Audit report generation method and device, storage medium and electronic equipment

Abstract

The invention discloses an audit report generation method and device, a storage medium and electronic equipment. The method comprises the steps that first auditing data, second auditing data and an auditing prompt template are obtained, the data format of the first auditing data is different from that of the second auditing data, the type of the first auditing data is the same as that of the second auditing data, and the auditing prompt template comprises a target prompt word; determining a plurality of groups of initial character strings according to the first audit data and the second audit data, and executing a target merging operation on the plurality of groups of initial character strings based on the target prompt word to obtain a target character string; and inputting the target character string and the audit prompt template into the target business model to generate a target audit report, and generating different parts in the target audit report by setting temperature parameters of different target business models. The technical problem that the efficiency of manually filling in the audit report is low is solved.

Description

Technical Field

[0001] The present application relates to the field of computers, and more specifically, to a method and device for generating an audit report, a storage medium, and an electronic device. Background Art

[0002] In current audit practices, auditors need to carefully review a large number of financial records, management documents and other relevant materials to evaluate the economic activities and responsibilities of the audited entities. The efficiency and quality of audit report generation are difficult to guarantee. That is, in existing technologies, audit practices are highly dependent on manual labor, which leads to technical problems such as low efficiency of audit reports.

[0003] To address the above-mentioned problems, no effective solutions have been proposed so far. Summary of the Invention

[0004] The embodiments of the present application provide a method and device for generating an audit report, a storage medium, and an electronic device to at least solve the technical problem of low efficiency in manually filling out an audit report.

[0005] According to one aspect of an embodiment of the present application, a method for generating an audit report is provided, comprising: obtaining first audit data, second audit data and an audit prompt template, wherein a data format of the first audit data and a data format of the second audit data are different, a type of the first audit data and a data type of the second audit data are the same, and the audit prompt template includes a target prompt word; determining multiple groups of initial character strings based on the first audit data and the second audit data, performing a target merging operation on the multiple groups of initial character strings based on the target prompt word to obtain a target character string; inputting the target character string and the audit prompt template into a target business model to generate a target audit report, wherein different temperature parameters of the target business model are set to generate different parts of the target audit report.

[0006] According to another aspect of an embodiment of the present application, an audit report generation device is also provided, including: an acquisition module, used to acquire first audit data, second audit data and an audit prompt template, wherein the data format of the first audit data and the data format of the second audit data are different, the type of the first audit data and the data type of the second audit data are the same, and the audit prompt template includes a target prompt word; an execution module, used to determine multiple groups of initial character strings based on the first audit data and the second audit data, and perform a target merging operation on the multiple groups of initial character strings based on the target prompt word to obtain a target character string; a generation module, used to input the target character string and the audit prompt template into a target business model to generate a target audit report, wherein different temperature parameters of the target business model are set to generate different parts of the target audit report.

[0007] Optionally, the device is used to determine multiple groups of initial character strings based on the first audit data and the second audit data in the following manner: converting the first audit data into a first character string, and converting the second audit data into a second character string; obtaining a target quantity in response to a target interaction operation; dividing the first character string into multiple first substrings, and dividing the second character string into multiple second substrings, wherein the sum of the number of the first substrings and the number of the second substrings is the target quantity; and generating multiple groups of the initial character strings based on multiple first substrings and multiple second substrings.

[0008] Optionally, the apparatus is configured to generate multiple groups of initial character strings based on multiple first substrings and multiple second substrings in the following manner, including at least one of the following: combining multiple first substrings and multiple second substrings in chronological order to generate multiple groups of initial character strings, wherein one group of initial character strings corresponds to a preset time interval; and combining multiple first substrings and multiple second substrings in a service execution order to generate multiple groups of initial character strings, wherein one group of initial character strings corresponds to a preset service link.

[0009] Optionally, the device is configured to perform a target merging operation on multiple groups of initial character strings based on the target prompt word to obtain a target character string by: determining a text similarity parameter corresponding to each group of the initial character strings in the multiple groups of the initial character strings based on the target prompt word; when the target text similarity parameter meets a target threshold, obtaining an intermediate character string corresponding to the target text similarity parameter, wherein the multiple groups of the initial character strings include the intermediate character string; and concatenating the intermediate character strings to obtain the target character string.

[0010] Optionally, the device is used to determine the text similarity parameters corresponding to each group of the initial character strings in the multiple groups of initial character strings based on the target prompt word in the following manner, including at least one of the following: performing a text similarity matching operation on the target prompt word and each group of the initial character strings in the multiple groups of initial character strings in turn to obtain a plurality of the text similarity parameters; performing a semantic vector distance matching operation on the target prompt word in the audit prompt template and each group of the initial character strings in the multiple groups of initial character strings in turn to obtain a plurality of the text similarity parameters.

[0011] Optionally, the device is used to input the target character string into a target business model to generate a target audit report in the following manner: setting the value of the temperature parameter of the target business model to a first value, so as to process the target character string using the target business model and obtain first report data, wherein the first report data is used to generate a data analysis text in the target audit report; setting the value of the temperature parameter of the target business model to a second value, so as to process the target character string using the target business model and obtain second report data, wherein the second report data is used to generate an audit situation text in the target audit report, and the second value is greater than the first value; setting the value of the temperature parameter of the target business model to a third value, so as to process the target character string using the target business model and obtain third report data, wherein the third report data is used to generate an audit risk text in the target audit report, and the third value is greater than the second value.

[0012] According to another aspect of the embodiments of the present application, a computer-readable storage medium is provided, in which a computer program is stored, wherein the computer program is configured to execute the above-mentioned method for generating an audit report when running.

[0013] According to another aspect of the present application, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the above-described method for generating an audit report.

[0014] According to another aspect of an embodiment of the present application, an electronic device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to execute the above-mentioned method for generating an audit report through the computer program.

