Automatic auditing analysis method and device based on business data and computer equipment

Abstract

The invention relates to an automatic auditing analysis method and device based on business data and computer equipment, and relates to the technical field of artificial intelligence. The method comprises the following steps: acquiring business data; identifying unstructured service data in the service data, mapping the unstructured service data to a preset structured template to perform format conversion on the service data to obtain converted service data; constructing metadata and a knowledge base according to the converted business data; and based on the metadata and the knowledge base, performing various data analysis on the converted business data to obtain a business analysis result. By adopting the method, the efficiency and quality of data analysis can be improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to an automatic audit analysis method, device and computer equipment based on business data. Background Art

[0002] With the rapid development of information technology, the amount of business data of enterprises and organizations has shown explosive growth, and the data types have become more complex and diverse. Traditional manual auditing methods are unable to cope with massive amounts of data. They are not only inefficient but also prone to errors.

[0003] Existing technologies provide business data model construction and analysis functions, can automatically extract and analyze data, and have certain collaborative working capabilities with audit management systems. However, business data audits are difficult to meet the actual needs of users, and can only process structured data. It is difficult to analyze and process unstructured and semi-structured data, and the analysis results are not ideal; the response time and quality of existing technologies are difficult to meet user needs, and performance and results are difficult to guarantee.

[0004] It can be seen that the current business data analysis method has the problems of low efficiency and poor quality of data analysis. Summary of the invention

[0005] Based on this, it is necessary to provide an automatic audit analysis method, device, computer equipment, computer-readable storage medium and computer program product based on business data that can improve the efficiency and quality of data analysis in response to the above technical problems.

[0006] In a first aspect, an embodiment of the present application provides an automatic audit analysis method based on business data. The method comprises:

[0007] Obtaining business data; the business data is data recorded in the business system and used to characterize the details of various business operations;

[0008] Identifying unstructured business data in the business data, and mapping the unstructured business data to a preset structured template to perform format conversion on the business data to obtain converted business data;

[0009] Constructing metadata and a knowledge base based on the converted business data; the metadata is data used to describe the content, structure and characteristics of the business data, and the knowledge base is a database obtained by classifying the business data;

[0010] Based on the metadata and the knowledge base, various data analyses are performed on the converted business data to obtain business analysis results; the business analysis results characterize the presentation rules of the business data and the correlation relationships between the various business data, and the business analysis results are used to verify the business data.

[0011] In one embodiment, the business analysis result includes a statistical analysis result, and the business analysis result obtained by performing various data analyses on the converted business data includes:

[0012] Predicting the relationship between the dependent variable and each variable in the business data to obtain a regression analysis result;

[0013] Identify common factors between variables in the business data, update the dimensions of the business data according to the common factors, and obtain factor analysis results;

[0014] According to the known classified variables in the business data, the classification information of the new sample is predicted to obtain the discriminant analysis result;

[0015] The regression analysis result, the factor analysis result and the discriminant analysis result are combined to obtain the statistical analysis result.

[0016] In one embodiment, the business analysis result includes a cluster analysis result, and the business analysis result is obtained by performing various data analyses on the converted business data, including:

[0017] Identifying data features of the business data, and determining a clustering algorithm according to the data features;

[0018] Calculating the total square error under different numbers of clusters, and determining the number of clusters according to the total square error;

[0019] Based on the clustering algorithm and the number of clusters, clustering the business data to obtain a clustering result;

[0020] Regularity analysis, nonspecific analysis, and outlier detection are performed on the clustering results to determine the clustering analysis results.

[0021] In one embodiment, mapping the unstructured business data to a preset structured template to convert the format of the business data to obtain the converted business data includes:

[0022] Performing word segmentation on the unstructured business data to obtain a word segmentation result;

[0023] Performing data mining on the word segmentation results to obtain subject information corresponding to the business data;

[0024] Classify the business data according to the subject information to obtain the classified business data;

[0025] Identify association rules of the classified business data, and map the unstructured business data to the structured template according to the association rules to obtain converted business data.

[0026] In one embodiment, the acquiring of business data includes:

[0027] Acquire the initial business data, perform integrity check and accuracy check on the initial business data, and obtain checked data;

[0028] Identifying different data types in the verified data, and performing standardization processing on the data of different data types to obtain standardized data;

[0029] Performing data cleaning on the standardized data to obtain cleaned data, integrating and recording data sources corresponding to the cleaned data to obtain integrated data;

[0030] The format of the text characters of the integrated data is converted to obtain the converted business data.

[0031] In one embodiment, after the step of performing various data analyses on the converted business data based on the metadata and the knowledge base to obtain business analysis results, the method further includes:

[0032] Generate a report template corresponding to the design requirement information according to the design requirement information;

[0033] Generate analysis process information of the business data according to the business analysis results, and generate working paper information based on the analysis process information and the business analysis results;

[0034] Generate a data analysis report corresponding to the business data based on the working paper information and the report template.

[0035] In a second aspect, the present application also provides an automatic audit analysis device based on business data. The device comprises:

[0036] A business data acquisition module is used to acquire business data; the business data is data recorded in the business system and used to characterize the details of various business operations;

[0037] A data conversion module, used for identifying unstructured business data in the business data, mapping the unstructured business data to a preset structured template, so as to perform format conversion on the business data to obtain converted business data;

[0038] A metadata and knowledge base construction module, used to construct metadata and a knowledge base according to the converted business data; the metadata is data used to describe the content, structure and characteristics of the business data, and the knowledge base is a database obtained by classifying the business data;

[0039] A data analysis module is used to perform various data analyses on the converted business data based on the metadata and the knowledge base to obtain business analysis results; the business analysis results characterize the presentation rules of the business data and the correlation relationships between the various business data, and the business analysis results are used to verify the business data.

[0040] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0041] Obtaining business data; the business data is data recorded in the business system and used to characterize the details of various business operations;

[0042] Identifying unstructured business data in the business data, mapping the unstructured business data to a preset structured template, so as to perform format conversion on the business data to obtain converted business data;

[0043] Constructing metadata and a knowledge base according to the converted business data; the metadata is data used to describe the content, structure and characteristics of the business data, and the knowledge base is a database obtained by classifying the business data;

[0044] Based on the metadata and the knowledge base, various data analyses are performed on the converted business data to obtain business analysis results; the business analysis results characterize the presentation rules of the business data and the correlation relationships between the various business data, and the business analysis results are used to verify the business data.

