Intelligent data filling and reporting system

Abstract

The invention provides an intelligent data filling and reporting system, which belongs to the technical field of data filling and reporting, and comprises a first determination module for obtaining data characteristics and data types of target data, and determining corresponding technical stacks according to the data characteristics and the data types; the first acquisition module is used for acquiring an automatic filling suggestion based on the filling prompt according to the technical stack; the checking module is used for setting a multi-level data verification mechanism and checking the format and logic of the data according to the verification mechanism; the detection module is used for filling the target data according to the automatic filling suggestion and the verification result, and detecting potential errors and inconsistency based on an artificial intelligence technology; and the correction module is used for correcting errors and inconsistency, displaying the data in a visual mode and realizing intelligent filling and reporting of the data. The problems that a large number of errors occur and the filling efficiency is reduced due to manual data filling, and meanwhile, the filling process becomes tedious and the data accuracy is reduced due to lack of a multi-layer verification mechanism are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of data reporting, and in particular to an intelligent data reporting system. Background Art

[0002] With the rapid development of my country's economy and the continuous improvement of science and technology, data collection and processing have become increasingly important.

[0003] Traditional data reporting methods cannot automatically report data and need to be done manually, which will result in a large number of errors, reduce reporting efficiency, and increase labor costs. At the same time, the lack of a multi-layer verification mechanism makes the reporting process cumbersome and reduces the accuracy and precision of the data.

[0004] Therefore, the present invention proposes an intelligent data reporting system. Summary of the invention

[0005] The present invention provides an intelligent data reporting system to solve the defects in the prior art that data needs to be reported manually, a large number of errors will occur, the reporting efficiency will be reduced, and the labor cost will be increased. At the same time, the lack of a multi-layer verification mechanism makes the reporting process cumbersome and reduces the accuracy and precision of the data.

[0006] In one aspect, the present invention provides an intelligent data reporting system, comprising: The first determination module: obtains data characteristics and data types of target data, and determines the corresponding technology stack according to the data characteristics and data types; A first acquisition module: acquiring auto-fill suggestions based on the filling prompt according to the technology stack; Verification module: setting up multiple levels of data verification mechanisms, and verifying the format and logic of data according to the verification mechanisms; Detection module: filling in the target data according to the automatic filling suggestions and verification results, and detecting potential errors and inconsistencies based on artificial intelligence technology; Correction module: correct the errors and inconsistencies, and display the data in a visual way to achieve intelligent data reporting.

[0007] According to an intelligent data reporting system provided by the present invention, a first determination module includes: A first acquisition unit: analyzing the target data to acquire data distribution and development trend of the target data; A second acquisition unit: acquiring data features of target data according to the data distribution and development trend; Extraction unit: extracts keywords of target data according to a preset method, and determines detailed information of target data according to the data source; A first determining unit: determining the data type of the target data according to the keyword and detailed information of the data; The third acquisition unit: obtains data processing requirements and storage requirements, and obtains data scale according to data characteristics and data types; The second determining unit determines a corresponding technology stack based on the data processing requirements, storage requirements, and data scale.

[0008] According to an intelligent data reporting system provided by the present invention, a first acquisition module includes: Processing unit: Use the word segmentation model to perform word segmentation on the existing data and filled-in prompts to obtain words or character sequences; Conversion unit: converts each generated word or character sequence into the corresponding word embedding vector; Learning unit: learning the mapping relationship between the word embedding vector and the corresponding label by establishing a neural network model; Prediction unit: predicts possible labels of target data through a neural network model based on the technology stack; Selection unit: Based on the probability distribution of each label in the prediction results, select the label with the highest probability distribution as the auto-fill suggestion; If multiple labels have the same Gaussian kernel density, one of the labels can be randomly selected.

[0009] According to the present invention, an intelligent data reporting system and a verification module include: Definition unit: define data format rules and logic validation rules according to the requirements of the reporting system; The third determination unit is: writing a corresponding nested verification function, and inputting the target data into the nested verification function, obtaining a corresponding plurality of Boolean values, and determining the validity of the data according to the plurality of Boolean values; Classification and labeling unit: classify and label the data according to the validity; Verification unit: Verify the format and logic of the data based on the classification and labeling results and the data format rules and logic verification rules.