[0015] In an embodiment of the present application, a method is adopted to obtain first audit data, second audit data, and an audit prompt template, wherein the data format of the first audit data is different from the data format of the second audit data, the type of the first audit data is the same as the data type of the second audit data, and the audit prompt template includes a target prompt word; multiple groups of initial character strings are determined based on the first audit data and the second audit data, and a target merge operation is performed on the multiple groups of initial character strings based on the target prompt word to obtain a target character string; the target character string and the audit prompt template are input into a target business model to generate a target audit report, wherein, by setting different temperature parameters of the target business model to generate different parts of the target audit report, through automated processing and the deep learning capabilities of the intelligent model, rapid integration of audit data and automatic generation of audit reports are achieved, which greatly shortens the time period for report preparation and reduces the workload of auditors. At the same time, the dynamic adjustment of the model temperature parameters ensures that each part of the report can maintain professional rigor while adapting to the expression requirements of different audit topics, realizing the intelligence and efficiency of the audit report generation process, thereby solving the technical problem of low efficiency of manually filling out audit reports. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0017] Figure 1 This is a schematic diagram of an application environment of an optional audit report generation method according to an embodiment of the present application;

[0018] Figure 2 This is a flowchart of an optional method for generating an audit report according to an embodiment of the present application;

[0019] Figure 3 is a schematic diagram of an optional method for generating an audit report according to an embodiment of the present application;

[0020] Figure 4 is a schematic diagram of another optional method for generating an audit report according to an embodiment of the present application;

[0021] Figure 5 is a schematic diagram of another optional method for generating an audit report according to an embodiment of the present application;

[0022] Figure 6 is a schematic diagram of another optional method for generating an audit report according to an embodiment of the present application;

[0023] Figure 7 is a schematic diagram of another optional method for generating an audit report according to an embodiment of the present application;

[0024] Figure 8 is a schematic diagram of another optional method for generating an audit report according to an embodiment of the present application;

[0025] Figure 9 This is a schematic structural diagram of an optional audit report generation device according to an embodiment of the present application;

[0026] Figure 10 This is a schematic diagram of the structure of an optional audit report generation product according to an embodiment of the present application;

[0027] Figure 11 It is a schematic structural diagram of an optional electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0028] In order to enable those skilled in the art to better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this application.

[0029] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in a sequence other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0030] The present application will be described below with reference to the following embodiments:

[0031] According to one aspect of the embodiment of the present application, a method for generating an audit report is provided. Optionally, in this embodiment, the method for generating an audit report can be applied to Figure 1 In the hardware environment composed of the server 101 and the terminal device 103 shown in FIG. Figure 1As shown, the server 101 is connected to the terminal device 103 via a network and can be used to provide services for the terminal device or the application 107 installed on the terminal device. The application can be a video application, instant messaging application, browser application, educational application, game application, etc. A database 105 may be set up on the server or independently of the server to provide data storage services for the server 101, for example, a game data storage server. The above-mentioned network may include but is not limited to: a wired network, a wireless network, wherein the wired network includes: a local area network, a metropolitan area network and a wide area network, and the wireless network includes: Bluetooth, WIFI and other networks that realize wireless communication. The terminal device 103 may be a terminal configured with an application, and may include but is not limited to at least one of the following: a mobile phone (such as an Android phone, an iOS phone, etc.), a laptop computer, a tablet computer, a PDA, a MID (Mobile Internet Devices), a PAD, a desktop computer, a smart TV, an intelligent voice interaction device, a smart home appliance, a vehicle-mounted terminal, an aircraft, a virtual reality (VR) terminal, an augmented reality (AR) terminal, a mixed reality (MR) terminal and other computer devices. The above-mentioned server may be a single server, a server cluster consisting of multiple servers, or a cloud server.

[0032] Combine Figure 1 As shown, the above-mentioned method for generating an audit report can be executed by an electronic device, which can be a terminal device or a server. The above-mentioned method for generating an audit report can be implemented by the terminal device or the server separately, or by the terminal device and the server together.

[0033] The above is only an example and is not specifically limited in this embodiment.

[0034] Alternatively, as an optional implementation, Figure 2 As shown, the method for generating the above audit report includes:

[0035] S202, obtaining first audit data, second audit data, and an audit prompt template, wherein the data format of the first audit data is different from the data format of the second audit data, the type of the first audit data is the same as the data type of the second audit data, and the audit prompt template includes a target prompt word;

[0036] S204, determining multiple groups of initial character strings based on the first audit data and the second audit data, and performing a target merging operation on the multiple groups of initial character strings based on the target prompt word to obtain a target character string;

[0037] S206, inputting the target character string and the audit prompt template into the target business model to generate a target audit report, wherein different temperature parameters of different target business models are set to generate different parts of the target audit report.

[0038] Optionally, in an embodiment of the present application, the first audit data and the second audit data are composed of audit materials from different sources or in different formats, including but not limited to financial statements, management reports, compliance documents, etc., to cover various information sources required for the audit. At the same time, the audit prompt template refers to a structured framework for guiding the large model to generate an audit report, which contains target prompt words. These prompt words are used to identify and classify key information points in the audit data, such as "financial analysis", "risk assessment", "compliance check", etc.

[0039] Optionally, in this embodiment of the present application, the target business model refers to a large-scale language model trained in a specific domain, used to process and generate professional audit reports. Target business models include, but are not limited to, deep learning models and neural network models, and possess powerful natural language understanding and generation capabilities. Based on the input audit data and audit prompt templates, they can adjust temperature parameters to control the structure, style, and content depth of the generated report, thereby ensuring the accuracy and professionalism of the report.

[0040] It should be noted that although the first and second audit data differ in data format, they maintain consistency in data type, ensuring the feasibility of the merge operation and the uniformity of the generated report. This application does not specifically limit the format differences of such data. As long as the data can be converted into a unified text format through appropriate conversion tools or technical means, this method can be applied.

[0041] For example, in an embodiment of the present application, heterogeneous audit data is first acquired through a data input component, converted into string form, and then, based on predefined rules or parameters, the string is segmented into multiple groups of initial strings. When executing a target merge operation, the initial string group that best matches the target prompt word in the audit prompt template is identified, and the target string is formed through intelligent combination or data fusion.

[0042] In an exemplary embodiment, using a financial audit application scenario as an example, the latest financial statements of the audited entity are first obtained as the first audit data, while the financial analysis reports for the past year are collected as the second audit data. After determining multiple sets of initial character strings, a target merge operation is performed based on the target prompt word "financial analysis," integrating character strings related to financial status, revenue trends, cost control, and so on into a set of target character strings. This set of target character strings, along with the audit prompt template, is then input into the pre-trained target business model. By setting appropriate temperature parameters, a rich and clearly structured financial analysis report is generated. This report covers the necessary financial details and analytical conclusions, laying a solid foundation for the complete preparation of the audit report.

[0043] In a specific embodiment, it is assumed that an economic responsibility audit report of a corporate executive needs to be generated, including but not limited to:

[0044] The first audit data: the executive's performance report, which details his decision-making ideas, business results and risk control measures during his tenure, is in the format of a Word document.

[0045] Second audit data: quantitative financial data and operating indicators, such as revenue, profit, cost and market share, provided in Excel form.

[0046] First, the performance report and financial data are obtained through the Word file input component and Excel file input component respectively. At this time, the first audit data (performance report) and the second audit data (financial data) have obvious differences in data format, but they belong to the same audit business data type.