[0045] In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0046] Obtaining business data; the business data is data recorded in the business system and used to characterize the details of various business operations;

[0047] Identifying unstructured business data in the business data, mapping the unstructured business data to a preset structured template, so as to perform format conversion on the business data to obtain converted business data;

[0048] Constructing metadata and a knowledge base according to the converted business data; the metadata is data used to describe the content, structure and characteristics of the business data, and the knowledge base is a database obtained by classifying the business data;

[0049] Based on the metadata and the knowledge base, various data analyses are performed on the converted business data to obtain business analysis results; the business analysis results characterize the presentation rules of the business data and the correlation relationships between the various business data, and the business analysis results are used to verify the business data.

[0050] In a fifth aspect, the present application further provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0051] Obtaining business data; the business data is data recorded in the business system and used to characterize the details of various business operations;

[0052] Identifying unstructured business data in the business data, mapping the unstructured business data to a preset structured template, so as to perform format conversion on the business data to obtain converted business data;

[0053] Constructing metadata and a knowledge base according to the converted business data; the metadata is data used to describe the content, structure and characteristics of the business data, and the knowledge base is a database obtained by classifying the business data;

[0054] Based on the metadata and the knowledge base, various data analyses are performed on the converted business data to obtain business analysis results; the business analysis results characterize the presentation rules of the business data and the correlation relationships between the various business data, and the business analysis results are used to verify the business data.

[0055] The above-mentioned automatic audit analysis method, device, computer equipment, storage medium and computer program product based on business data first obtains relevant data recorded by the business system, such as transaction records and customer information, for subsequent analysis; then identifies unstructured business data, maps it to structured data using preset templates, ensures that the data is formatted for analysis, and then builds metadata and knowledge bases, and standardizes the management of collected data by defining data structures, creating data dictionaries, and maintaining and updating; finally, based on metadata and knowledge bases, multi-dimensional analysis of structured data is performed to identify data patterns and potential risks; the implementation of this solution achieves data accuracy and consistency, provides data-based support for decision-making, thereby improving audit quality, ensuring the effectiveness of internal control, and significantly improving the quality and efficiency of data analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related technologies, the drawings required for use in the embodiments or the related technical descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0057] FIG1 is a diagram of an application environment of an automatic audit analysis method based on business data in one embodiment;

[0058] Figure 2 A flowchart of an automatic audit analysis method based on business data in one embodiment;

[0059] Figure 3 A flowchart of an automatic audit analysis method based on business data in another embodiment;

[0060] Figure 4 is a structural block diagram of an automatic audit analysis device based on business data in one embodiment;

[0061] FIG5 is a diagram showing the internal structure of a computer device in one embodiment;

[0062] FIG. 6 is a diagram showing the internal structure of a computer device in another embodiment. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0064] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.

[0065] The automatic audit analysis method based on business data provided in the embodiment of the present application can be applied in the application environment shown in Figure 1. The terminal 102 communicates with the server 104 through a network.

[0066] The terminal 102 may be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart car-mounted devices, etc. The portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers.

[0067] In one embodiment, Figure 2 As shown, an automatic audit analysis method based on business data is provided, which is described by taking the method applied to the terminal 102 in FIG. 1 as an example, and includes the following steps:

[0068] S201, obtaining business data.

[0069] Among them, business data is data recorded in the business system and used to characterize the details of various business operations. Specifically, business data refers to specific information related to various business activities recorded in the business system. These data usually include transaction records, customer information, financial data, operation logs, etc., which can show the business processing process and results in detail. By analyzing these data, potential risks, anomalies and trends can be identified to provide support for decision-making.

[0070] For example, business data in a sales system may include sales, customer information, transaction time, and payment method for each transaction; data in a financial system may include bill details, income and expenditure records, and cost analysis. These data provide a basis for audits, enabling effective risk assessment and internal control analysis.

[0071] S202, identifying unstructured business data in the business data, and mapping the unstructured business data to a preset structured template to convert the format of the business data to obtain converted business data.

[0072] Among them, unstructured data refers to data that does not conform to a predefined structure, such as text, images, and audio, which are difficult to store and analyze directly.

[0073] For example, in order to convert it into structured data, first identify and extract key content, then map the extracted information to a preset structured template, and convert the data into a format such as JSON or CSV through format conversion to ensure the accuracy and consistency of the data for subsequent analysis and application.

[0074] S203, constructing metadata and knowledge base according to the converted business data.

[0075] Among them, metadata is data used to describe the content, structure and characteristics of business data, and the knowledge base is a database obtained by classifying business data.

[0076] Exemplarily, the metadata management process includes: (1) defining the data structure: first, the data structure and type of the audit data need to be defined, which usually involves creating database tables, fields and indexes; (2) data dictionary creation: building a data dictionary to record the definition, data type, source and purpose of all data elements; (3) maintenance and update: regularly updating metadata to reflect changes in data sources to ensure data accuracy and consistency; the knowledge base management process includes: based on the classified data, knowledge is classified according to type, field and importance to build different business knowledge bases.

[0077] S204, based on the metadata and the knowledge base, various data analyses are performed on the converted business data to obtain business analysis results.

[0078] Among them, the business analysis results represent the presentation rules of business data and the correlation between various business data. The business analysis results are used to verify business data.

[0079] Exemplarily, the data analysis process uses statistical analysis, cluster analysis, and association analysis to identify patterns and potential risks in business data, and uses regression, factor, and discriminant analysis to reveal variable relationships and determine the number and characteristics of clusters.

[0080] In the above-mentioned automatic audit analysis method based on business data, first, relevant data recorded by the business system, such as transaction records and customer information, is obtained for subsequent analysis; then, unstructured business data is identified and mapped to structured data using preset templates to ensure that the data is formatted for analysis, and then metadata and knowledge bases are constructed. By defining data structures, creating data dictionaries, and maintaining and updating, the collected data is managed in a standardized manner; finally, based on metadata and knowledge bases, structured data is analyzed in multiple dimensions to identify data patterns and potential risks; the implementation of this solution achieves data accuracy and consistency, provides data-based support for decision-making, thereby improving audit quality, ensuring the effectiveness of internal control, and significantly improving the quality and efficiency of data analysis.

[0081] In one embodiment, the business analysis results include statistical analysis results, and various data analyses are performed on the converted business data to obtain business analysis results, including: predicting the relationship between the dependent variable and the respective variables in the business data to obtain regression analysis results; identifying common factors between the variables in the business data, updating the dimensions of the business data according to the common factors, and obtaining factor analysis results; predicting classification information of new samples according to known classified variables in the business data to obtain discriminant analysis results; and combining regression analysis results, factor analysis results, and discriminant analysis results to obtain statistical analysis results.

