation so that all cells on the page are in an editable state, and at the same time, an empty cell is displayed after the last cell "Assets: China: Investment" to provide users with a new accounting subject. After the user modifies the content in the cell or / and enters a new accounting entry in the empty cell, the user clicks the confirmation button 441, and the mobile phone responds to the operation and updates the accounting chart in the memory. Other categories can be modified through similar operations, and then the close button 446 next to the subject can be clicked to exit the setting page of the accounting chart.
[0102] If the user has some questions on the Chart of Accounts page, for example, the user is not clear about the meaning of "China: Debit" in the asset, the user can directly Figure 4 In the text input area 444 shown in the interface 440, input "Assets in assets: China: Debit What does it mean?" and click the send button 445 to send. In response to the click operation, the mobile phone will switch to the accounting assistant page (such as Figure 5 The interface 510 shown in (1) is a interface in which the user can communicate with the accounting assistant directly through natural language.
[0103] Exemplarily, after receiving the text input by the user, the mobile phone parses the text content and triggers the loading of the accounting assistant page. Using the retrieval enhancement generation technology, similarity retrieval is performed through the vector database to retrieve the information most relevant to the semantics of the input text from the knowledge base and generate retrieval results. Combined with the retrieval results, the input text, the chart of accounts, and the generation requirements for generating answers, such as "give a brief explanation and give specific suggestions, no more than 200 words", as context input to the generation model, the model infers and generates a help message based on the input content, and finally the help message is displayed in the display area, such as Figure 5The message 511 shown in the interface 510 in (1) is shown in FIG. The generation model may use the DeepSeek large language model, and the vector database may use FAISS to achieve efficient semantic retrieval and storage management.
[0104] For example, after reading the explanation, if the user has no further questions, he or she can directly click the "return" button 512 in the accounting assistant message, and the mobile phone will respond to the operation by returning to the previous page; if the user has further questions, he or she can continue to communicate with the accounting assistant.
[0105] For example, if the user is not satisfied with the help message of this accounting assistant, he can click Figure 5 In the message 511 shown in the interface 510 in (1), the mobile phone clicks the regenerate button in the control group 513 (from left to right, the regenerate button, the like button, and the dislike button) to regenerate the answer to the question using the search enhancement generation technology. The user can also use the like button and the dislike button in the control group 513 to evaluate the help message. Further information can also be provided to the accounting assistant, and the mobile phone can generate further help messages using the search enhancement generation technology in response to the new information.
[0106] For example, the user clicks the close button 514 on the right side of the accounting assistant theme, and the mobile phone responds to the operation and returns to the following Figure 4 After viewing the interface 440 shown in the figure, the user feels that most of this asset classification has met his needs, but he hopes to modify "Assets: China: Debit" to a more specific classification that suits his situation. The user's current deposits are mainly placed in the debit card of China Merchants Bank, so the user needs to modify this classification accordingly and change it to "Assets: China: China Merchants Debit Card". The user can click Figure 4 The user clicks the edit button 442 on the middle interface 440, and the mobile phone responds to the operation so that all cells on the page are editable. At this time, the user modifies the content in the cell 443 to "Assets: China: Merchants Debit Card", and clicks the confirm button 441. The mobile phone responds to the operation to modify the corresponding accounting account in the accounting account table stored in the memory. For example, the user also wants to add a category "Assets: China: WeChat", and the user can click the following Figure 4 Click the edit button 442 on the middle interface 440. The mobile phone responds to this operation so that all cells on the page are editable and an empty cell is displayed below "Assets: China: Investment". The user directly enters "Assets: China: WeChat" into the cell and clicks the confirm button 441. The mobile phone responds to this operation and adds the accounting account to the chart of accounts in the memory. After completing the setting of the chart of accounts, click the close button 446 on the right side of the topic. The mobile phone responds to this operation and returns to the default accounting dialogue interface, such as Figure 5The interface 520 shown in (1) can then be used to formally record accounts.
[0107] For example, let's assume that a user likes photography very much and wants to buy a SLR camera after work. The user has been eyeing Canon's EOS 600D18-55m on JD.com for a while as a gift for his newborn. After the funds are in place, he decisively places an order. The user wants to record this account, and can use one of the following common methods to record it.