[0047] The long text segmentation component is used to segment the performance report into several first substrings, each containing an independent description of management decisions. Financial data and operating indicators are also converted into second substrings, ensuring that each data description is associated with a specific financial indicator or operating result.

[0048] Next, based on the target prompt words in the audit prompt template, such as "revenue recognition", "cost control" and "compliance check", multiple first substrings and second substrings are traversed and analyzed and vector searched to identify the substring group most relevant to the target prompt word. Data cleaning and expert knowledge embedding are performed to ensure that the final merged string content is both accurate and comprehensive.

[0049] Finally, the target string obtained through the target merging operation, together with the audit prompt template, is input into the pre-trained target business model. The model temperature parameter plays a key role. Different temperature values ​​are set according to different parts of the audit report, such as background description, financial analysis, risk assessment, and audit recommendations.

[0050] For example, a lower temperature value (such as 0.3) is used in the background description section to ensure that the generated text is strictly based on the input data and the language expression is accurate and correct; while in the audit recommendation section, a higher temperature value (such as 0.6) is used to enhance the creativity and richness of the generated content and provide a more multidimensional perspective and strategic recommendations for the audit work.

[0051] Through the embodiment of the present application, the first audit data, the second audit data and the audit prompt template are obtained, wherein the data format of the first audit data and the data format of the second audit data are different, the type of the first audit data and the data type of the second audit data are the same, and the audit prompt template includes a target prompt word; multiple groups of initial character strings are determined based on the first audit data and the second audit data, and a target merge operation is performed on the multiple groups of initial character strings based on the target prompt word to obtain a target character string; the target character string and the audit prompt template are input into the target business model to generate a target audit report, wherein, by setting different temperature parameters of the target business model to generate different parts of the target audit report, through automated processing and the deep learning capabilities of the intelligent model, the rapid integration of audit data and the automatic generation of audit reports are achieved, which greatly shortens the time period for report preparation and reduces the workload of auditors. At the same time, the dynamic adjustment of the model temperature parameters ensures that each part of the report can maintain professional rigor and adapt to the expression requirements of different audit topics, realizes the intelligence and efficiency of the audit report generation process, and thus solves the technical problem of low efficiency of manually filling out audit reports.

[0052] As an optional solution, the above-mentioned determining multiple groups of initial character strings based on the above-mentioned first audit data and the above-mentioned second audit data includes: converting the above-mentioned first audit data into a first character string, and converting the above-mentioned second audit data into a second character string; obtaining a target quantity in response to a target interaction operation; dividing the above-mentioned first character string into multiple first substrings, and dividing the above-mentioned second character string into multiple second substrings, wherein the sum of the number of the above-mentioned first substrings and the number of the above-mentioned second substrings is the above-mentioned target quantity; generating multiple groups of the above-mentioned initial character strings based on the multiple first substrings and the multiple second substrings.

[0053] Optionally, in an embodiment of the present application, the target interaction operation refers to parameters set or adjusted by the user through a human-computer interface, including but not limited to a target quantity, that is, the total number of substrings that need to be generated.

[0054] It should be noted that various technical means can be employed during the conversion between the first and second strings, including but not limited to natural language processing (NLP) techniques for text parsing or the extraction of key numerical information using data mining algorithms. The length, content integrity, and processing efficiency of the converted strings can be flexibly adjusted based on the requirements of the specific audit task, and this application does not impose any restrictions on this.

[0055] Exemplarily, the first audit data and the second audit data are first converted into a first string and a second string, respectively. Then, in response to a target interaction operation, a target number is determined, and based on this number, substring division of the first and second strings is performed. The first string is divided into a predetermined number of first substrings, each focusing on a different content paragraph of the performance report; the second string is correspondingly divided into multiple second substrings, each containing a different aspect of the financial data.

[0056] In an exemplary embodiment, the economic responsibility audit of a company's personnel is used as an application scenario, and the company's personnel's performance reports (first audit data) and the company's annual financial statements (second audit data) are obtained. First, the performance reports and financial statements are converted into first and second character strings, respectively. The user sets the target number to 20 in the target interaction operation to ensure the comprehensiveness of the report. Next, the system divides the first character string into 10 first substrings, each of which focuses on the description of different business areas in the report, such as strategic execution, market development, and project management; at the same time, the second character string is divided into 10 second substrings, which correspond to the core aspects of financial indicators such as revenue, profit, cost, and cash flow. Ultimately, these 20 groups of substrings constitute the initial character string set, providing an information source for report generation.

[0057] Through the embodiments of the present application, the target quantity-oriented substring division is adopted according to the character string converted from the audit data, thereby achieving the technical effect of automatic generation of audit reports, achieving the purpose of efficiently processing heterogeneous audit information, improving the speed of audit report generation, and preserving data integrity and comprehensiveness of audit content.

[0058] As an optional solution, the generating of multiple groups of the above-mentioned initial character strings based on the multiple first substrings and the multiple second substrings includes at least one of the following: combining the multiple first substrings and the multiple second substrings in chronological order to generate multiple groups of the above-mentioned initial character strings, wherein a group of the above-mentioned initial character strings corresponds to a preset time interval; combining the multiple first substrings and the multiple second substrings in a business execution order to generate multiple groups of the above-mentioned initial character strings, wherein a group of the above-mentioned initial character strings corresponds to a preset business link.

[0059] Optionally, in an embodiment of the present application, the preset time interval refers to a time period division set according to audit business requirements, used to organize and combine the first substring and the second substring, including but not limited to annual, quarterly, monthly, etc., to ensure that the generated initial string can reflect the economic responsibility audit information within a specific time range.

[0060] Similarly, the preset business link refers to the combination strategy of the first substring and the second substring based on the process sequence of the enterprise's business activities, such as market research, product development, sales execution, after-sales service, etc., to ensure that the initial string covers the performance and data of the audit objectives at different business stages. This application does not make specific restrictions on this and allows for flexible adjustment of the time interval and business link settings based on the specific audit objectives and enterprise characteristics.

[0061] Exemplarily, the first and second audit data are received and processed, converted into first and second character strings, and then these two types of character strings are segmented to obtain multiple first substrings and multiple second substrings. The substrings are then combined according to a preset time sequence or business execution sequence. In the time sequence combination, the first substring in the performance report is paired with the second substring in the financial data according to key time nodes on the timeline, ensuring that each set of initial strings reflects the company's economic activities and management results within a specific time period. In the business sequence combination, substrings corresponding to various stages in the company's value chain are identified and combined to ensure that the content of the audit report covers all key aspects of the company's operations.