[0082] In data analysis, statistical analysis identifies audit focus through regression analysis, factor analysis and discriminant analysis. Among them, regression analysis is used to explore the relationship between dependent variables and independent variables. The process includes selecting a suitable model, determining independent variables, fitting models, conducting statistical tests, and interpreting regression coefficients and prediction accuracy to identify suspicious points; factor analysis helps reduce data dimensions and identify abnormal samples by determining data applicability, extracting and rotating factors, calculating factor scores, and explaining the meaning of factors; discriminant analysis is applied to samples with known classifications. By selecting suitable model training data, testing model accuracy, analyzing discriminant functions, and identifying samples whose predictions do not match the actual classifications.

[0083] In this embodiment, firstly, the relationship between the dependent variable and the independent variable in the business data is predicted by regression analysis, and a suitable model is selected and variable selection and model fitting are performed to finally obtain the regression analysis results; then, factor analysis is used to identify the common factors between the variables, and factors are extracted after data applicability test, and the data dimension is reduced to obtain the factor analysis results; then, discriminant analysis is applied to known classified samples, and model training and validity test are performed to predict the classification information of new samples and obtain the discriminant analysis results; finally, the results of regression analysis, factor analysis and discriminant analysis are integrated to form a comprehensive statistical analysis result. The implementation of this series of analysis processes not only improves the ability to identify potential risks and abnormal behaviors of business data, but also enhances the pertinence and accuracy of audit focus, thereby providing data support for decision-making and ensuring the compliance and effectiveness of business activities.

[0084] In one embodiment, the business analysis results include clustering analysis results, and various data analyses are performed on the converted business data to obtain the business analysis results, including: identifying data features of the business data, and determining the clustering algorithm based on the data features; calculating the total square error under different numbers of clusters, and determining the number of clusters based on the total square error; clustering the business data based on the clustering algorithm and the number of clusters to obtain clustering results; performing regularity analysis, non-specific analysis, and outlier detection on the clustering results to determine the clustering analysis results.

[0085] Exemplarily, the process of cluster analysis includes: Algorithm selection: Select appropriate clustering algorithms according to the characteristics of the data, such as K-means, hierarchical clustering (such as AGNES), density-based clustering (such as DBSCAN), etc. The selected models are all pre-trained models using data; Determine the number of clusters by calculating the total sum of squared errors (SSE) under different numbers of clusters, and select the points where the rate of decrease of SSE slows down significantly as the number of clusters; Cluster analysis, cluster the data with the selected clustering algorithm, check the clustering results, and analyze the characteristics and size of each cluster; Analyze regularity and non-specific analysis. Internal consistency analysis, check whether the data points within each cluster have similar characteristics, and check whether the data points between different clusters have obvious differences. Analyze the center point or main features of each cluster to determine whether there are obvious patterns or trends; Outlier detection: Identify outliers or small clusters in the cluster, which may indicate abnormalities or special cases.

[0086] In this embodiment, first, a suitable clustering algorithm is selected according to the characteristics of the business data. Then, the optimal number of clusters is identified by calculating the sum of squared total errors under different numbers of clusters. Usually, a point where the rate of decrease of the squared total error slows down significantly is selected. Then, based on the determined number of clusters and algorithm, the converted business data is clustered to obtain corresponding clustering results. Then, through regularity analysis and non-specific analysis, the consistency of data points within each cluster and the differences between different clusters are checked. At the same time, the center points of the clusters and their main characteristics are analyzed to identify patterns or trends. This technical solution effectively improves the ability to process business data, enhances the ability to identify abnormal behaviors, provides a clearer and more targeted basis for decision-making, and improves the depth and breadth of business analysis.

[0087] In one embodiment, unstructured business data is mapped to a preset structured template to convert the format of the business data to obtain converted business data, including: segmenting the unstructured business data to obtain segmentation results; performing data mining on the segmentation results to obtain subject information corresponding to the business data; classifying the business data according to the subject information to obtain classified business data; identifying association rules of the classified business data, and mapping the unstructured business data to the structured template according to the association rules to obtain converted business data.

[0088] The word segmentation process includes: loading the word segmentation library, performing word segmentation on all data, building a domain dictionary, and loading the domain dictionary for processing to improve the accuracy of word segmentation.

[0089] The data mining process includes: conducting lexical statistical analysis on the segmented data, extracting key information using TFIDF, statistically analyzing word frequency information, constructing LDA thesaurus based on words, obtaining archive subject information, and statistically analyzing archive subject information.

[0090] Among them, data classification includes the classification of unstructured data such as project audit reports, contract documents, and financial statements; knowledge association refers to finding frequent item sets and association rules.

[0091] Among them, data structuring includes: template definition: defining XML or JSON structured templates for audit data; data mapping: mapping unstructured data to defined structured templates; conversion encoding: using libraries of programming languages ​​(such as Python) for data conversion; structured storage: storing the converted data in a database or file system for subsequent processing.

[0092] This embodiment realizes format conversion by mapping unstructured business data to a preset structured template. First, the unstructured data is segmented, and the segmentation library and domain dictionary are loaded to improve the accuracy of the segmentation. Then, the segmentation results are data mined, and the word frequency statistics and TFIDF methods are used to build the LDA subject vocabulary and extract the subject information. Then, the business data is classified according to the extracted subject information, including types such as project audit reports and contract documents. Subsequently, the association rules of the classified data are identified to find frequent item sets. Finally, the unstructured data is mapped to a defined XML or JSON structured template, and the coding conversion is performed using a programming language to ensure the standardized storage of the data. This technical solution significantly improves the processing efficiency and accuracy of unstructured data, making it convenient for subsequent analysis and decision-making, while enhancing the availability and structured management capabilities of the data, providing more efficient and reliable support for business analysis.

[0093] In one embodiment, obtaining business data includes: obtaining initial business data, performing integrity check and accuracy check on the initial business data to obtain verified data; identifying different data types in the verified data, and standardizing data of different data types respectively to obtain standardized data; performing data cleansing on the standardized data to obtain cleansed data, integrating and recording data sources corresponding to the cleansed data to obtain integrated data; converting the format of text characters of the integrated data to obtain converted business data.

[0094] First, the business project data of the audit management system is obtained by calling the API interface. In the process of data collection, the integrity, accuracy, availability and security of the acquired data need to be guaranteed. The data integrity is confirmed by comparing the system data with the imported data. At the same time, data accuracy verification is required, and data validation rules are used to check whether the field content conforms to the predefined format.

[0095] Among them, standardization processing refers to standardizing different types of unstructured data such as text, images, audio, and video. For text, images, PPT and other data, parsing algorithms and their OCR (Optical Character Recognition) technology are used to fully obtain the hidden information in the data, and the data is saved in the form of Excel and text. For audio and video data, it is necessary to combine OCR and speech recognition technology to extract hidden information and save the information. The saved information needs to be associated with the original data.