[0010] According to an intelligent data reporting system provided by the present invention, a detection module includes: A fourth acquisition unit: analyzing the automatic filling suggestion and the verification result to acquire each step and required variables; Reporting unit: reporting the target data according to each step and required variables; Detection unit: pre-process the reported data based on natural language processing, and detect potential errors and inconsistencies in the reported data based on the pre-processing results.

[0011] According to the present invention, an intelligent data reporting system and a correction module are provided, comprising: Analysis unit: analyzing the errors and inconsistencies, and obtaining the causes of the problems according to the analysis results; A fourth determining unit: determining the problem type according to the cause of the problem; Correction unit: determines a corresponding repair solution according to the problem type, corrects the errors and inconsistencies according to the repair solution, and displays the corrected data in a visual manner.

[0012] According to an intelligent data reporting system provided by the present invention, a definition unit includes: The first determination subunit: determines the type of reporting data format according to the requirements of the reporting system, and defines the data format rules according to the type of reporting data format and the conversion parameters between various data formats; The second determination subunit: determines the condition parameters to be met by the reported content, and determines the verification logic conditions according to the condition parameters to be met; Generate subunit: Generate logical validation rules based on validation logic conditions.

[0013] An intelligent data reporting system provided by the present invention further includes: The second determination module: obtains the basic information of the reported data, and determines the macro-dimensional data labels and micro-dimensional data labels according to the basic information; The second acquisition module: obtains the label data corresponding to the macro-dimensional data label and the micro-dimensional data label and integrates them to obtain a label data set; Training module: preprocess the labeled data set and use the preprocessed labeled data set to train the logistic regression model; Generation module: obtains data screening quantitative indicators and data screening qualitative indicators through the trained logistic regression model, and generates data screening rules according to the data screening quantitative indicators and data screening qualitative indicators; Retrieval module: connects to the big database in a preset way and generates data retrieval requests based on the data retrieval function of the big database; The third acquisition module: serializes the data screening rules to generate a rule file, imports the rule file into the trigger server of the database, searches the database through the trigger server, and obtains multiple data folders; Setting module: Set the data storage path, extract source data from multiple data folders and store them in the data storage path; The third determination module: obtain the business object of each source data, determine the business attributes of the business object, determine the consistency between the business attributes and the basic information of the reported content, and select the target source data as the data to be retrieved for reporting based on the consistency.

[0014] Compared with the prior art, the beneficial effects of this application are as follows: By determining the technology stack based on the data characteristics and data types of the target data, combining the existing data with filling prompts, obtaining automatic filling suggestions, and reporting the target data based on the multi-layer verification mechanism, the reporting errors are determined, which can improve reporting efficiency and reduce labor costs. At the same time, adding a multi-layer verification mechanism can streamline the reporting process and improve the accuracy and precision of the data. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 It is a structural diagram of an intelligent data reporting system provided by an embodiment of the present invention; Figure 2 It is a structural diagram of a first determination module provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] Embodiment 1: An intelligent data reporting system provided by an embodiment of the present invention is as follows: Figure 1 As shown, the method mainly includes the following steps: The first determination module: obtains data characteristics and data types of target data, and determines the corresponding technology stack according to the data characteristics and data types; A first acquisition module: acquiring auto-fill suggestions based on the filling prompt according to the technology stack; Verification module: setting up multiple levels of data verification mechanisms, and verifying the format and logic of data according to the verification mechanisms; Detection module: filling in the target data according to the automatic filling suggestions and verification results, and detecting potential errors and inconsistencies based on artificial intelligence technology; Correction module: correct the errors and inconsistencies, and display the data in a visual way to achieve intelligent data reporting.

[0019] In this embodiment, the data characteristics include the frequency and format of the data.

[0020] In this embodiment, the technology stack can use Python programming language and deep learning frameworks such as TensorFlow / PyTorch to implement functions such as automatic completion and prediction.

[0021] In this embodiment, auto-fill is a predictive technique that fills in default values ​​for missing values ​​by analyzing existing data sets and patterns.