[0108] Method 1: The user clicks the plus button in the input area, selects the "Photo" option in the pop-up menu, and takes a photo of the order information on the shopping page or uploads and sends the screenshot of the order directly through the input area to record the transaction. Figure 5 The JD.com order shown in (2) can be placed by clicking Figure 4 After selecting "Photo" in the pop-up menu by clicking the plus button 416 in the middle interface 410, the mobile phone responds to the operation by displaying the following Figure 4 The input display area 421 on the middle interface 420; you can also take a screenshot of the order first, and then upload it by selecting the "Attach File" option through the plus button in the input area. A similar image is also displayed in the input display area 421. Then click the send button 425, and the mobile phone responds to the operation, uses OCR to process the photo to extract the original text, processes the extracted original text, removes invalid characters and advertisements and other information irrelevant to the transaction, and obtains the pre-processed transaction text. The text is semantically parsed through NPL to identify and extract key information in the transaction record, such as "Product: Canon EOS 600D 700D 750D 760D Entry-level SLR Camera Rear Sealed High-definition Student Travel Novice 600D18-55mm Daily Lens Official Standard Store Warranty for Five Years; Product Quantity: 1; Actual Payment: ¥2518.00", and the transaction text will be used to generate accounting entries. Using the retrieval enhancement generation technology, similarity retrieval is performed through the vector database. Based on the transaction text, information related to the input description is retrieved from the knowledge base; the retrieval results, the transaction text, the chart of accounts, and the generation requirements of accounting entries that meet the double-entry bookkeeping requirements are input into the generation model as context to generate accounting entries that meet user requirements, add a unique Nano ID code to the accounting entry, and display the final accounting entry on the display area of the user dialogue interface, such as Figure 6 As shown in the middle interface 610.
[0109] Exemplarily, the requirements for generating accounting entries that meet the requirements of double-entry bookkeeping may be "Based on the input transaction text, generate accounting entries that comply with the double-entry bookkeeping principle. The first line of the accounting entry contains the date and time. If the transaction text does not contain the date and / or time, the current date and / or time is automatically used; the last line is a description, the content of which contains information in the transaction text other than that already included in other parts of the accounting entry. If the transaction text is lengthy, a summary is made; between the two are debit or credit entries, each entry occupies one line, and the accounting subject in each debit and credit entry must select the most appropriate category from the defined chart of accounts. If the transaction text does not contain information on the payment method, it is deemed by default to be paid through WeChat.".
[0110] Wherein, the knowledge base includes accounting books, charts of accounts, accounting-related knowledge, question and answer records and a vector database. The vector database stores embedded vectors to support efficient similarity retrieval. The embedded vector is calculated by a pre-trained embedding model based on the accounting books, the charts of accounts, the accounting-related knowledge and the question and answer record data. The chart of accounts includes all pre-defined accounting accounts. The accounting-related knowledge includes but is not limited to accounting principles and financial statement preparation texts. The pre-trained embedding model can use DeepSeek-Embedding to convert the above knowledge base data into embedded vectors for storage in a vector database (such as FAISS) and support efficient retrieval based on vector similarity.
[0111] In this case, for example, the user thinks that the automatically extracted description is too long and unnecessary, and wants to make some modifications, so he can directly copy the following: Figure 6 The message 611 shown in the middle interface 610 is pasted in Figure 6 In the text input area 612 shown in the middle interface 610, the description is deleted and the send button is clicked to send it out. The mobile phone responds to the operation and displays it in the display area, such as Figure 6 As shown in the message 621 in the middle interface 620, the mobile phone determines that the accounting entry is already stored in the accounting book based on the input information through its Nano ID, so it finds the accounting entry stored in the accounting book through the Nano ID and updates its content according to the latest input, and then displays a prompt message of the completion of the update in the display area, such as Figure 6 As shown in message 622 in middle interface 620.
[0112] In this case, for example, if the user does not want the accounting entry, the user can directly copy the complete accounting entry or the Nano ID information of the accounting entry, paste it in the text input area, and add the word "delete" at the beginning and send it out, such as Figure 6As shown in the message 623 sent out in the middle interface 620, the mobile phone responds to the operation by deleting the accounting entry and its corresponding transaction text stored in the accounting book, and displays a prompt message of the completion of the deletion in the display area, such as Figure 6 As shown in message 624 in middle interface 620.