[0096] Among them, after data cleaning, it is necessary to integrate data from different sources: by merging and associating different data sets, create a unified view for cross-source analysis; finally, convert the data: such as converting date strings into date formats, unifying currency units, classifying data codes, etc., to normalize the data.

[0097] In this embodiment, the API interface is first used to obtain the initial business data from the audit management system, and the integrity and accuracy are checked to ensure the integrity, accuracy and availability of the data. Then, the different data types in the verified data are identified, and the text, image, audio and video are standardized respectively. The hidden information is extracted using the parsing algorithm and optical character recognition technology, and saved as text or Excel format. Then, the standardized data is cleaned, data from different sources are integrated, and a unified view is created through merging and association to support cross-source analysis. Finally, the integrated data is converted into a format, such as the standardization of dates and currency units, to ensure the consistency and standardization of the data. This technical solution effectively improves the efficiency and accuracy of data processing, enables different types of data to be effectively integrated and analyzed, enhances the basis for business decision-making, and ensures the security and traceability of the data.

[0098] In one embodiment, after performing various data analyses on the converted business data based on metadata and knowledge base to obtain business analysis results, the method also includes: generating a report template corresponding to the design requirement information based on the design requirement information; generating analysis process information of the business data based on the business analysis results, and generating working paper information based on the analysis process information and the business analysis results; generating a data analysis report corresponding to the business data based on the working paper information and the report template.

[0099] Among them, the configuration process of the report template includes: template design: designing the template according to the requirements of the audit report, including the format, content and layout; customization: allowing the report template to be customized according to the needs of different audit projects.

[0100] Among them, the process of generating working papers includes: recording the analysis process: recording each step of the analysis process in detail, including the scripts used, parameter settings, analysis results and conclusions drawn, to ensure the integrity of the audit trail; saving analysis scripts and results: saving all analysis scripts and generated result files in the tool, and organizing them according to the format requirements of the audit working papers.

[0101] Among them, the generation process of the data analysis report includes: using the reporting function to prepare an audit report: selecting an appropriate report template to integrate the analysis results, audit conclusions and recommendations into a unified format and detailed audit report; ensuring that the report content is clear: the report should have clear logic and concise language, avoid the use of professional terms or abbreviations, and ensure that non-professionals can also understand the audit findings and recommendations.

[0102] In this embodiment, first, a corresponding report template is generated according to the design requirements. The template design includes format, content and layout, and supports customization according to different audit projects. Secondly, each step in the analysis process is recorded to ensure the integrity of the audit trail, and the analysis script and results are organized into working papers. Finally, based on the working papers and report templates, a detailed and logically clear audit report is compiled to ensure that even non-professionals can understand it. This technical solution optimizes the report generation process, improves the readability and traceability of business analysis results, and provides more efficient and standardized support for audit work.

[0103] In another embodiment, Figure 3 As shown, an automatic audit analysis method based on business data is provided, comprising the following steps:

[0104] S301, obtaining initial business data, performing integrity check and accuracy check on the initial business data, and obtaining checked data.

[0105] S302, identifying different data types in the verified data, and performing standardization processing on the data of different data types to obtain standardized data.

[0106] S303, performing data cleaning on the standardized data to obtain cleaned data, integrating and recording data sources corresponding to the cleaned data to obtain integrated data.

[0107] S304, converting the format of the text characters of the integrated data to obtain converted business data.

[0108] S305, identifying unstructured business data in the business data, and mapping the unstructured business data to a preset structured template to convert the format of the business data to obtain converted business data.

[0109] S306, constructing metadata and a knowledge base according to the converted business data, and performing various data analyses on the converted business data based on the metadata and the knowledge base to obtain business analysis results.

[0110] S307, generating a report template corresponding to the design requirement information according to the design requirement information, generating analysis process information of the business data according to the business analysis results, and generating working draft information based on the analysis process information and the business analysis results.

[0111] S308, generating a data analysis report corresponding to the business data based on the working paper information and the report template.

[0112] It should be noted that the specific limitations of the above steps can be found in the above specific limitations of an automatic audit analysis method based on business data, which will not be repeated here.

[0113] Intelligent audit data analysis tools can help auditors deeply explore the value of data, quickly build and understand business data models through more detailed data analysis, automate data analysis processes, and improve audit efficiency. Audit data analysis tools in existing technologies provide business data model construction and analysis functions, can automatically extract and analyze data, and have certain collaborative working capabilities with audit management systems. However, these tools have many technical limitations:

[0114] (1) Business data auditing is difficult to meet the actual needs of users and can only process structured data. It is difficult to analyze and process unstructured and semi-structured data, and the analysis effect is not ideal; (2) The response time and quality of technical support are difficult to meet user needs, and performance and results are difficult to guarantee; (3) Audit analysis methods are backward and the results are poor. There is a lack of technical iteration and updates, and the data processing capabilities are weak. It is difficult to handle very complex data processing tasks and it is difficult to meet users' rapidly changing audit needs; (4) There is a lack of customized analysis and processing for specific needs; (5) Audit analysis is difficult to achieve across platforms, data structures, and fields, and it is difficult to achieve real-time auditing; (6) There is a lack of complete audit technology standards.

[0115] Based on this, the embodiment of the present application provides an automatic audit analysis method based on business data, and the automatic audit analysis method based on business data is described in detail with a specific embodiment below. It is worth understanding that the following description is only an exemplary description, and not a specific limitation of the application.

[0116] This application proposes an audit analysis method based on business data, which can process unstructured and semi-structured multi-type data. Semi-structured data includes: html, JSON and XML, and unstructured data includes text, audio and video. Provide a business data model for the implementation stage of the audit project, acquire, analyze, refine, summarize and apply massive business data from multiple angles, combine big data and artificial intelligence methods to conduct intelligent analysis of business data, and provide a more customized audit analysis model and function.

[0117] This application proposes an audit analysis method based on business data with the following characteristics: (1) Efficient and complex data processing capabilities. By using big data processing technology, high-performance processing and analysis of massive structured and semi-structured multi-type data can be achieved to improve audit efficiency; (2) Intelligent audit analysis capabilities. An intelligent anomaly detection and predictive analysis algorithm model is designed. By combining machine learning models to learn abnormal patterns in data, the accuracy and efficiency of audits are improved, and data is analyzed and predicted to provide decision support for auditors to assess future risks. (3) Customized audit analysis models, such as industry-specific models: audit analysis models customized for different industries are provided to better meet the needs of specific business scenarios. User-defined models: allow auditors to customize analysis models and rules based on their own experience and needs. (4) Automated audit of business data processes. Standardize and streamline business data audit processes, automate workflows, and achieve full automation from data extraction, cleaning to analysis; achieve coordination between audit processes and audit management systems to achieve closed-loop management of audit processes. (5) Data security and compliance assurance. Data encryption technology is applied during data transmission and storage to ensure data security, and a compliance check mechanism is built in to ensure that the audit process complies with relevant laws, regulations and industry standards.