[0062] In an exemplary embodiment, taking the application scenario of an enterprise's economic responsibility audit as an example, the performance report (first audit data) and annual financial statements (second audit data) of the enterprise's senior executives are obtained. First, the enterprise system divides the performance report into four first substrings in chronological order, each substring describing the management activities of a quarter; at the same time, the financial statement data is decomposed into four second substrings by quarter. Then, the enterprise system combines the first substrings and the second substrings in chronological order to generate four groups of initial strings, each group of strings corresponding to the comprehensive audit information of a quarter. In another embodiment, the enterprise system combines the first substring about product development in the performance report with the second substring about R&D expenditure in the financial data in the order of business execution to generate an initial string for R&D activities. Similarly, other business links are also combined accordingly to generate an audit report that comprehensively reflects the implementation of the enterprise's economic responsibilities.

[0063] Through the embodiments of this application, by combining substrings in chronological order or by business execution order, the generation of audit reports is organized and targeted, accurately reflecting the auditee's fulfillment of its economic responsibilities within a specific time period or business phase. This strategy ensures a clear structure and comprehensive content for audit reports, while providing auditors with a basis for in-depth analysis and effectively improving the professionalism and efficiency of audit work.

[0064] As an optional solution, the above-mentioned target merging operation is performed on multiple groups of the above-mentioned initial character strings based on the above-mentioned target prompt word to obtain the target character string, including: determining the text similarity parameters corresponding to each group of the above-mentioned initial character strings in the multiple groups of the above-mentioned initial character strings based on the above-mentioned target prompt word; when the target text similarity parameters meet the target threshold, obtaining the intermediate character strings corresponding to the above-mentioned target text similarity parameters, wherein the multiple groups of the above-mentioned initial character strings include the above-mentioned intermediate character strings; and splicing the above-mentioned intermediate character strings to obtain the above-mentioned target character string.

[0065] Optionally, in this embodiment of the present application, the text similarity parameter refers to a quantitative indicator that measures the degree of semantic similarity between the initial character string and the target prompt word, including but not limited to similarity scores calculated based on advanced models such as word vectors, TF-IDF, or BERT. These parameters are used to determine which initial character strings are most relevant to a specific audit topic, thereby effectively screening information.

[0066] It should be noted that the target threshold can be set flexibly and diversely, depending on the level of detail required for the audit report, the audit focus, and the preferences of the auditors. For example, a higher threshold can ensure that the merged text information is highly relevant, while a lower threshold may include a wider range of information, providing a more comprehensive audit perspective. This application does not impose specific numerical limits on the target threshold; its setting should be based on audit objectives and reporting requirements.

[0067] For example, the present embodiment first calculates text similarity parameters between each set of initial character strings and target prompt words, such as "profit analysis," "investment decision," or "cost control." These parameters are then checked to see if they meet a preset target threshold, and the initial character strings that meet the criteria are identified as intermediate character strings. Finally, all intermediate character strings are concatenated into a single entity, forming a final target character string that is closely related to the target prompt word. This final target character string serves as the basis for the input of the larger model, which is then used to automatically generate the audit report.

[0068] In an exemplary embodiment, using the application scenario of a bank's economic responsibility audit as an example, the system acquired multiple first and second substrings related to the bank's operations. The target prompt was set to "credit risk." By calculating the text similarity parameter between each set of substrings and "credit risk," the system identified intermediate strings that detailed loan strategies, changes in non-performing loan ratios, and risk management measures. After ensuring that the text similarity parameters of all intermediate strings exceeded the target threshold, these strings were concatenated to generate a target string containing data and descriptions related to the bank's credit risk. This provided accurate input for the large model used to generate the audit report portion of the credit risk analysis.

[0069] Through the embodiments of the present application, text similarity parameter calculation and intermediate character string screening based on target prompt words and preset thresholds are adopted to achieve the technical effect of accurate information positioning and integration in the automatic generation of audit reports, thereby achieving the purpose of improving the professionalism and pertinence of audit reports, reducing interference from irrelevant information, optimizing the overall structure of reports and improving reading experience, thereby improving the efficiency and effectiveness of the entire audit process.

[0070] As an optional solution, the above-mentioned determination of the text similarity parameters corresponding to each of the multiple groups of initial character strings based on the above-mentioned target prompt words includes at least one of the following: performing text similarity matching operations on the above-mentioned target prompt words and each of the multiple groups of initial character strings in turn to obtain multiple text similarity parameters; performing semantic vector distance matching operations on the above-mentioned target prompt words in the above-mentioned audit prompt template and each of the multiple groups of initial character strings in turn to obtain multiple text similarity parameters.

[0071] Optionally, in an embodiment of the present application, the text similarity matching operation refers to comparing the semantic association between the target prompt word and each group of initial character strings through an algorithm, and specific implementation methods include but are not limited to using term frequency-inverse document frequency (TF-IDF), cosine similarity, etc.

[0072] It should be noted that audit prompt templates can be customized based on different audit objectives and enterprise characteristics, and target prompt words can be pre-set by auditors based on audit priorities or automatically generated by the system based on big data analysis. The calculation method and matching strategy of text similarity parameters can be flexibly selected based on actual needs and are not limited by this application.

[0073] For example, the target prompt word is first matched against each set of initial strings. A text similarity parameter is calculated to quantify the semantic relevance between the initial strings and the audit topic. If cosine similarity matching is used, the target prompt word is converted into a vector representation and compared with the vector representation of each set of initial strings. This yields multiple text similarity parameters that reflect the degree of match between the initial strings and the audit topic.

[0074] In an exemplary embodiment, taking the application scenario of an economic responsibility audit of a manufacturing enterprise as an example, the auditor set "production cost efficiency" as the target prompt word. Through a text similarity matching operation, the similarity between the target prompt word and multiple collected first substrings (related to production processes, cost control, etc.) and second substrings (including specific financial data) was calculated. In this process, a deep learning model was used to quantify the relationship between each substring and the target prompt word based on the semantic vector distance matching strategy, and a series of text similarity parameters were obtained. These parameters were then used to filter content that was highly relevant to "production cost efficiency" to ensure that the generated audit report section accurately reflected the actual cost control situation of the enterprise.

[0075] Through the embodiments of the present application, a text similarity matching operation based on target prompt words is adopted to achieve accurate identification and efficient screening of audit information, thereby achieving the purpose of improving the accuracy and professionalism of automatic generation of audit reports, effectively avoiding the interference of irrelevant information, ensuring that the content of the audit report is closely centered on the audit theme, and meeting the actual needs of the audit work.