[0022] In this embodiment, the data validation mechanism is a mechanism for detecting errors and anomalies in input data and providing feedback and suggestions so that users can correct errors and make corrections.

[0023] In this embodiment, artificial intelligence technology includes: machine learning and natural language processing.

[0024] The beneficial effects of the above technical solution are: determining the technology stack through the data characteristics and data types of the target data, combining the existing data and filling prompts to obtain automatic filling suggestions, and filling in the target data according to the multi-layer verification mechanism to determine the filling errors, which can improve the filling efficiency and reduce labor costs. At the same time, adding a multi-layer verification mechanism can make the filling process streamlined and improve the accuracy and precision of the data.

[0025] Embodiment 2: Based on Embodiment 1, the first determining module of the embodiment of the present invention includes: A first acquisition unit: analyzing the target data to acquire data distribution and development trend of the target data; A second acquisition unit: acquiring data features of target data according to the data distribution and development trend; Extraction unit: extracts keywords of target data according to a preset method, and determines detailed information of target data according to the data source; A first determining unit: determining the data type of the target data according to the keyword and detailed information of the data; The third acquisition unit: obtains data processing requirements and storage requirements, and obtains data scale according to data characteristics and data types; The second determining unit determines a corresponding technology stack based on the data processing requirements, storage requirements, and data scale.

[0026] In this embodiment, the data feature of the data refers to useful information describing the data set, such as the distribution of the data and the degree of discreteness of the data.

[0027] In this embodiment, there are usually multiple ways to extract keywords of target data, such as: Rule-based approach: Set keywords for each category by manually writing rules or referring to an existing rule library.

[0028] Machine learning-based methods: Use existing machine learning algorithms (such as SVM, Naive Bayes, decision tree, etc.) to classify the target data and extract keywords from it.

[0029] Deep learning-based methods: Use deep learning models (such as neural networks) to classify target data and extract keywords from them.

[0030] In this embodiment, the data source refers to the channels and means for obtaining data, such as: internal questionnaires, public databases, social media platforms, and government public data.

[0031] In this embodiment, the detailed information may include the format, metadata, and column description of the data.

[0032] In this embodiment, the data type refers to the object or attribute represented by the data, such as: numerical type, Boolean type, text type, date type.

[0033] In this embodiment, data scale refers to the size or quantity of data, which is usually measured in bytes, megabytes, or gigabytes.

[0034] The beneficial effects of the above technical solution are: obtaining data characteristics of target data through data distribution and development trends, determining the data type of target data based on keywords and detailed information of the data, thereby obtaining the data scale, and determining the corresponding technology stack in combination with data processing requirements and storage requirements. Determining the correct data characteristics and types can avoid unnecessary data conversion, cleaning and preprocessing tasks, thereby improving work efficiency. At the same time, obtaining the corresponding technology stack can ensure that the results of data processing are more reliable and accurate, and improve data quality.

[0035] Embodiment 3: Based on Example 2, the first acquisition module of the embodiment of the present invention includes: Processing unit: Use the word segmentation model to perform word segmentation on the existing data and filled-in prompts to obtain words or character sequences; Conversion unit: converts each generated word or character sequence into the corresponding word embedding vector; Learning unit: learning the mapping relationship between the word embedding vector and the corresponding label by establishing a neural network model; Prediction unit: predicts possible labels of target data through a neural network model based on the technology stack; Selection unit: Based on the probability distribution of each label in the prediction results, select the label with the highest probability distribution as the auto-fill suggestion; If multiple labels have the same Gaussian kernel density, one of the labels can be randomly selected.

[0036] In this embodiment, the word segmentation model is an algorithm for analyzing words or subwords in text or spoken language, which can automatically segment the input text into meaningful words or phrases and retain the grammatical and semantic information of the original text as much as possible.

[0037] In this embodiment, the character sequence refers to a continuous sequence consisting of a series of characters, which may be a simple text consisting of a single character, or a sentence, paragraph, or document consisting of multiple characters.

[0038] In this embodiment, the word embedding vector is a technology that maps words to a high-dimensional space, which can convert words into numerical vectors and preserve the semantic relationship between words in the vector space.