[0118] The intelligent audit analysis method based on business data provided by this application specifically includes the following steps:

[0119] Step 1. Data Collection

[0120] Obtain business project data of the audit management system by calling the API interface.

[0121] In the process of data collection, the integrity, accuracy, availability and security of the acquired data need to be ensured. The data integrity is confirmed by comparing the system data and the imported data. At the same time, data accuracy verification is required, and data validation rules are used to check whether the field content conforms to the predefined format. The data is encrypted during the collection process to effectively ensure the security of the data. In the process of data collection, the generated data needs to be tracked, carefully collected, shared, and judged to ensure that the information is obtained from the source.

[0122] Step 2. Data preparation

[0123] Standardization of data, standardization of different types of unstructured data such as text, images, audio, video, etc.

[0124] For text (word, PDF), image, PPT and other data, parsing algorithms and their OCR (Optical Character Recognition) technology are used to completely obtain the hidden information in the data and save the data in the form of Excel and text.

[0125] For audio and video data, it is necessary to combine OCR and speech recognition technology to extract hidden information and save the information. The saved information needs to be associated with the original data.

[0126] OCR implementation process:

[0127] Combined with OCR, the text contained in the image is recognized, and the valuable information contained in the image is mined by combining image-related technologies. OCR recognition mainly includes two stages: text detection and text recognition, which are implemented by DBNet and CRNN respectively.

[0128] (1) DBnet text detection

[0129] The DBNet model is used for feature extraction in the image text position detection stage. First, the image is input into the ResNet backbone network with a feature pyramid. The feature pyramid is upsampled from top to bottom, and the upsampled features are cascaded with features of the same size to obtain a feature map (F). The feature map is used to predict the probability map (P) and the threshold map (T). Then, an approximate binary map is calculated using P and F. The probability map (P) indicates the probability that a pixel is text, and the threshold map (T) indicates whether each pixel is text. Each pixel is adaptively binarized. The binarization threshold is not pre-set but learned by the network. The binarization step is added to the network for training. In this way, the final output image will have very good robustness to the threshold, which simplifies post-processing while improving the effect of text detection. The network structure of DBNet is as follows: Figure 3 shown.

[0130] In the segmentation network of DBNet, the binarization operation is added to the network for joint training, so that the network can adaptively learn the threshold of the pixel corresponding to each position in the image, which can more accurately separate text and non-text. The standard binarization function does not have the characteristic of gradient differentiability, so it cannot be added to the network for training. Therefore, a DB (Differentiable Binarization) module is proposed to implement differentiable binarization and solve the problem of non-differentiable gradients. The DB module is differentiable, which allows the binarization process to be added to the training process of the end-to-end model.

[0131] (2) CRNN text recognition

[0132] The CRNN neural network model is used to implement the archive text content recognition stage. CRNN is a neural network model that combines CNN and RNN. The significant advantage of the CRNN model is that it does not need to label each character to achieve sequence learning and has no constraints on the length of sequence-like objects. The CRNN text recognition model uses a convolutional neural network to extract the deep features of the input image, and sends it to a bidirectional recurrent neural network to predict the label distribution of each frame sequence. Finally, CTC is used to convert the feature sequence into a label sequence. The CRNN model uses the CTC operation to only care about which text appears in the image, without having to care about the specific position and width of the text. The CTC loss is differentiable and will guide the NN model to train. However, in the problem of scene text recognition, there is some semantic information between each time slice. CTC has no restrictions on the length of the recognized text and can learn the correlation information of the text sequence.

[0133] CRNN is mainly divided into three components, the model structure is as follows Figure 4 As shown in the figure, from bottom to top, they are the CNN network, the RNN network, and the transcription layer. The workflow is as follows: (1) The input image first passes through the convolution layer to extract a feature sequence; (2) Then the RNN predicts each frame of the feature sequence; (3) Finally, the output layer transcribes the prediction results of each frame of the RNN to obtain a final label sequence, that is, each character predicted by the RNN is combined to obtain a complete word.

[0134] Combined with OCR, the text contained in the image can be identified, including: important event names, conference names, competition names, event time, etc. By combining image-related technologies, valuable information contained in the image can be mined.

[0135] The steps for implementing image data OCR recognition are as follows:

[0136] Use the DBNet text detection model to detect the text location in the archive image data. The DBnet text detection calculation formula is as follows:

[0137]

[0138] Where k is the amplification factor, is the probability, Represents the adaptive threshold learned from the network.

[0139] Use the CRNN text content recognition model to identify the text content corresponding to the text image detected in the previous step, and finally obtain the text position and corresponding text content in the image. The CRNN calculation formula is as follows:

[0140]

[0141]

[0142] in The mapping function B from sequence to sequence is transformed into the set of all paths of text l, and π is one of the paths.

[0143] Step 3. Data cleaning

[0144] Perform data cleaning, data integration and data conversion on the collected data.

[0145] (1) First, data cleaning is required: data cleaning should be carried out in an appropriate manner, such as removing null values, correcting data type errors, deleting duplicate records, etc. Missing data should be filled, estimated, or deleted. Systematic review should be conducted on massive data information with different structures and large differences, and data information with duplicate content, incomplete information, and irregular format should be eliminated. Audit information with accurate content, complete content, and consistent format should be saved to improve the quality of audit data.

[0146] (2) Secondly, it is necessary to integrate data from different sources: by merging and associating different data sets, a unified view can be created to facilitate cross-source analysis. Through big data, unstructured data such as video, audio, and images related to the audit analysis program, as well as semi-structured data such as web pages and logs, can be centrally integrated and processed to prevent audit data from being tampered with.

[0147] (3) Finally, convert the data: for example, convert date strings into date formats, unify currency units, classify data codes, etc.; unify text formats, such as unifying to lowercase or uppercase, and process simplified and traditional Chinese conversion in Chinese text to standardize data.

[0148] Step 4: Intelligent data analysis

[0149] Since unstructured data has diverse data types, complex relationships, and hidden knowledge that is difficult to mine, an intelligent data analysis module is designed to process the data and structure it.