[0076] As an optional solution, the above-mentioned input of the target character string into the target business model to generate a target audit report includes: setting the value of the above-mentioned temperature parameter of the above-mentioned target business model to a first value, so as to use the above-mentioned target business model to process the above-mentioned target character string to obtain first report data, wherein the above-mentioned first report data is used to generate the data analysis text in the above-mentioned target audit report; setting the value of the above-mentioned temperature parameter of the above-mentioned target business model to a second value, so as to use the above-mentioned target business model to process the above-mentioned target character string to obtain second report data, wherein the above-mentioned second report data is used to generate the audit situation text in the above-mentioned target audit report, and the above-mentioned second value is greater than the above-mentioned first value; setting the value of the above-mentioned temperature parameter of the above-mentioned target business model to a third value, so as to use the above-mentioned target business model to process the above-mentioned target character string to obtain third report data, wherein the above-mentioned third report data is used to generate the audit risk text in the above-mentioned target audit report, and the above-mentioned third value is greater than the above-mentioned second value.

[0077] For example, the temperature parameter of the target business model is set to a value between 0.3 and 0.6. For example, a low temperature value (0.3) can effectively reduce the randomness of the generated content, ensuring that the model focuses on accurate statements with high probability and avoids subjective speculation or vague expressions. A high temperature value (0.6) can enrich the diversity of the text and improve the readability of the report.

[0078] In an exemplary embodiment, the first value is 0.3, which ensures that core content such as financial data and change trends are completely generated based on input information to avoid fictitious or biased information.

[0079] In an exemplary embodiment, the second value is greater than 0.3 and less than or equal to 0.5, maintaining language neutrality to meet the objectivity requirements of the economic responsibility report.

[0080] In an exemplary embodiment, the third value is set to 0.6 to enhance the creativity of the model and improve the richness of audit recommendations, while the model Top-k and Top-p adopt the model default values.

[0081] In an exemplary embodiment, considering that the traditional economic responsibility audit process usually includes multiple steps such as data collection, data analysis, problem identification, and report writing, the entire process is not only time-consuming, but also relies on the professional judgment and experience of auditors, and has limitations in understanding and analyzing complex indicator data analysis, and often requires deep manual intervention. In order to improve the large model's understanding of complex performance reports and indicator data, the model's intelligent analysis capabilities are enhanced through deep learning and natural language processing technologies, and key text information is extracted and summarized in an automated manner; existing automatic audit report systems usually use fixed templates to generate reports, lack adaptability to the characteristics of different organizations and industries, and cannot meet personalized audit needs. Based on the above-mentioned audit report generation method, the embodiment of the present application has developed a method for generating customized reports based on the specific requirements of different enterprises or industries, so that the reports are more in line with the actual audit situation and increase the relevance and practicality of the reports.

[0082] Furthermore, the existing system lacks user interaction. Automatically generating reports using large models is time-consuming and laborious, and it struggles to meet the requirements of responsible audit reports. To improve the system's interactivity and ease of use, users only need to upload documents and enter a few required fields according to prompts. Without requiring model training, multiple audit reports are automatically generated for auditors' reference, enhancing trust and acceptance.

[0083] To sum up, in order to overcome the shortcomings of the existing technology in terms of intelligent analysis capabilities, personalized report generation, interactivity and operability, the embodiment of the present application greatly improves the generation efficiency and quality of economic responsibility audit reports by utilizing the target business model, making it more adaptable to the needs of actual audit work.

[0084] For example, the target business model described above can be a large model. A large model refers to an important method or technology in the fields of machine learning and artificial intelligence for processing large-scale data and complex models. It is typically characterized by large datasets and complex model structures, aiming to improve the model's accuracy and generalization capabilities. While there is currently no unified official definition, large models generally refer to deep learning models with tens or even hundreds of millions of parameters. These models require massive amounts of data and computing power during training and are widely used in fields such as natural language processing, image recognition, financial forecasting, medical imaging analysis, and autonomous driving.

[0085] Specifically, the embodiments of the present application can solve the problems of traditional economic responsibility audit report preparation being time-consuming, inefficient, and highly subjective, and improve the quality and efficiency of audit work by building an efficient and automated economic responsibility audit report generation tool.

[0086] For example, first, by performing data cleaning on the recalled chunks, key indicators and content that need to be analyzed are extracted to avoid redundant data contaminating the output content of the large model, while adding expert knowledge to ensure the accuracy and relevance of the output content.

[0087] In an exemplary embodiment, Figure 3 This is a schematic diagram of an optional method for generating an audit report according to an embodiment of the present application, such as Figure 3 As shown, it includes data acquisition module, data preprocessing module, text traversal analysis module and report generation module.

[0088] S1, data acquisition module, Figure 4 This is a schematic diagram of another optional method for generating an audit report according to an embodiment of the present application, such as Figure 4 As shown, the required data is collected from multiple sources such as the auditee's performance report, the auditee's institution annual summary, previous audit findings, and economic responsibility indicator tables, and the authenticity and completeness of the data are ensured.

[0089] S2, data preprocessing module, Figure 5 This is a schematic diagram of another optional method for generating an audit report according to an embodiment of the present application, such as Figure 5As shown, it consists of a Word file input component / PDF file input component / Excel file input component and a long text segmentation component. It is responsible for converting performance reports, economic and responsibility indicators and other materials in different formats into character strings, and then segmenting the character strings into text segments that can be supported by the large model.

[0090] S3, text retrieval module:

[0091] Mode 1: Figure 6 This is a schematic diagram of another optional method for generating an audit report according to an embodiment of the present application, such as Figure 6 As shown, text traversal retrieval consists of a large model component (the target business model described above), a segment text traversal component, and an audit prompt template component. It is responsible for traversing the segmented text segments and extracting the target paragraphs as the strings to be processed. This module combines natural language understanding and generation technologies to identify the target paragraphs (the target strings described above) based on prompt word constraints.

[0092] Mode 2: Figure 7 This is a schematic diagram of another optional method for generating an audit report according to an embodiment of the present application, such as Figure 7 As shown, text vector retrieval, consisting of a large model component, text vector retrieval, and an audit prompt template component, is responsible for extracting target paragraphs as strings to be processed based on vector distance from the segmented text. This module uses semantic vector distance for judgment, enabling faster target paragraph identification compared to mode 1. Furthermore, data cleaning is performed on the target paragraphs to remove content not needed for large model analysis and prevent data contamination.

[0093] S4, report generation module, Figure 8 This is a schematic diagram of another optional method for generating an audit report according to an embodiment of the present application, such as Figure 8 As shown, it consists of a large model component, an audit prompt template component, a text input executor component, and an output to Word template component. It is responsible for processing the character string to be processed according to the audit prompt template structure and passing it to the large model. The large model analyzes and processes it and then outputs it to the Word template to ensure the integrity and readability of the report.