[0039] In this embodiment, the mapping relationship between the word embedding vector and the corresponding label refers to a method of associating a word with a specific label. Generally, there is a simple linear relationship between the word embedding vector and the corresponding label, that is, the mapping of a word to a label can be represented by a scalar, which is called "label bias" or "label weight". For example, if a word "dog" has a corresponding label "animal", their mapping relationship can be expressed as [0.5, 0.7], where 0.5 represents a close relationship between the word and the "animal" label, and 0.7 represents a closer relationship between the word and the "plant" label.

[0040] In this embodiment, the technology stack can use Python programming language and deep learning frameworks such as TensorFlow / PyTorch to implement functions such as automatic completion and prediction.

[0041] In this embodiment, Gaussian kernel density is an algorithm for calculating data similarity. The Gaussian kernel density algorithm converts the similarity measure in two data sets into an integral form of a kernel function, and then integrates the two data sets to calculate the similarity between them.

[0042] The beneficial effects of the above technical solution are: performing word segmentation on existing data and filling prompts, obtaining words or character sequences, and converting them into word embedding vectors, learning the mapping relationship between word embedding vectors and labels based on the model, predicting the possible labels of the target data through a neural network model based on the technology stack, and selecting the most likely label as the automatic filling suggestion, which can improve work efficiency and accuracy, avoid duplication of work, and reduce the error rate.

[0043] Embodiment 4: Based on Example 3, the inspection module of the embodiment of the present invention includes: Definition unit: define data format rules and logic validation rules according to the requirements of the reporting system; The third determination unit is: writing a corresponding nested verification function, and inputting the target data into the nested verification function, obtaining a corresponding plurality of Boolean values, and determining the validity of the data according to the plurality of Boolean values; Classification and labeling unit: classify and label the data according to the validity; Verification unit: Verify the format and logic of the data based on the classification and labeling results and the data format rules and logic verification rules.

[0044] In this embodiment, defining data format rules is to describe the structure and constraints of data, including: Data type: Determines the type of data, such as integer, floating point number, string, etc.

[0045] Data Length: Specify the maximum or minimum length of each data item.

[0046] Data format: describes how data is represented, such as date format, time format, currency format, etc.

[0047] In this embodiment, the logic validation rules are defined as follows: Input format verification: Check whether the input data is submitted in accordance with the specified format requirements, for example, only numbers and letters can be entered in the text box.

[0048] Data integrity check: Ensure that the input data has not been maliciously tampered with or modified, such as by checking through code whether the input value is within a predetermined range.

[0049] Non-empty validation: Ensure that the input data cannot be empty, because empty data may cause unknown errors.

[0050] In this embodiment, the nested verification function is used to call another verification function inside a verification function to complete a more complex logic verification task.

[0051] In this embodiment, a Boolean value is a data type representing true and false values ​​in computer programming. A Boolean value usually uses the number 0 to represent a false value and the number 1 to represent a true value.

[0052] In this embodiment, data classification is a process of grouping a set of data according to their common features or attributes. The purpose of classification is to divide the data into different categories or groups.

[0053] The beneficial effects of the above technical solution are: obtaining the corresponding Boolean value according to the nested verification function to determine the validity of the data, classifying and labeling the valid data, and checking the format and logic of the data based on the data format rules and logical verification rules, which can ensure that all data is formatted according to the same standards, and check the consistency of the data to ensure the accuracy of the data. At the same time, the verification mechanism can help identify malicious input and improve the security of the system.

[0054] Embodiment 5: Based on Example 4, the detection module of the embodiment of the present invention includes: A fourth acquisition unit: analyzing the automatic filling suggestion and the verification result to acquire each step and required variables; Reporting unit: reporting the target data according to each step and required variables; Detection unit: pre-process the reported data based on natural language processing, and detect potential errors and inconsistencies in the reported data based on the pre-processing results.

[0055] In this embodiment, each step of the auto-fill suggestion includes, for example: Collecting User Information: Users will be asked to enter details related to their personal information such as name, contact information, email address, etc.

[0056] Identify goal: Based on the information provided by the user, the system will select an appropriate suggested goal, such as losing weight, saving money, or investing.