[0150] (1) Tokenization

[0151] Load the jieba word segmentation library, perform word segmentation on all data, build domain dictionaries (archives, electricity, etc.), and load domain dictionaries for processing to improve word segmentation accuracy.

[0152] (2) Data mining

[0153] The data after word segmentation is processed by vocabulary statistics. TFIDF is used to extract key information, count word frequency information, and build LDA subject word library based on words to obtain archive subject information and statistically analyze archive subject information.

[0154] The TFIDF calculation process is: ;

[0155] in, is the frequency of word i appearing in document j, is the sum of the frequency of all words in document j, D is the total number of documents, is the total number of documents containing the term.

[0156] LDA is an unsupervised machine learning technology. In the field of text mining, it can be used to identify the hidden topic information in large-scale document sets or corpora. In the LDA model, the number of topics needs to be set in advance. In order to obtain the most appropriate number of topics, Python is used to build a perplexity solution model to obtain the optimal number of topics. The calculation formula is:

[0157] ;

[0158] (3) Data classification

[0159] Classification of unstructured data such as project audit reports, contract documents, financial statements, etc.

[0160] (4) Knowledge association

[0161] Use the Apriori algorithm or FP-growth algorithm to find frequent item sets and association rules. Result analysis: Analyze the mined rules, such as discovering the association between certain contract terms and audit risks. Identify key entities and entity relationships from audit data, and determine the relationships between entities, such as investment relationships, contractual relationships, etc. Build a graph: Use a graph database (such as Neo4j) to build a knowledge graph, storing entities and relationships as nodes and edges. Apply reasoning: Perform reasoning on the knowledge graph to discover potential knowledge associations.

[0162] (5) Data structuring

[0163] Define Template: Define a structured template in XML or JSON for audit data.

[0164] Data mapping: Mapping unstructured data into defined structured templates.

[0165] Transcoding: Use libraries in programming languages ​​such as Python to perform data transformations.

[0166] Structured storage: Store the converted data in a database or file system for subsequent processing.

[0167] Step 5: Metadata Management

[0168] (1) Define data structure: First, you need to define the data structure and type of the audit data, which usually involves creating database tables, fields, and indexes.

[0169] (2) Data dictionary creation: Build a data dictionary to record the definition, data type, source, and purpose of all data elements.

[0170] (3) Maintenance and update: Regularly update metadata to reflect changes in data sources and ensure data accuracy and consistency

[0171] Step 6: Knowledge Base Management

[0172] Based on the classified data, knowledge is classified according to type, field and importance to build different business knowledge bases.

[0173] Step 7. Audit data analysis:

[0174] (1) Statistical analysis: mainly using regression analysis, factor analysis, discriminant analysis and other methods to list suspicious key points as the audit focus;

[0175] Regression analysis: Regression analysis is used to predict the relationship between a dependent variable and one or more independent variables. Detailed analysis process:

[0176] ①Select model: Select an appropriate regression model according to the type of dependent variable (continuous or categorical), such as linear regression, logistic regression, etc.

[0177] ②Variable selection: Select independent variables through correlation analysis, stepwise regression and other methods.

[0178] ③Model fitting: Fit the regression model using the selected independent variables.

[0179] ④Model testing: Conduct statistical tests on the model, such as t-test and F-test, to evaluate the significance of the model.

[0180] ⑤ Result interpretation: Analyze the regression coefficients to determine which independent variables have a significant impact on the dependent variable.

[0181] ⑥ Prediction and evaluation: Use the model to make predictions and evaluate the accuracy of the predictions.

[0182] ⑦ Suspicious point identification: Identify variables with abnormally large regression coefficients, which may indicate abnormal behavior. Check the residuals. Abnormally large residuals may indicate that the data point does not fit the model and requires further investigation.

[0183] Factor analysis: Factor analysis is used to identify common factors among multiple variables and reduce the dimensionality of the data. Detailed analysis process:

[0184] ① Data suitability test: Use Bartlett's sphericity test and KMO test to determine whether the data is suitable for factor analysis.

[0185] ② Extract factors: Use principal component analysis and other methods to extract factors.

[0186] ③Factor rotation: Use orthogonal rotation or oblique rotation method to make the factor structure clearer.

[0187] ④Factor score: Calculate the score of each sample on each factor.

[0188] ⑤Result interpretation: Based on the factor loading, explain the meaning of each factor.

[0189] ⑥ Identification of suspicious points: Check samples with abnormally high factor scores, which may indicate that these samples have abnormal behavior on specific factors.

[0190] Discriminant analysis: Discriminant analysis is used to predict the classification of new samples based on known classified variables. Detailed analysis process:

[0191] ①Select model: Select linear discriminant analysis (LDA) or quadratic discriminant analysis (QDA) according to the data type.

[0192] ②Variable selection: Select variables that have a significant impact on classification.

[0193] ③Model training: Use data with known classification to train the discriminant model.

[0194] ④Model verification: Use methods such as cross-validation to test the accuracy of the model.

[0195] ⑤Result interpretation: Analyze the discriminant function to determine which variables are most important for classification.

[0196] ⑥ Identification of suspicious points: Identify samples where the model’s predicted classification does not match the actual classification, which may be the focus of the audit.

[0197] (2) Cluster analysis: Analyze whether the data is regular and non-specific;

[0198] ① Algorithm selection: Select a suitable clustering algorithm based on the characteristics of the data, such as K-means, hierarchical clustering (such as AGNES), density-based clustering (such as DBSCAN), etc. The selected models are all pre-trained models using data.

[0199] ② Determine the number of clusters by calculating the sum of squared errors (SSE) under different numbers of clusters, and select the point where the rate of SSE decreases significantly slows down as the number of clusters.

[0200] ③ Cluster analysis: cluster the data using the selected clustering algorithm, check the clustering results, and analyze the characteristics and size of each cluster.

[0201] ④Analysis of regularity and non-specificity. Internal consistency analysis, check whether the data points within each cluster have similar characteristics, and check whether the data points between different clusters have obvious differences. Analyze the center point or main characteristics of each cluster to determine whether there are obvious patterns or trends.

[0202] ⑤Outlier detection: Identify outliers or small clusters in the cluster, which may indicate anomalies or special cases

[0203] (3) Correlation analysis: Through correlation analysis, abnormal correlation data can be discovered and in-depth analysis of the abnormal data can be performed.

[0204] (4) Analyze trends, patterns, and relationships: Use the tool’s data visualization tools, such as charts and scatter plots, to reveal trends, patterns, and relationships in the data to discover audit clues.