[0113] For example, in another case, Figure 6 After the message 611 is given on the middle interface 610, the user wishes to adjust the accounting subject of the transaction of purchasing the camera. By default, the transaction will be classified into the "Expenses: Living: Digital Products" account, but the user considers that there may be more camera-related expenses in the future and hopes to track such transactions separately. In addition, since products such as cameras are of great value, the user wishes to create an independent accounting subject, such as "Expenses: Living: Camera". Since this category does not yet exist in the chart of accounts, in this case, the user needs to first create an "Expenses: Living: Camera" in the chart of accounts. The user clicks on Figure 6 The menu key 613 in the interface 610, the mobile phone responds to the user's operation and pops up a pop-up window in the display area such as Figure 6 In the menu shown in interface 630, the user clicks on the "chart of accounts" menu item, and the mobile phone displays the following in response to the user's click: Figure 6 In the interface 640 shown in the figure, click the Modify button 641 in the figure, and an empty cell will appear after the last cell "Expense: Insurance: Insurance", and the user enters "Expense: Accommodation: Camera" in it, as shown in the new cell 642 in the interface 640, and clicks the Confirm button on the page. In response to this operation, the mobile phone will add the accounting subject to the chart of accounts. The user clicks the Close button 643 on the right side of the subject, and the mobile phone will respond to this operation by closing the chart of accounts page and returning to the interface 610. Similar to the operation of modifying accounting entries described above, the user directly copies the following Figure 6 The message 611 in the interface 610 is pasted in the article input area 612 shown in the interface 610, the debit entry is modified and the description is deleted, and then the send button is clicked to send it out. In response to the operation, the mobile phone displays the modified accounting entry in the display area, such as Figure 7 As shown in message 711 in interface 710, the mobile phone determines that the accounting entry is already stored in the accounting book based on the input information through its Nano ID, so it finds the accounting entry stored in the accounting book through the Nano ID and replaces its content according to the latest input, and then displays a prompt message of the completion of the update in the display area, such as Figure 7 As shown in message 721 in interface 720.
[0114] Method 2: The user clicks the plus button in the input area, selects the "Take a photo" option in the pop-up menu, takes a photo of the order information on the shopping page, and simultaneously enters the payment information not included in the order information through text.
[0115] For example, the user directly clicks on Figure 4 The plus button 416 in the middle interface 410 selects "Photo" in the pop-up menu. After the photo is taken, the image is displayed as shown below. Figure 4 In the input display area 421 shown in the middle interface 420, further directly input the text "China Merchants Bank Debit Card Payment" in the text input area 423, and then click the send button 425 on the right side of the input area to send. Afterwards, in the same way, the mobile phone responds to the operation by using OCR to process the photo, obtains the pre-processed text, merges the input plain text "China Merchants Bank Debit Card Payment" to obtain the transaction text, and performs semantic analysis on the text through NPL to identify and extract key information in the transaction record, such as "Product: Canon EOS 600D 700D 750D 760D Entry-level SLR Camera Sealed HD Student Travel Novice 600D18-55mm Daily Lens Official Standard Store Warranty for Five Years; Product Quantity: 1; Actual Payment: ¥2518.00; China Merchants Bank Debit Card Payment", the transaction text will be used to generate accounting entries. Similar to the above, the accounting entries are generated using retrieval enhancement generation technology and a unique code Nano ID is added, and the final accounting entries are displayed on the display area of the user dialogue interface, such as Figure 7 As shown in message 731 in the middle interface 730, the detailed process will not be repeated.
[0116] Method 3: The user clicks the plus button in the input area, selects the "Take a photo" option in the pop-up menu, takes a photo of the order information on the shopping page, and inputs the payment information not included in the order information by voice.
[0117] That is, replace the text input "CMB debit card payment" with voice input, analyze the voice content through automatic speech recognition technology, and process the extracted original text, including but not limited to normalization, removal of irrelevant characters and spelling correction, to obtain the pre-processed text. The subsequent processing process is similar to method 2 and will not be repeated here. Method 4: The user enters the complete transaction information in the input area through text. Figure 7 As shown in the middle interface 740, the process is similar to the above process and will not be described again here.
[0118] Method 5: The user inputs the complete transaction information in the input area by voice. This is similar to the above process and will not be described here.