[0057] Provide personalized recommendations: The system will provide personalized recommendations based on the user's personal information and goals. This may include recommending specific exercise plans, budget management tools, or investment strategies.

[0058] In this embodiment, the variables required for the automatic filling suggestion generally include, for example: name, address, company, and position.

[0059] In this embodiment, natural language processing is the process of enabling a computer to process human text input and perform tasks related to human language, such as answering questions, generating text, or conducting conversations.

[0060] In this embodiment, preprocessing may include: word segmentation, part-of-speech tagging, named entity recognition, and semantic analysis.

[0061] The beneficial effects of the above technical solution are: filling in the target data according to the automatic filling suggestions and verification results, and detecting potential errors and inconsistencies based on artificial intelligence technology, which can ensure the speed of the filling process and the accuracy of the filling results, and improve the filling efficiency and the accuracy of the reported data.

[0062] Embodiment 6: Based on Example 5, the correction module of this embodiment of the present invention includes: Analysis unit: analyzing the errors and inconsistencies, and obtaining the causes of the problems according to the analysis results; A fourth determining unit: determining the problem type according to the cause of the problem; Correction unit: determines a corresponding repair solution according to the problem type, corrects the errors and inconsistencies according to the repair solution, and displays the corrected data in a visual manner.

[0063] In this embodiment, the cause may be: programming error, input error, algorithm error.

[0064] In this embodiment, the problem type may be: data input error, calculation error, logic error.

[0065] In this embodiment, the repair solution may be: code modification, input verification, and algorithm optimization.

[0066] The beneficial effects of the above technical solution are: analyzing the errors, obtaining the problem type, determining the repair plan according to the problem type, repairing the errors, and displaying the corrected data in a visual way, which can improve the quality and integrity of the data, reduce the time for manual modification, and improve the efficiency of the reporting process.

[0067] Embodiment 7: Based on Example 6, the definition unit of this embodiment of the present invention includes: The first determination subunit: determines the type of reporting data format according to the requirements of the reporting system, and defines the data format rules according to the type of reporting data format and the conversion parameters between various data formats; The second determination subunit: determines the condition parameters to be met by the reported content, and determines the verification logic conditions according to the condition parameters to be met; Generate subunit: Generate logical validation rules based on validation logic conditions.

[0068] In this embodiment, the types of reporting data formats include: Structured data formats: such as tables, database records, etc., in which data is organized and presented in a predetermined structure and columns.

[0069] Unstructured data formats: such as text files, pictures, audio, video, etc., in which data is stored in an unstructured form without a predetermined format and structure.

[0070] Binary data formats: such as bytecode, executable files, machine code, etc., where data is stored and transmitted in binary form.

[0071] Graphic data format: such as vector graphics, bitmap images, etc., in which data is represented in a graphic format.

[0072] In this embodiment, conversion between different data formats requires the use of specific conversion parameters, which determine how to convert one data format into another data format, including: field name, data type, primary key, cell reference, and data type.

[0073] In this embodiment, the condition parameters to be satisfied include: Filter condition parameter: specifies the filter condition to be applied, for example, a filter of "only containing specific text".

[0074] Sort Criteria parameter: Specifies the sort criteria to be applied. For example, a sorter of "sort by rating descending".

[0075] Grouping Condition Parameter: Specifies the grouping condition to be applied. For example, a grouper of "Group by Region".

[0076] The beneficial effects of the above technical solution are: defining data format rules according to the types of reported data formats and the conversion parameters between various data formats, and determining verification logic conditions according to the conditional parameters to be met by the reported content, thereby generating logical verification rules, which can ensure the consistency and accuracy of the rules, thereby avoiding data inconsistency problems caused by human errors and improving efficiency and accuracy.