[0205] (5) Risk Assessment Analysis: Apply analysis commands and scripts: Write or use existing tool scripts to perform specific audit tests, such as account balance analysis, transaction frequency analysis, etc., to identify potential risk points. Use the tool's risk assessment tools to assess the risk level in the data based on preset risk indicators and determine which areas require further investigation.

[0206] (6) Audit suspicion anomaly identification: Identify anomalies based on preset rules and standards: Set specific audit rules, such as transaction amounts exceeding certain thresholds, abnormal transaction frequencies, etc., to automatically mark abnormal data. Investigate anomalies: Conduct in-depth analysis of marked abnormal data and determine whether the anomaly constitutes an audit issue by comparing historical data, industry standards, and internal control requirements.

[0207] (7) Analyze a large amount of data of the audited unit through stage characteristic analysis, distribution trend analysis, comparative analysis, etc., so as to make the internal audit more targeted.

[0208] (8) Summary analysis;

[0209] (9)Customized analysis.

[0210] Step 8. Analyze data visualization

[0211] Visualizing massive data helps auditors to conduct quick and effective interactive analysis of massive data, thereby enhancing audit insight; the application of cloud computing and other technologies in practice can effectively analyze audit data and realize real-time online auditing. Specifically, it includes:

[0212] (1) Structured Data Visualization

[0213] Data extraction: Extract the data that needs to be visualized from the structured database.

[0214] Design charts: Choose appropriate chart types based on the purpose of analysis, such as bar charts, line charts, pie charts, etc.

[0215] Configure visualization: Use libraries such as ECharts to configure chart parameters, such as axes, data series, legends, etc.

[0216] Display interaction: Embed charts into web pages and implement interactive functions such as data filtering and drilling, so that users can deeply analyze data.

[0217] Step 9. Report template configuration

[0218] (1) Template design: Design the template according to the requirements of the audit report, including format, content and layout.

[0219] (2) Customization: Allows report templates to be customized according to the needs of different audit projects.

[0220] Step 10. Preparation of working papers

[0221] Record the analysis process: Record each step of the analysis process in detail, including the scripts used, parameter settings, analysis results and conclusions drawn, to ensure the integrity of the audit trail.

[0222] Save analysis scripts and results: Save all analysis scripts and generated result files in the tool and organize them according to the format requirements of audit working papers.

[0223] Step 11. Result evaluation

[0224] Verify the accuracy of analytical results: Verify the accuracy of analytical results by independent review, re-performing the analytical process, or comparing with the explanation provided by the audited entity.

[0225] Communicate with the audited entity: Discuss key findings in the analysis results with the audited entity’s management or relevant departments to clarify questions and obtain additional explanations or evidence.

[0226] Step 12. Risk warning management

[0227] Potential risks are identified through audit models and data analysis, and warning thresholds are set according to risk levels. When the risk exceeds the threshold, the system automatically issues a warning and initiates the corresponding response mechanism.

[0228] Step 13. Report preparation

[0229] Prepare audit reports using the reporting function: select an appropriate report template to combine analysis results, audit conclusions and recommendations into a consistent and detailed audit report.

[0230] Ensure that the report is clear: The report should be logically clear and concise, avoiding the use of professional terms or abbreviations, to ensure that non-professionals can understand the audit findings and recommendations.

[0231] Step 14. Follow up

[0232] Collaborate with the auditee to resolve issues: Collaborate with the auditee to develop an action plan to address issues and recommendations raised in the audit report.

[0233] Follow-up Audits: Follow-up audits are conducted after a period of time to confirm whether previously identified issues have been resolved and whether new controls are effective.

[0234] Step 15. Continue monitoring

[0235] Set up an ongoing monitoring program: Set up an ongoing monitoring program for key business processes or high-risk areas, defining the monitoring frequency and conditions.

[0236] Real-time or periodic detection of anomalies: Run monitoring scripts regularly to check data in real time or periodically. Once an anomaly is found, a report is immediately generated and relevant personnel are notified to take action.

[0237] Step 16. Audit analysis job management

[0238] (1) Standardization of audit analysis processes: Different audit analysis processes are formulated for different audit business data, and the processes are standardized in this module.

[0239] (2) Assign audit analysis tasks to different members based on the audit plan and resources.

[0240] (3) Track the completion status of audit analysis tasks in real time to ensure that the audit analysis work proceeds as planned.

[0241] (4) Quality control: Ensure that the quality of audit analysis work meets audit standards and requirements.

[0242] This application starts from the source of business data, and through data collection, data preparation, data cleaning, intelligent data analysis, metadata management, knowledge base construction, audit data analysis, analysis data visualization, analysis report, risk warning, continuous monitoring, and operation management, it fully covers the business data analysis needs in the audit implementation work. It greatly solves the problems existing in traditional data analysis tools, and the technical effects achieved include at least:

[0243] (1) This application supports structured, unstructured, and semi-structured data analysis, including image, audio, and video data, solving the problem that traditional analysis tools only support structured data. (2) This application supports data correlation processing, data classification, and other processing, solving the data island problem of single data source analysis. Improve the reliability and accuracy of data analysis. (3) This application supports a large number of big data analysis models such as regression, clustering, decision trees, and association analysis, solving the problem that traditional analysis tools use relatively backward analysis methods such as query and filtering, have weak data processing capabilities, are unable to handle very complex data processing tasks, and are unable to meet the rapid changes in audit analysis needs. (4) This application supports highly customized analysis algorithms and flexible parameter adjustment, solving the problem that traditional data analysis tools cannot meet the customized analysis and processing needs of specific needs. (5) This application supports the collection and organization of data across platforms, data structures, and business fields, and ensures the real-time nature of the data. It solves the problem that traditional data analysis tools cannot be directly implemented, require a large amount of manual preprocessing, and are inefficient, as well as the timeliness of data. (6) This application comprehensively formulates the process and technical standards for audit data analysis from data collection to analysis reports. Combined with the audit analysis job management function, a standard solution is provided for data analysis during the audit implementation phase.

[0244] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0245] Based on the same inventive concept, the embodiment of the present application also provides an automatic audit analysis device based on business data for implementing the automatic audit analysis method based on business data involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in the embodiments of one or more automatic audit analysis devices based on business data provided below can refer to the limitations of the automatic audit analysis method based on business data above, and will not be repeated here.