[0119] For example, let's assume a new scenario where a user receives a task to purchase a batch of printing supplies for the company where he works, but the user needs to pay for it first and then ask the company's accountant to reimburse him. In order to track the money he spent on the company and make it easier to reimburse him later, the user wants to record these transactions, but the user has never kept similar accounts and doesn't know how to do it. So the user clicks on Figure 7 The accounting assistant icon 732 in the upper left corner of the middle interface 730, the mobile phone responds to the operation to enter the accounting assistant page, such as Figure 8 As shown in the middle interface 810, the user enters in the input area "I need to purchase a batch of printing supplies for the company now and I need to advance the funds. I want to set up an account to track the purchases for the company. What account do I need to add? If I need to be reimbursed later, how should I record the account? Please try to use the accounts that have been established. If you need to create a new accounting subject, please give suggestions and explanations in advance." and clicks the send button, as shown in the message 812 of the middle interface 810. In response to this operation, the mobile phone uses the message as input, generates suggestions using the search enhancement generation technology, and displays the suggestions in the display area, such as Figure 8 The reply message shown in the middle interface 810 and 820 gives the specific accounting subject "Assets: China: Company Advance Payment" that needs to be added, as well as how to record the advance payment and reimbursement accounts, and prompts the user to directly click the underlined "Assets: China: Company Advance Payment" link 821 to directly enter the accounting chart page for confirmation or further modification and confirmation. The user clicks link 821, and the mobile phone responds to the operation, displays the asset page of the accounting chart, and automatically adds a new cell "Assets: China: Company Advance Payment" at the end, as shown in the figure below. Figure 8 The user thinks that no modification is needed, so he clicks Figure 8 The confirmation button on the middle interface 830 is then clicked, and the close button 831 on the right side of the topic is then clicked to close the chart of accounts page. The mobile phone responds to this operation, and the chart of accounts in the memory is updated. At the same time, the user returns to the bookkeeping assistant page, and the user automatically replies "Add: Assets: China: Company Advance Payment". Figure 8 As shown in interface 820, the user clicks the close button 823 on the right side of the accounting assistant theme again to close the accounting assistant page and return to the accounting dialogue page.
[0120] For example, if the user is not satisfied with the help message of this accounting assistant, he can click Figure 8 The corresponding buttons in the control group 822 (from left to right, the regenerate button, the like button and the dislike button) included in the message shown in the middle interface 820 are used to execute the corresponding functions. Figure 5 The description in the middle interface 510 is similar, and the detailed process is not repeated here.
[0121] For example, after the user purchases printer ink and printing paper at an offline store, the user enters in the input area "Purchasing printer ink and printing paper at the HP store for a total of 500 yuan, paid in cash, and paid for the company." and clicks Send. The user can also press and hold the voice button on the right to record a voice with the same content, and then click Send. Voice input requires the use of automatic speech processing to recognize the voice and convert it into text. The subsequent processing of the two texts is similar to the previous description and will not be repeated here. Furthermore, the mobile phone uses retrieval enhancement generation technology to generate accounting entries, adds Nano ID to the accounting entries, saves them in the memory, and displays them as output in the display area of the accounting dialogue, such as Figure 8 As shown in the middle interface 840, the detailed process is similar to the above and will not be repeated here.
[0122] For example, after the reimbursement is completed, you only need to enter the information directly through the mobile phone, such as "reimburse the company for the 500 yuan spent on purchasing printer ink and printing paper from HP, and deposit the reimbursement into the China Merchants Bank debit card." The mobile phone responds to the input and can automatically generate accounting entries and display them in the display area.
[0123] The above examples explain how users use the accounting method described in this application in real life scenarios through specific scenarios. In daily life, most of the scenarios where the accounting application is used are to record current transactions. Therefore, whether the path from inputting transactions to generating qualified accounting entries is convenient will greatly affect the user's accounting efficiency and usage experience. The software uses multimodal input and parsing methods, fully utilizing retrieval enhancement generation technology to quickly generate accounting entries, improving the flexibility and timeliness of transaction input, while helping users solve specific problems encountered in accounting, making the entire accounting experience smoother and thus improving the efficiency of accounting.
[0124] The purpose of bookkeeping is to understand your income and expenditure, as well as your assets and liabilities, so that you can make further financial decisions based on these. Therefore, the accounting books formed by the recorded accounting entries generate relevant financial statements so that users can understand their specific financial situation over a certain period of time, such as a week, a month, a year, etc. Users can click on Figure 6 The "Financial Statement" option in the drop-down menu in the upper right corner of the middle interface 630, the mobile phone responds to the operation and displays the financial statement page, such as Fig. 9As shown in (1). For example, the page in the figure shows two card items. The current page shows the income and expenditure table page, which can be switched by clicking on the name of the card. The current figure uses a sunburst chart to show the various categories of expenditure or income and their proportions. This is just a simple example. In fact, various display methods can be used. Mobile phones are electronic devices with relatively small screens. Due to the limitation of the screen, they can only display relatively limited information and functions, but they can fully meet the daily query needs of users.