[0077] Embodiment 8: Based on Example 7, the embodiment of the present invention further includes: The second determination module: obtains the basic information of the reported data, and determines the macro-dimensional data labels and micro-dimensional data labels according to the basic information; The second acquisition module: obtains the label data corresponding to the macro-dimensional data label and the micro-dimensional data label and integrates them to obtain a label data set; Training module: preprocess the labeled data set and use the preprocessed labeled data set to train the logistic regression model; Generation module: obtains data screening quantitative indicators and data screening qualitative indicators through the trained logistic regression model, and generates data screening rules according to the data screening quantitative indicators and data screening qualitative indicators; Retrieval module: connects to the big database in a preset way and generates data retrieval requests based on the data retrieval function of the big database; The third acquisition module: serializes the data screening rules to generate a rule file, imports the rule file into the trigger server of the database, searches the database through the trigger server, and obtains multiple data folders; Setting module: Set the data storage path, extract source data from multiple data folders and store them in the data storage path; The third determination module: obtain the business object of each source data, determine the business attributes of the business object, determine the consistency between the business attributes and the basic information of the reported content, and select the target source data as the data to be retrieved for reporting based on the consistency.

[0078] In this embodiment, the macro-dimensional data labels of the reported data refer to labels added to the chart after the data is classified, grouped and summarized, such as: category labels, value labels, percentage labels, and total labels.

[0079] In this embodiment, the micro-dimensional data labels of the reported data refer to labels used for more fine-grained analysis and processing of the data, such as custom labels, subcategory labels, filter labels, and calculation labels.

[0080] In this embodiment, the labeled data set refers to a collection of data related to a certain topic, where each data has a relevant label.

[0081] In this embodiment, the logistic regression model is a supervised learning algorithm used in classification and regression tasks. The model is trained to learn the logical relationship between input features and output results, and this knowledge is used to classify or predict new inputs.

[0082] In this embodiment, the data screening quantitative indicators are generally used to screen and evaluate a set of data to determine which data points meet specific criteria or thresholds. These criteria can be based on various factors, such as numerical size, percentage, statistics, outliers, etc., such as: mean, median, mode, standard deviation, etc.

[0083] In this embodiment, data screening qualitative indicators are typically used to screen and evaluate a set of data to determine which data points meet specific classification or text criteria. These criteria can be based on semantic, logical, grammatical, common sense and other factors, such as the occurrence of certain words, the range of certain numbers, and the characteristics of certain texts, such as: regular expression matching, bag of words model, natural language processing rule matching, and word segmenter.

[0084] In this embodiment, the way to connect to the large database can be: using SQL direct connection, using ODBC or JDBC driver.

[0085] In this embodiment, the data retrieval function refers to the system capability of acquiring and processing data from a database or other data source.

[0086] In this embodiment, the purpose of serialization of the data screening rule is to convert the data into a character string in a specific format so as to be used in the program.

[0087] In this embodiment, the data storage path refers to where the data is stored, such as: a local file system, a network server, a cloud storage or other data storage platform.

[0088] In this embodiment, the business object of the source data refers to the actual business entity or concept related to the source data. For example, in an order management data, the source data may include order number, customer name, product name, etc., and the business object may include order, customer, product, etc.

[0089] In this embodiment, a business attribute refers to a specific feature or property that describes a business entity. For example, in an order management data, an order entity may have the following business attributes: order number, customer number, order date, and order status.

[0090] The beneficial effects of the above technical solution are: obtaining the business attributes of the business objects of each source data, determining the degree of consistency between the business attributes and the basic information of the reported content, and selecting the target source data as the data to be retrieved for reporting based on the degree of consistency, which can improve the degree of automation of filling in the reporting form. Automation can help reduce the possibility of manual input errors, ensure the accuracy and completeness of the reported data, and save a lot of time.

[0091] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0092] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent data reporting system, characterized in that: include: The first determination module: obtains data characteristics and data types of target data, and determines the corresponding technology stack according to the data characteristics and data types; A first acquisition module: acquiring auto-fill suggestions based on the filling prompt according to the technology stack; Verification module: setting up multiple levels of data verification mechanisms, and verifying the format and logic of data according to the verification mechanisms; Detection module: filling in the target data according to the automatic filling suggestions and verification results, and detecting potential errors and inconsistencies based on artificial intelligence technology; Correction module: correct the errors and inconsistencies, and display the data in a visual way to achieve intelligent data reporting.