[0246] In one embodiment, Figure 3As shown, an automatic audit analysis device based on business data is provided, including: a business data acquisition module 301, a data conversion module 302, a metadata and knowledge base construction module 303, and a data analysis module 304, wherein:

[0247] The business data acquisition module 301 is used to acquire business data; business data is data recorded in the business system and used to represent the details of each business transaction;

[0248] The data conversion module 302 is used to identify unstructured business data in the business data, map the unstructured business data to a preset structured template, and perform format conversion on the business data to obtain converted business data;

[0249] The metadata and knowledge base construction module 303 is used to construct metadata and knowledge base according to the converted business data; metadata is data used to describe the content, structure and characteristics of business data, and the knowledge base is a database obtained by classifying business data;

[0250] The data analysis module 304 is used to perform various data analyses on the converted business data based on the metadata and knowledge base to obtain business analysis results; the business analysis results characterize the presentation rules of the business data and the correlation between various business data, and the business analysis results are used to verify the business data.

[0251] In one embodiment, the device is used to: identify data features of business data and determine a clustering algorithm based on the data features; calculate the total square error under different numbers of clusters and determine the number of clusters based on the total square error; cluster the business data based on the clustering algorithm and the number of clusters to obtain clustering results; perform regularity analysis, non-specific analysis, and outlier detection on the clustering results to determine the clustering analysis results.

[0252] In one embodiment, the device is used to: perform word segmentation on unstructured business data to obtain word segmentation results; perform data mining on the word segmentation results to obtain subject information corresponding to the business data; classify the business data according to the subject information to obtain classified business data; identify association rules of the classified business data, and map the unstructured business data to a structured template according to the association rules to obtain converted business data.

[0253] In one embodiment, the device is used to: obtain initial business data, perform integrity check and accuracy check on the initial business data to obtain verified data; identify different data types in the verified data, and standardize the data of different data types respectively to obtain standardized data; perform data cleansing on the standardized data to obtain cleansed data, integrate and record the data sources corresponding to the cleansed data to obtain integrated data; convert the format of the text characters of the integrated data to obtain converted business data.

[0254] In one embodiment, the device is used to: generate a report template corresponding to the design requirement information based on the design requirement information; generate analysis process information of the business data based on the business analysis results, and generate working paper information based on the analysis process information and the business analysis results; generate a data analysis report corresponding to the business data based on the working paper information and the report template.

[0255] Each module in the automatic audit analysis device based on business data can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0256] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be shown in FIG5. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, an automatic audit analysis method based on business data is implemented.

[0257] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be shown in FIG6. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit, and an input device. The processor, the memory, and the input / output interface are connected via a system bus, and the communication interface, the display unit, and the input device are connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner may be implemented through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, an automatic audit analysis method based on business data is implemented. The display unit of the computer device is used to form a visually visible picture, which may be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0258] Those skilled in the art will understand that the structures shown in FIGS. 5 and 6 are merely block diagrams of partial structures related to the scheme of the present application, and do not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device may include more or fewer components than those shown in the figures, or combine certain components, or have a different arrangement of components.

[0259] In one embodiment, a computer device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, the steps in the above method embodiments are implemented.

[0260] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0261] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0262] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0263] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0264] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. An automatic audit analysis method based on business data, characterized in that: The method comprises: Obtaining business data; the business data is data recorded in the business system and used to characterize the details of various business operations; Identifying unstructured business data in the business data, and mapping the unstructured business data to a preset structured template to perform format conversion on the business data to obtain converted business data; Constructing metadata and a knowledge base based on the converted business data; the metadata is data used to describe the content, structure and characteristics of the business data, and the knowledge base is a database obtained by classifying the business data; Based on the metadata and the knowledge base, various data analyses are performed on the converted business data to obtain business analysis results; the business analysis results characterize the presentation rules of the business data and the correlation relationships between the various business data, and the business analysis results are used to verify the business data.

2. The method according to claim 1, characterized in that The business analysis result includes a statistical analysis result, and the business analysis result obtained by performing various data analyses on the converted business data includes: Predicting the relationship between the dependent variable and each variable in the business data to obtain a regression analysis result; Identify common factors between variables in the business data, update the dimensions of the business data according to the common factors, and obtain factor analysis results; According to the known classified variables in the business data, the classification information of the new sample is predicted to obtain the discriminant analysis result; The regression analysis result, the factor analysis result and the discriminant analysis result are combined to obtain the statistical analysis result.

3. The method according to claim 1, characterized in that The business analysis result includes a cluster analysis result, and the business analysis result obtained by performing various data analyses on the converted business data includes: Identifying data features of the business data, and determining a clustering algorithm according to the data features; Calculating the total square error under different numbers of clusters, and determining the number of clusters according to the total square error; Based on the clustering algorithm and the number of clusters, clustering the business data to obtain a clustering result; Regularity analysis, nonspecific analysis, and outlier detection are performed on the clustering results to determine the clustering analysis results.

4. The method according to claim 1, characterized in that: Mapping the unstructured business data to a preset structured template to convert the format of the business data to obtain the converted business data includes: Performing word segmentation on the unstructured business data to obtain a word segmentation result; Performing data mining on the word segmentation results to obtain subject information corresponding to the business data; Classify the business data according to the subject information to obtain the classified business data; Identify association rules of the classified business data, and map the unstructured business data to the structured template according to the association rules to obtain converted business data.

5. The method according to claim 1, characterized in that: The obtaining of business data includes: Acquire the initial business data, perform integrity check and accuracy check on the initial business data, and obtain checked data; Identifying different data types in the verified data, and performing standardization processing on the data of different data types to obtain standardized data; Performing data cleaning on the standardized data to obtain cleaned data, integrating and recording data sources corresponding to the cleaned data to obtain integrated data; The format of the text characters of the integrated data is converted to obtain the converted business data.

6. The method according to claim 1, characterized in that After the step of performing various data analyses on the converted business data based on the metadata and the knowledge base to obtain business analysis results, the method further includes: Generate a report template corresponding to the design requirement information according to the design requirement information; Generate analysis process information of the business data according to the business analysis results, and generate working paper information based on the analysis process information and the business analysis results; Generate a data analysis report corresponding to the business data based on the working paper information and the report template.

7. An automatic audit analysis device based on business data, characterized in that: The device comprises: A business data acquisition module is used to acquire business data; the business data is data recorded in the business system and used to characterize the details of various business operations; A data conversion module, used for identifying unstructured business data in the business data, mapping the unstructured business data to a preset structured template, so as to perform format conversion on the business data to obtain converted business data; A metadata and knowledge base construction module, used to construct metadata and a knowledge base according to the converted business data; the metadata is data used to describe the content, structure and characteristics of the business data, and the knowledge base is a database obtained by classifying the business data; A data analysis module is used to perform various data analyses on the converted business data based on the metadata and the knowledge base to obtain business analysis results; the business analysis results characterize the presentation rules of the business data and the correlation relationships between the various business data, and the business analysis results are used to verify the business data.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, 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. 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.

Images