[0125] If the user wishes to conduct further inquiries and edits on the report, he or she may log in to the web page on a large-screen electronic device, such as a laptop or desktop computer, to query the financial statements, detailed accounting entries, or further edit certain accounting entries, or further display other statistical information, such as the usage of each different category of accounting subjects, etc. Fig. 9 (2) is only an example and does not represent all functions. By combining the instant response and convenient operation of small-screen devices with the advantages of large-screen devices in data display capacity and batch operation, users can give full play to the characteristics of different terminal devices, realize the seamless connection between real-time recording of accounting data and report generation, and significantly improve accounting efficiency and data processing accuracy. Based on the above description, Fig.10 As shown, an embodiment of the present application provides a conversational accounting method, which may include the following steps:
[0126] Step 101: display a dialog user interface including a display area and an input area, wherein the dialog user interface supports multimodal interaction.
[0127] like Figure 4 The dialog user interfaces shown in the middle interfaces 410 and 420 include an input area and a display area.
[0128] Users mainly enter information through the input area of the conversational user interface, such as Figure 4 The input area 415 shown in the middle interface 410 includes a plus button 416, a text input area 417, a voice button 418 and a send button 419. The information input function in the input area is completed through the cooperation of these buttons. The input area supports input types including text, voice, image and other electronic vouchers for accounting. It can be a single type of input or a combination of the above types of input, and has multi-modal input parsing capabilities. The user can also click on Figure 4 The accounting assistant button 411 and the menu button 413 in the interface 410 open the accounting assistant page, the setting chart of accounts page, and the query financial report page respectively.
[0129] The display area supports display content including but not limited to single-type display of text, voice, image and control or combination of multiple single-type display. Figure 5 The message shown in the middle interface 510 includes the display of text and button 512, such as Figure 8 The message display shown in the middle interface 820 includes a link 821 .
[0130] The display area or the entire dialog user interface can be switched to a web page display according to user input to load or embed web page content and support interactive operations of web page content. Figure 4 The display area shown in the middle interface 430 and 440 displays the accounting account chart page, and the user can edit and modify the accounting accounts through this page. Fig. 9 The financial statement page shown in (1) above is where users can view financial statements.
[0131] Step 102, receiving transaction information input by the user through the dialogue user interface input area.
[0132] Receive transaction information entered by the user in the input area, including text, voice, image and electronic accounting voucher. Specifically, refer to any one of the input methods from method 1 to method 5 above, which will not be repeated here.
[0133] Step 103, generating a corresponding transaction text by parsing the transaction information, and generating an accounting entry based on the transaction text.
[0134] The received new transaction information is parsed and finally converted into transaction text. The transaction text is used as input and the accounting entries that meet the requirements of double-entry bookkeeping are generated using retrieval enhancement generation technology. For details, please refer to the detailed description of the parsing and generation of accounting entries in the above method 1, which will not be repeated here.
[0135] Step 104, storing the transaction text and the corresponding accounting entries in the accounting book, and displaying the execution results to the user in the display area.
[0136] The transaction text and the corresponding accounting entries are stored in the accounting book, which is stored in the storage module of the mobile phone and can also be uploaded to the remote server for storage. The generated accounting entries are displayed in the display area, such as Figure 6 Message 611 shown in the middle interface 610.
[0137] The above steps are the core process of adding new accounting entries in this conversational accounting method. On this basis, it also includes modifying accounting entries, deleting accounting entries, setting up chart of accounts, accounting assistant and querying financial statements. These functions are described in detail in the above specific scenarios of purchasing SLR cameras and purchasing printing consumables for the company, so they will not be repeated here.
[0138] Another embodiment of the present application provides an accounting device, such as Fig.11 As shown, the accounting device is used to implement the method described in the above method embodiment, and includes: an input and output module 1101, a processing module 1102 and a storage module 1103.