[0094] When performing text analysis, prompt words use a few-shot prompt word structure. By adding structural examples of the output content in the prompt words, the big model can understand the output structure and content composition. At the same time, expert knowledge is embedded in the prompt words. For example, "Improving digital business capabilities mainly includes fintech-related content" allows the big model to accurately analyze the specified dimensions, thereby improving the accuracy and relevance of the model output content.

[0095] Setting the temperature parameter of the target business model to a value between 0.3 and 0.6, for example, a low temperature (0.3) can effectively reduce the randomness of generated content, ensuring that the model focuses on high-probability accurate representations and avoids subjective assumptions or vague expressions. A high temperature (0.6) can enrich the diversity of textual expression and improve the readability of the report.

[0096] In an exemplary embodiment, the first value is 0.3, which ensures that core content such as financial data and change trends are completely generated based on input information to avoid fictitious or biased information.

[0097] In an exemplary embodiment, the second value is greater than 0.3 and less than or equal to 0.5, maintaining language neutrality to meet the objectivity requirements of the economic responsibility report.

[0098] In an exemplary embodiment, the third value is set to 0.6 to enhance the creativity of the model and improve the richness of audit recommendations, while the model Top-k and Top-p adopt the model default values.

[0099] This embodiment of the application only requires the client to provide a performance report and business indicator data, as well as input a small amount of text information. Using the large model, an audit report that meets the requirements of a responsible audit report can be generated. The generation process is simple, easy to operate, and provides a strong user experience. By selecting appropriate, specific, professional, and accurate prompt words, continuous optimization and adjustment are carried out to guide and train the large model, improve the accuracy and relevance of the model results, and generate a clear, logical, and audit-compliant audit report.

[0100] By introducing algorithms specific to financial accountability audits, the performance of large models in processing complex audit texts has been significantly improved, reducing errors in reports and enhancing their professionalism and credibility. By enhancing the understanding of enterprise- and industry-specific rules, the system's customization capabilities have been strengthened. Users can customize the constraints for change prompts based on their own audit reporting requirements, ensuring that the generated audit reports are more tailored to the actual needs of different enterprises.

[0101] In addition to the large-scale model-based technical solution, there are other alternative solutions that can achieve the same or similar goals, namely, improving the efficiency and quality of generating economic responsibility audit reports and reducing reliance on auditors' subjective judgment. This application does not limit these solutions, and these alternative solutions can be approached from different perspectives, such as adopting different technical paths, improving the application of existing technologies, or combining advanced technologies from other fields to optimize the audit report generation process.

[0102] For example, hybrid AI models can be employed, combining rule-based systems (RBS) and machine learning (ML). RBS can set fixed rules based on audit standards, while ML models handle unstructured data and complex analysis. This combination provides more comprehensive coverage of various audit scenarios and improves the accuracy and professionalism of reports.

[0103] Another example is the synergy between reinforcement learning and expert systems. A model trained using reinforcement learning (RL) is combined with an expert system. The reinforcement learning model can continuously attempt to generate reports in a simulated environment, gradually optimizing its performance through a feedback mechanism. The expert system provides support for domain-specific knowledge, such as audit rules and industry best practices. This approach leverages the power of machine learning while also leveraging the knowledge of human experts to improve the accuracy and applicability of reports.

[0104] Another example is the use of cloud computing and distributed processing. Given the massive amount of data processing, computing tasks can be deployed to the cloud, leveraging the powerful computing power and storage resources of cloud computing. Distributed processing accelerates data preprocessing and analysis while ensuring system scalability and reliability. Furthermore, cloud computing platforms can easily integrate a variety of large-scale model services and tools, further improving the efficiency and quality of report generation.

[0105] For example, by employing cross-domain data fusion and analysis, the aforementioned first and second audit data not only incorporate business data from the auditee's institutions but also integrate cross-domain data such as market data. Processing this diverse data through deep learning models enables a more comprehensive understanding of a company's operational status and potential risks, resulting in richer and more accurate audit reports.

[0106] Furthermore, improving the human-computer interaction interface and designing a more intuitive and efficient report editing and review tool can significantly improve audit efficiency. By using features like intelligent prompts and automated proofreading, auditors’ workload can be reduced and reporting accuracy and speed can be improved.

[0107] Through the embodiments of the present application, the degree of automation is high, and the large model can automatically identify and process large amounts of data, reducing the manual burden; through deep learning of historical data, the model can intelligently identify trends, discover anomalies, and conduct cause analysis; it can also generate reports with specific formats and contents according to different audit needs to meet diverse needs; it has strong natural language processing capabilities, can understand complex audit indicators and expression requirements, automatically generate standardized and rigorous audit report texts, and use artificial intelligence technology to improve the accuracy and efficiency of audit work. The full process automation from data collection and analysis to report generation is realized, which greatly reduces the audit cost. The use of natural language processing technology enables machines to understand and generate highly professional audit reports, filling the technical gap in this field.

[0108] In summary, the embodiments of the present application not only improve the efficiency of audit work, but also enhance the accuracy of audit, while expanding the application scope of artificial intelligence technology in different business fields, improving accuracy and professionalism, enhancing personalized customization functions, realizing a rapid response mechanism, and improving the work efficiency of auditors.

[0109] It is understandable that in the specific implementation of this application, related data such as user information is involved. When the above embodiments of this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.

[0110] It should be noted that for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that this application is not limited by the order of the actions described, because according to this application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by this application.

[0111] According to another aspect of the embodiment of the present application, there is also provided an audit report generation device for implementing the above-mentioned audit report generation method. Figure 9 As shown, the device includes:

[0112] Acquisition module 902 is configured to acquire first audit data, second audit data, and an audit prompt template, wherein the data format of the first audit data is different from the data format of the second audit data, the type of the first audit data is the same as the data type of the second audit data, and the audit prompt template includes a target prompt word;

[0113] An execution module 904 is configured to determine multiple groups of initial character strings based on the first audit data and the second audit data, and perform a target merging operation on the multiple groups of initial character strings based on the target prompt word to obtain a target character string;

[0114] The generation module 906 is used to input the above-mentioned target character string and the above-mentioned audit prompt template into the target business model to generate a target audit report, wherein different temperature parameters of the above-mentioned target business model are set to generate different parts of the above-mentioned target audit report.

[0115] As an optional solution, the above-mentioned device is used to determine multiple groups of initial character strings based on the above-mentioned first audit data and the above-mentioned second audit data in the following manner: converting the above-mentioned first audit data into a first character string, and converting the above-mentioned second audit data into a second character string; obtaining a target quantity in response to a target interaction operation; dividing the above-mentioned first character string into multiple first substrings, and dividing the above-mentioned second character string into multiple second substrings, wherein the sum of the number of the above-mentioned first substrings and the number of the above-mentioned second substrings is the above-mentioned target quantity; generating multiple groups of the above-mentioned initial character strings based on the multiple first substrings and the multiple second substrings.