2. The intelligent data reporting system according to claim 1, characterized in that: The first determination module includes: A first acquisition unit: analyzing the target data to acquire data distribution and development trend of the target data; A second acquisition unit: acquiring data features of target data according to the data distribution and development trend; Extraction unit: extracts keywords of target data according to a preset method, and determines detailed information of target data according to the data source; A first determining unit: determining the data type of the target data according to the keyword and detailed information of the data; The third acquisition unit: obtains data processing requirements and storage requirements, and obtains data scale according to data characteristics and data types; The second determining unit determines a corresponding technology stack based on the data processing requirements, storage requirements, and data scale.

3. The intelligent data reporting system according to claim 1, characterized in that: The first acquisition module includes: Processing unit: Use the word segmentation model to perform word segmentation on the existing data and filled-in prompts to obtain words or character sequences; Conversion unit: converts each generated word or character sequence into the corresponding word embedding vector; Learning unit: learning the mapping relationship between the word embedding vector and the corresponding label by establishing a neural network model; Prediction unit: predicts possible labels of target data through a neural network model based on the technology stack; Selection unit: Based on the probability distribution of each label in the prediction results, select the label with the highest probability distribution as the auto-fill suggestion; If multiple labels have the same Gaussian kernel density, one of the labels can be randomly selected.

4. The intelligent data reporting system according to claim 1, characterized in that: Inspection modules include: Definition unit: define data format rules and logic validation rules according to the requirements of the reporting system; The third determination unit is: writing a corresponding nested verification function, and inputting the target data into the nested verification function, obtaining a corresponding plurality of Boolean values, and determining the validity of the data according to the plurality of Boolean values; Classification and labeling unit: classify and label the data according to the validity; Verification unit: Verify the format and logic of the data based on the classification and labeling results and the data format rules and logic verification rules.

5. The intelligent data reporting system according to claim 1, characterized in that: Detection module, including: A fourth acquisition unit: analyzing the automatic filling suggestion and the verification result to acquire each step and required variables; Reporting unit: reporting the target data according to each step and required variables; Detection unit: pre-process the reported data based on natural language processing, and detect potential errors and inconsistencies in the reported data based on the pre-processing results.

6. The intelligent data reporting system according to claim 1, characterized in that: Correction modules, including: Analysis unit: analyzing the errors and inconsistencies, and obtaining the causes of the problems according to the analysis results; A fourth determining unit: determining the problem type according to the cause of the problem; Correction unit: determines a corresponding repair solution according to the problem type, corrects the errors and inconsistencies according to the repair solution, and displays the corrected data in a visual manner.

7. The intelligent data reporting system according to claim 4, characterized in that: Define the unit, including: The first determination subunit: determines the type of reporting data format according to the requirements of the reporting system, and defines the data format rules according to the type of reporting data format and the conversion parameters between various data formats; The second determination subunit: determines the condition parameters to be met by the reported content, and determines the verification logic conditions according to the condition parameters to be met; Generate subunit: Generate logical validation rules based on validation logic conditions.

8. The intelligent data reporting system according to claim 1, characterized in that: Also includes: The second determination module: obtains the basic information of the reported data, and determines the macro-dimensional data labels and micro-dimensional data labels according to the basic information; The second acquisition module: obtains the label data corresponding to the macro-dimensional data label and the micro-dimensional data label and integrates them to obtain a label data set; Training module: preprocess the labeled data set and use the preprocessed labeled data set to train the logistic regression model; Generation module: obtains data screening quantitative indicators and data screening qualitative indicators through the trained logistic regression model, and generates data screening rules based on the data screening quantitative indicators and data screening qualitative indicators; Retrieval module: connects to the big database in a preset way and generates data retrieval requests based on the data retrieval function of the big database; The third acquisition module: serializes the data screening rules to generate a rule file, imports the rule file into the trigger server of the database, searches the database through the trigger server, and obtains multiple data folders; Setting module: Set the data storage path, extract source data from multiple data folders and store them in the data storage path; The third determination module: obtain the business object of each source data, determine the business attributes of the business object, determine the consistency between the business attributes and the basic information of the reported content, and select the target source data as the data to be retrieved for reporting based on the consistency.

Images