[0139] Among them, the input and output module 1101 is used to allow users to input relevant data through sensors such as touch, microphone, camera module, etc., and output feedback to users through screens, speakers, etc. The input and output module specifically includes an input module and a display module, and the display module is used to display interactive information with the user. In physical entities, the display module and the input module are specifically embodied as touch screens. The input module is used for user input, and supports single-type or multi-modal combination input. The types include text, voice, image and other electronic vouchers for accounting, and have a multi-modal input parsing function; the display module is used to display the results of the execution, and the supported content includes single-type display or multi-type combination display of text, voice, image and interactive components. Among them, the display area or the entire dialogue user interface can be switched to a web page display according to user input.
[0140] Processing module 1102 specifically processes the functions of extracting the content of input information, generating and adding accounting entries, modifying accounting entries, deleting accounting entries, setting the chart of accounts, running the bookkeeping assistant, and querying financial statements. The processor can also communicate with the remote server through a network interface (not shown in the figure). In some cases, the processing module may include the computing power of the remote server so that some computing tasks (such as reasoning calculations based on large-scale neural networks) are performed on the remote server. For example, a large language model in an intelligent model may need to call a third-party API (such as DeepSeek's API) or deploy the corresponding model on the remote server. In this case, the local processor works in conjunction with the remote server to form a processing module.
[0141] The storage module 1103 is used to store data including temporarily stored data and data that needs to be stored for a long time, wherein the temporarily stored data includes the intermediate data required for the operation of the accounting device; the long-term stored data includes accounting books, charts of accounts, accounting-related knowledge, question and answer records, and embedded vectors, which are used to support retrieval enhancement generation technology. The storage module physically includes both the memory on the local electronic device and the memory on the remote server. Through the network interface (not shown in the figure), the processing module 1102 can obtain data from the storage module on the remote server. In the present invention, the data on the server may include all possible data mentioned above.
[0142] Furthermore, the processing module 1102 includes an extraction module, an instruction generation module, an intelligent model module and an execution module.
[0143] All relevant contents of each step involved in the above method embodiment can be referred to the functional description of the corresponding functional module, and will not be repeated here.
[0144] Another embodiment of the present application provides an electronic device, including a touch screen, one or more processors and one or more memories, wherein a computer program is stored in the memory, and the computer program includes computer instructions. When the computer program in the memory is executed by the processor of the electronic device, the electronic device can implement the accounting method described in any one of the above embodiments.
[0145] Another embodiment of the present application provides a client-server system, which includes an electronic device and a remote server, wherein the electronic device and the remote server respectively include one or more processors and one or more memories, wherein the memories store a computer program, and the computer program includes computer instructions. When the computer program is executed on the electronic device and the remote server, the electronic device and the remote server respectively implement at least a part of the accounting method described in any one of the above embodiments, and collaboratively complete the accounting method described in any one of the above embodiments.
[0146] An embodiment of the present application also provides a computer-readable storage medium, which includes a computer program. When the computer program is executed by a processor of an electronic device, the electronic device is enabled to implement the accounting method described in any one of the above embodiments; or, when the computer program is collaboratively executed on the electronic device and a remote server, the electronic device and the remote server respectively implement at least a portion of the accounting method described in any one of the above embodiments, and collaboratively complete the accounting method described in any one of the above embodiments.
[0147] An embodiment of the present application also provides a computer program product, which includes a computer program. When the computer program is executed by a processor in an electronic device, the electronic device is enabled to implement the accounting method described in any one of the above embodiments; or, when the computer program is collaboratively executed on the electronic device and the remote server, the electronic device and the remote server respectively implement at least a portion of the accounting method described in any one of the above embodiments, and collaboratively complete the accounting method described in any one of the above embodiments.
[0148] In the several embodiments provided in the present application, it should be understood that the described methods and devices can be implemented in other different ways. For example, the division of the above-mentioned functional modules is only schematic. During the specific implementation, multiple functional modules can be re-divided as needed, or integrated into a single module, or completed by multiple independent components. In addition, the connection between different functional modules can be wired or wireless, and can also be implemented through hardware interfaces, software interfaces or other forms.
[0149] The unit described as a separate component may or may not be physically separated. The unit may be a hardware module or multiple logical components, which may be located in the same device or distributed on different terminal devices or servers. The specific implementation method may be selected according to actual needs.
[0150] In various embodiments of the present application, each functional unit can be deployed independently or integrated in a processor or storage medium. The unit can be implemented in the form of hardware or its function can be completed by software. The software functional unit can be stored in a storage medium and executed by a processor. The method steps of the embodiments of the present application can be stored in a storage medium such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), etc. to implement the functions.