[0116] As an optional solution, the apparatus is configured to generate multiple groups of the initial character strings based on multiple first substrings and multiple second substrings in the following manner, including at least one of the following: combining multiple first substrings and multiple second substrings in chronological order to generate multiple groups of the initial character strings, wherein one group of the initial character strings corresponds to a preset time interval; or combining multiple first substrings and multiple second substrings in a service execution order to generate multiple groups of the initial character strings, wherein one group of the initial character strings corresponds to a preset service link.

[0117] As an optional solution, the above-mentioned device is used to perform a target merging operation on multiple groups of the above-mentioned initial character strings based on the above-mentioned target prompt word in the following manner to obtain the target character string: determine the text similarity parameter corresponding to each group of the above-mentioned initial character strings in the multiple groups of the above-mentioned initial character strings based on the above-mentioned target prompt word; when the target text similarity parameter meets the target threshold, obtain the intermediate character string corresponding to the above-mentioned target text similarity parameter, wherein the multiple groups of the above-mentioned initial character strings include the above-mentioned intermediate character string; splice the above-mentioned intermediate character strings to obtain the above-mentioned target character string.

[0118] As an optional solution, the above-mentioned device is used to determine the text similarity parameters corresponding to each group of the above-mentioned initial character strings in the multiple groups of the above-mentioned initial character strings based on the above-mentioned target prompt words in the following manner, including at least one of the following: performing text similarity matching operations on the above-mentioned target prompt words and each group of the above-mentioned initial character strings in the multiple groups of the above-mentioned initial character strings in turn to obtain multiple of the above-mentioned text similarity parameters; performing semantic vector distance matching operations on the above-mentioned target prompt words in the above-mentioned audit prompt template and each group of the above-mentioned initial character strings in the multiple groups of the above-mentioned initial character strings in turn to obtain multiple of the above-mentioned text similarity parameters.

[0119] As an optional solution, the above-mentioned device is used to input the above-mentioned target character string into the target business model to generate a target audit report in the following manner: setting the value of the above-mentioned temperature parameter of the above-mentioned target business model to a first value, so as to use the above-mentioned target business model to process the above-mentioned target character string and obtain first report data, wherein the above-mentioned first report data is used to generate the data analysis text in the above-mentioned target audit report; setting the value of the above-mentioned temperature parameter of the above-mentioned target business model to a second value, so as to use the above-mentioned target business model to process the above-mentioned target character string and obtain second report data, wherein the above-mentioned second report data is used to generate the audit situation text in the above-mentioned target audit report, and the above-mentioned second value is greater than the above-mentioned first value; setting the value of the above-mentioned temperature parameter of the above-mentioned target business model to a third value, so as to use the above-mentioned target business model to process the above-mentioned target character string and obtain third report data, wherein the above-mentioned third report data is used to generate the audit risk text in the above-mentioned target audit report, and the above-mentioned third value is greater than the above-mentioned second value.

[0120] In the embodiments of the present application, the term "module" or "unit" refers to a computer program or a part of a computer program that has a predetermined function and works together with other related parts to achieve a predetermined goal, and can be implemented in whole or in part by using software, hardware (such as processing circuits or memories) or a combination thereof. Similarly, a processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be part of an overall module or unit that includes the function of the module or unit.

[0121] Regarding the apparatus in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0122] According to one aspect of the present application, a computer program product is provided, which includes a computer program.

[0123] The serial numbers of the above embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0124] Figure 10 The block diagram schematically shows a computer system structure of an electronic device used to implement an embodiment of the present application.

[0125] It should be noted that Figure 10 The computer system 1000 of the electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0126] like Figure 10 As shown, the computer system 1000 includes a central processing unit (CPU) 1001, which can perform various appropriate actions and processes according to the program stored in the read-only memory 1002 (ROM) or the program loaded from the storage part 1008 into the random access memory 1003 (RAM). Various programs and data required for system operation are also stored in the random access memory 1003. The CPU 1001, the read-only memory 1002, and the random access memory 1003 are connected to each other via a bus 1004. An input / output interface 1005 (i.e., an I / O interface) is also connected to the bus 1004.

[0127] The following components are connected to the input / output interface 1005: an input section 1006 including a keyboard, a mouse, and the like; an output section 1007 including devices such as a cathode ray tube (CRT), a liquid crystal display (LCD), and a speaker; a storage section 1008 including a hard disk; and a communication section 1009 including a network interface card such as a local area network card or a modem. The communication section 1009 performs communication processing via a network such as the Internet. A drive 1010 is also connected to the input / output interface 1005 as needed. Removable media 1011, such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, is installed in the drive 1010 as needed, so that computer programs read therefrom can be installed into the storage section 1008 as needed.

[0128] In particular, according to an embodiment of the present application, the processes described in the various method flow charts can be implemented as computer software programs. For example, an embodiment of the present application includes a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for executing the methods shown in the flow charts. In such an embodiment, the computer program can be downloaded and installed from a network via the communication portion 1009 and / or installed from a removable medium 1011. When the computer program is executed by the central processing unit 1001, the various functions defined in the system of the present application are performed.

[0129] In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 1009, and / or installed from the removable medium 1011. When the computer program is executed by the central processing unit 1001, various functions provided by the embodiments of the present application are performed.

[0130] According to another aspect of the embodiment of the present application, an electronic device for implementing the above-mentioned method for generating an audit report is also provided. The electronic device may be Figure 1 The terminal device or server shown in FIG. This embodiment is described by taking the electronic device as a terminal device as an example. Figure 11 As shown, the electronic device includes a memory 1102 and a processor 1104. The memory 1102 stores a computer program, and the processor 1104 is configured to execute the steps in any of the above method embodiments through the computer program.

[0131] Optionally, in this embodiment, the electronic device may be located in at least one network device among a plurality of network devices of a computer network.

[0132] Optionally, in this embodiment, the above-mentioned processor can be configured to execute the methods in each embodiment of the present application through a computer program.

[0133] Alternatively, those skilled in the art will appreciate that Figure 11 The structure shown is for illustration only. Figure 11 It does not limit the structure of the above electronic device. For example, the electronic device may also include Figure 11 More or fewer components (such as network interfaces, etc.) as shown in, or with Figure 11 Different configurations shown.