[0151] The above contents are only embodiments of the present application, but any modification or replacement made by a person skilled in the art after reading the present application shall be deemed to be within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims
1. A conversational accounting method, applied to an electronic device, characterized in that: include: Displaying a conversational user interface (Conversational User Interface) including a display area and an input area, wherein the conversational user interface supports multimodal interaction (Multimodal Interaction); Receiving transaction information input by a user through the dialogue user interface input area; By parsing the transaction information, a corresponding transaction text is generated, and an accounting entry is generated based on the transaction text; The transaction text and the corresponding accounting entries are stored in an accounting book, and the execution results are displayed to the user in the display area.
2. The method according to claim 1, characterized in that: The input area of the conversational user interface supports input types including text, voice, image, and other electronic accounting vouchers, which can be either a single type of input or a combination of the above types of input, and has multi-modal input parsing capabilities; The display area of the dialog user interface supports display content including but not limited to a single type of display of text, voice, image, and control or a combination of multiple single types of display; Among them, the display area or the entire dialogue user interface can be switched to a web page display according to user input to load or embed web page content and support interactive operations of web page content.
3. The method according to claim 1, characterized in that The types of transaction information include: The single types of electronic accounting vouchers include text, voice, image and other types, or the combination of the above single types.
4. The method according to claim 1 or 3, characterized in that: The analysis of the input transaction information includes: when the transaction information is text, it can be directly used as the transaction text; When the transaction information is voice, it is converted into text by voice recognition technology, and the text is pre-processed to generate the transaction text; When the transaction information is an image, it is converted into text by using image recognition technology, and the text is pre-processed to generate the transaction text; When the transaction information is other electronic vouchers for accounting, it is converted into text by using corresponding electronic voucher parsing technology, and the text is pre-processed to generate the transaction text; When the transaction information is a combination of the above-mentioned multiple single types, they are parsed and preprocessed separately and then merged to generate the transaction text.
5. The method according to claim 1 or 4, characterized in that: The accounting entries are generated by using the Retrieval-Augmented Generation (RAG) technology, including: Performing a similarity search through a vector database to retrieve information related to the input description from a knowledge base based on the transaction text; Combining the search results, the transaction text, the chart of accounts, and the generation requirements of accounting entries that meet the double-entry bookkeeping requirements as context inputs to a generation model to generate accounting entries that meet the user's requirements; Wherein, the knowledge base includes: the accounting books, the chart of accounts, accounting related knowledge, question and answer records and vector database; The vector database stores embedded vectors to support efficient similarity retrieval; The embedding vector is calculated by a pre-trained embedding model (PEM) based on the accounting book, the chart of accounts, accounting-related knowledge and the question-and-answer record data; The chart of accounts includes all pre-defined accounting accounts and can be reset according to user needs; The accounting related knowledge includes but is not limited to accounting principles and financial statement preparation texts.
6. The method according to any one of claims 1 to 5, characterized in that: The method further includes: receiving other input information from the dialogue user interface input area, the input information including: text of an existing accounting entry to be modified, text of an existing accounting entry to be deleted, text or voice of setting a chart of accounts, text or voice of starting an accounting assistant, and text or voice of querying a financial statement; Based on the analysis of the input information in the input area, the operation corresponding to the input information is performed, and the operation includes: modifying the original accounting entries saved in the accounting book, and displaying the feedback results of the execution in the display area; deleting the original accounting entries and the corresponding transaction texts saved in the accounting book, and displaying the feedback results of the execution in the display area; opening the page for setting the chart of accounts, and adding, modifying and / or deleting specific accounting accounts in the chart of accounts by interacting with the chart of accounts page; opening the bookkeeping assistant page, and the user obtains bookkeeping-related suggestions and help by interacting with the bookkeeping assistant page; generating corresponding financial statements according to the principle of double-entry bookkeeping, opening the financial statement page, and the user obtains corresponding financial statement information by interacting with the financial statement page.