[0134] Among them, the memory 1102 can be used to store software programs and modules, such as the program instructions / modules corresponding to the method and device for generating an audit report in the embodiment of the present application. The processor 1104 executes various functional applications and data processing by running the software programs and modules stored in the memory 1102, that is, realizes the above-mentioned method for generating an audit report. The memory 1102 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 1102 may further include a memory remotely located relative to the processor 1104, and these remote memories may be connected to the terminal via a network. Examples of the above-mentioned networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. Among them, the memory 1102 can be used to store, but is not limited to, first audit data, second audit data and other information. As an example, if Figure 11 As shown, the memory 1102 may include, but is not limited to, the acquisition module 902, execution module 904, and generation module 906 of the audit report generation device. In addition, it may also include, but is not limited to, other module units of the audit report generation device, which will not be repeated in this example.

[0135] Optionally, the transmission device 1106 is configured to receive or send data via a network. Specific examples of the network may include a wired network and a wireless network. In one embodiment, the transmission device 1106 includes a network interface controller (NIC), which can be connected to other network devices and a router via a network cable to communicate with the Internet or a local area network. In one embodiment, the transmission device 1106 is a radio frequency (RF) module, which is configured to communicate with the Internet wirelessly.

[0136] In addition, the electronic device further includes: a display 1108 for displaying the target audit data; and a connection bus 1110 for connecting the various module components in the electronic device.

[0137] In other embodiments, the terminal device or server may be a node in a distributed system, wherein the distributed system may be a blockchain system, and the blockchain system may be a distributed system formed by connecting multiple nodes via network communication. The nodes may form a peer-to-peer network, and any computing device, such as a server, terminal, or other electronic device, may become a node in the blockchain system by joining the peer-to-peer network.

[0138] According to one aspect of the present application, a computer-readable storage medium is provided, and a processor of an electronic device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the electronic device executes the audit report generation method provided in various optional implementation methods of the above-mentioned audit report generation.

[0139] Optionally, in this embodiment, the above-mentioned computer-readable storage medium can be configured to store data for executing the methods in various embodiments of the present application.

[0140] Optionally, in this embodiment, a person of ordinary skill in the art may understand that all or part of the steps in the various methods of the above embodiments may be completed by instructing the hardware related to the terminal device through a program, and the program may be stored in a computer-readable storage medium, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0141] The serial numbers of the above embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0142] If the integrated units in the above embodiments are implemented in the form of software functional units and sold or used as independent products, they can be stored in the above-mentioned computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for causing one or more electronic devices to execute all or part of the steps of the method described in each embodiment of the present application.

[0143] In the above embodiments of the present application, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0144] In the several embodiments provided in this application, it should be understood that the disclosed applications can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0145] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0146] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0147] The above is only a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for generating an audit report, characterized in that: include: Acquire first audit data, second audit data, and an audit prompt template, wherein the data format of the first audit data is different from the data format of the second audit data, the type of the first audit data is the same as the data type of the second audit data, and the audit prompt template includes a target prompt word; determining a plurality of groups of initial character strings based on the first audit data and the second audit data, and performing a target merging operation on the plurality of groups of initial character strings based on the target prompt word to obtain a target character string; The target character string and the audit prompt template are input into the target business model to generate a target audit report, wherein different temperature parameters of the target business model are set to generate different parts of the target audit report.

2. The method according to claim 1, characterized in that The determining of multiple groups of initial character strings according to the first audit data and the second audit data includes: converting the first audit data into a first string, and converting the second audit data into a second string; In response to the target interaction operation, obtaining the target quantity; dividing the first character string into a plurality of first substrings, and dividing the second character string into a plurality of second substrings, wherein the sum of the number of the first substrings and the number of the second substrings is the target number; A plurality of groups of the initial character strings are generated according to a plurality of the first sub-character strings and a plurality of the second sub-character strings.

3. The method according to claim 2, characterized in that Generating a plurality of groups of the initial character strings according to the plurality of the first substrings and the plurality of the second substrings comprises at least one of the following: combining a plurality of the first substrings and a plurality of the second substrings in time order to generate a plurality of groups of the initial character strings, wherein a group of the initial character strings corresponds to a preset time interval; Multiple first substrings and multiple second substrings are combined in a service execution order to generate multiple groups of initial strings, wherein one group of initial strings corresponds to a preset service link.

4. The method according to claim 1, wherein The step of performing a target merging operation on the multiple groups of initial character strings based on the target prompt word to obtain a target character string includes: Determining a text similarity parameter corresponding to each group of the initial character strings in the plurality of groups of the initial character strings based on the target prompt word; When the target text similarity parameter satisfies a target threshold, obtaining an intermediate character string corresponding to the target text similarity parameter, wherein the plurality of groups of initial character strings include the intermediate character string; The intermediate character strings are concatenated to obtain the target character string.

5. The method according to claim 4, characterized in that The determining of the text similarity parameter corresponding to each of the multiple groups of initial character strings based on the target prompt word includes at least one of the following: Performing a text similarity matching operation on the target prompt word and each of the multiple groups of the initial character strings in sequence to obtain a plurality of the text similarity parameters; The target prompt word in the audit prompt template is sequentially matched with each group of the initial character strings by a semantic vector distance matching operation to obtain a plurality of the text similarity parameters.

6. The method according to claim 1, characterized in that The step of inputting the target character string into a target business model and generating a target audit report includes: Setting the value of the temperature parameter of the target business model to a first value, so as to process the target character string using the target business model to obtain first report data, wherein the first report data is used to generate data analysis text in the target audit report; Setting the value of the temperature parameter of the target business model to a second value, so as to process the target character string using the target business model to obtain second report data, wherein the second report data is used to generate the audit status text in the target audit report, and the second value is greater than the first value; The value of the temperature parameter of the target business model is set to a third value to process the target character string using the target business model to obtain third report data, wherein the third report data is used to generate the audit risk text in the target audit report, and the third value is greater than the second value.

7. A device for generating an audit report, characterized in that: include: an acquisition module, configured to acquire first audit data, second audit data, and an audit prompt template, wherein the data format of the first audit data is different from the data format of the second audit data, the type of the first audit data is the same as the data type of the second audit data, and the audit prompt template includes a target prompt word; an execution module, configured to determine a plurality of groups of initial character strings according to the first audit data and the second audit data, and perform a target merging operation on the plurality of groups of initial character strings based on the target prompt word to obtain a target character string; A generation module is used to input the target character string and the audit prompt template into a target business model to generate a target audit report, wherein different temperature parameters of the target business model are set to generate different parts of the target audit report.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein the computer program can be executed by an electronic device to perform the method according to any one of claims 1 to 6.

9. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to execute the method according to any one of claims 1 to 6 through the computer program.

Images