7. The method according to any one of claims 1 to 6, characterized in that: The method further includes: receiving a control touch input from the dialog user interface, the control touch input including: a touch on a chart of accounts control, a touch on a bookkeeping assistant control, and a touch on a financial statement query control; based on the received control touch input, performing operations corresponding to the controls, the operations including: opening a page for setting a chart of accounts, and adding, modifying, and / or deleting specific accounting accounts in the chart of accounts by interacting with the page; opening a bookkeeping assistant page, and the user obtains bookkeeping-related suggestions and help messages by interacting with the bookkeeping assistant page; generating corresponding financial statements according to the principle of double-entry bookkeeping, and opening a financial statement page, and the user obtains corresponding financial statement information by interacting with the financial statement page; Among them, the accounting assistant includes: converting the information input by the user into question text, combining with the knowledge base, generating the accounting-related suggestions and help messages through the generation model, the accounting-related suggestions and help messages include evaluation controls, and the user can evaluate the quality of the generated messages through the evaluation controls. The question text, the accounting suggestions and help messages, and the evaluation of the accounting suggestions and help messages will all be stored in the question and answer record.
8. The method according to claim 5 or 7, characterized in that: The knowledge base performs an update operation at a predetermined time interval or under a specific trigger condition according to the update status of the accounting book, the chart of accounts, the accounting-related knowledge and the question and answer record, and the update operation includes recalculating the embedding vector of the above data and storing it in the vector database to improve the accuracy of the accounting entry generation and the accounting-related suggestions and help messages provided by the accounting assistant; When embedding the question and answer records, high-quality question and answer records are screened out as embedded texts based on the user's evaluation of each record through the control.
9. The method according to any one of claims 1 to 8, characterized in that: The method further comprises: The financial report page design is optimized for small-screen electronic devices and large-screen electronic devices. For small-screen electronic devices, the page displays simplified core financial information, including but not limited to expenditure statistics, income statistics and balance sheets, to ensure efficient information presentation within limited screen space and facilitate query at any time; For large-screen electronic devices, the page displays more comprehensive financial information, including not only the basic content on small-screen electronic devices, but also a more detailed display of financial data. It also supports but is not limited to the functions of querying, modifying, and deleting accounting entries, as well as statistical analysis of the usage of accounting subjects, thereby providing users with a more intuitive, flexible, and in-depth financial management experience on large-screen devices.
10. An accounting device, characterized in that: include: An input / output module, configured to display the dialog user interface supporting multimodal interaction, receive information input by a user through the input area or touch control on a control on the dialog user interface, and display execution results; A processing module, for processing including: parsing the information in the input area or responding to the touch of the control on the dialog user interface, and performing corresponding operations according to the specific content, including but not limited to: adding, modifying and deleting accounting entries, setting the chart of accounts, starting the accounting assistant function and querying financial statements; A storage module is used to store data including temporarily stored data and data that needs to be stored for a long time, wherein the temporarily stored data includes intermediate data required for the operation of the accounting device; the long-term stored data includes the accounting book, the chart of accounts, the accounting-related knowledge, the question and answer record, and the embedded vector, which is used to support retrieval enhancement generation technology.
11. An electronic device, characterized in that: include: One or more processors and one or more memories, the memory is coupled to the processor, the memory stores computer program code, the computer program code includes computer instructions, when the processor executes the computer instructions, the electronic device executes the accounting method according to any one of claims 1 to 9.
12. A client-server system, characterized in that: include: An electronic device and a remote server, wherein the electronic device and the remote server respectively include one or more processors and one or more memories, wherein the memories store a computer program, and the computer program includes computer instructions. When the computer program is executed on the electronic device and the remote server, the electronic device and the remote server respectively execute at least a part of the steps of the accounting method, and collaboratively complete the accounting method as described in any one of claims 1 to 9.
13. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a computer program, which, when executed on an electronic device, enables the electronic device to perform the accounting method as described in any one of claims 1 to 9; or, when the computer program is executed collaboratively on an electronic device and a remote server, enables the electronic device and the remote server to respectively perform at least a portion of the steps of the accounting method and collaboratively complete the accounting method as described in any one of claims 1 to 9.
14. A computer program product, characterized in that The computer product includes a computer program, which, when executed on an electronic device, enables the electronic device to perform the accounting method as described in any one of claims 1 to 9; or, when the computer program is executed collaboratively on an electronic device and a remote server, enables the electronic device and the remote server to respectively perform at least a portion of the steps of the accounting method and collaboratively complete the accounting method as described in any one of claims 1 to 9.
Citation Information
Cited By
Discrete commodity transaction processing method and system
CN121169412A
A method and system for processing discrete commodity transactions
CN121169412B