Data credibility evaluation system and method for whole-process flow mining

Abstract

The invention relates to the technical field of data credibility evaluation, and discloses a whole-process flow mining data credibility evaluation system, which comprises a processor used for controlling the operation of the system; the data source credibility evaluation module is used for objectively evaluating the credibility of the data source; the data quality evaluation module is used for evaluating the consistency, integrity, accuracy and the like of the data by using objective data quality evaluation indexes; the data cleaning and preprocessing module is used for cleaning and preprocessing the mining data according to the evaluation result; and the data mining algorithm module is used for mining the processed data by using a data mining algorithm. Through objective data quality evaluation indexes, the consistency, integrity, accuracy and the like of the data are evaluated, and the influence of subjective judgment is reduced. And different judgments on the quality and credibility of the data due to subjective judgments of the user are avoided, so that the certainty of a system evaluation result is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data credibility evaluation, and in particular to a data credibility evaluation system and method for full-process flow mining. Background Art

[0002] The data credibility evaluation system refers to a system that evaluates the credibility of the data used in the data mining process. In the data mining process, the quality and credibility of the data are crucial to the accuracy and reliability of the mining results. The credibility evaluation system can help analysts better evaluate the credibility of the data used when performing data mining tasks, thereby improving the accuracy and credibility of data mining results. In the exploration of natural resources such as oil, natural gas, and minerals, a large amount of geological exploration, geophysical exploration, and geochemical exploration data is required. The data credibility evaluation system for the whole process mining can be used to evaluate the quality and credibility of these exploration data to ensure the accuracy and reliability of exploration decisions.

[0003] In the current mining data credibility evaluation system, data quality evaluation and credibility evaluation are often affected by the subjective judgment of analysts, and there is a certain degree of subjectivity and human factors. Different analysts may have different judgments on the quality and credibility of data, resulting in uncertainty in the evaluation results. To this end, the present invention proposes a data credibility evaluation system and method for the whole process of process mining. Summary of the invention

[0004] In view of the deficiencies of the prior art, the present invention provides a data credibility evaluation system and method for the whole process process mining, which solves the problem that the mining data credibility evaluation system in the prior art has different judgments on the quality and credibility of the data, resulting in uncertainty in the evaluation results.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a data credibility evaluation system for full-process process mining, comprising:

[0006] A processor for controlling the operation of the system;

[0007] Data source credibility evaluation module, used to objectively evaluate the credibility of data sources;

[0008] Data quality evaluation module, which is used to evaluate the consistency, completeness, accuracy, etc. of data using objective data quality evaluation indicators;

[0009] Data cleaning and preprocessing module, used to clean and preprocess the mining data according to the evaluation results;

[0010] A data mining algorithm module is used to mine the processed data using a data mining algorithm;

[0011] The mining result verification module is used to verify and evaluate the mining results.

[0012] Preferably, the data source credibility evaluation module includes:

[0013] A data collection method analysis unit, used for analyzing the data collection method;

[0014] Data source verification unit, used to verify the source and provider of data;

[0015] The data credibility assessment unit is used to comprehensively consider the evaluation results of data collection methods, data sources, data integrity and data accuracy.

[0016] Preferably, the data quality assessment module includes:

[0017] Data consistency assessment unit, used to assess data consistency and check whether data is consistent across different dimensions or data sources;

[0018] Data integrity assessment unit, used to assess the integrity of data and check whether the data is missing important information;

[0019] Data accuracy assessment unit, which is used to assess the accuracy of data and check whether the data contains errors, deviations or inconsistencies;

[0020] Data consistency repair unit, used to clean and repair data consistency issues;

[0021] Data integrity repair unit, used to clean and fill data for data integrity issues;

[0022] The data accuracy repair unit is used to clean and repair data for data accuracy issues.

[0023] Preferably, the data cleaning and preprocessing module includes:

[0024] An outlier identification unit, used to identify outliers in the data and mark these outliers;

[0025] An outlier processing unit, used for processing the identified outliers;

[0026] Missing value identification unit, used to identify missing values ​​in the data and mark these missing values;

[0027] A missing value filling unit is used to fill the identified missing values;

[0028] The data normalization unit is used to normalize the data, map the data into a unified numerical range, and eliminate the dimensional impact between different indicators;

[0029] The data conversion unit is used to convert the data.

[0030] Preferably, the data mining algorithm module includes:

[0031] Business requirements analysis unit, used to analyze users' business requirements and explore targets, and determine the problems and targets that need to be solved;

[0032] Algorithm selection unit, used to use data mining algorithms according to business requirements and data characteristics;

[0033] Parameter optimization unit, used to optimize and adjust the parameters of the data mining algorithm to obtain the best algorithm performance;

[0034] A model evaluation unit, used to evaluate the algorithm model after parameter optimization;

[0035] The algorithm integration unit is used to integrate multiple algorithm results to improve the accuracy and robustness of mining.

[0036] Preferably, the mining result verification module includes:

[0037] The data preprocessing unit is used to preprocess the mining results before verifying and evaluating the results;

[0038] A mining result comparison and verification unit is used to compare the mining results with the actual situation and verify the accuracy of the mining results;

[0039] The result interpretation and analysis unit is used to interpret and analyze the mining results and convert the mining results into a language or form that can be understood by the business.

[0040] A data credibility evaluation method for whole-process process mining includes the following steps:

[0041] S1. First, collect information about the data source, and then apply objective evaluation indicators and algorithms to evaluate the credibility of the data source based on the collected information;

[0042] S2. Objectively evaluate the quality of data based on objective data quality evaluation indicators and statistical methods;

[0043] S3. Based on the data quality evaluation results, data cleaning and preprocessing are performed to convert the data into a form suitable for mining and analysis to improve the quality and credibility of the data;

[0044] S4. Use data mining algorithms based on the characteristics of the data and the results of quality evaluation, and adjust the algorithm parameters according to the credibility of the data to optimize the algorithm;

[0045] S5. Verify and evaluate the mining results to verify the credibility of the mining results.

[0046] Preferably, the evaluation indicators in the SI include:

[0047] Data completeness indicators,expressed as: missing value proportion;

[0048] Data accuracy indicators,expressed as: error value ratio;

[0049] Data consistency index,expressed as: data consistency score;

[0050] Data uniqueness indicators,expressed as: the proportion of duplicate records;

[0051] Data timeliness indicator, expressed as: data update frequency.

[0052] Preferably, the objective evaluation in S2 is: analyzing based on the indicators calculated in S1, evaluating the performance of the data in various quality aspects, comprehensively considering the results of various indicators, objectively evaluating the overall quality of the data, and forming a data quality report or evaluation conclusion.

[0053] Preferably, the data mining algorithm in S4 is a decision tree, and the splitting criterion information gain adopted is the Gini index.

[0054] The present invention provides a data credibility evaluation system and method for whole-process process mining. It has the following beneficial effects:

[0055] 1. The present invention uses objective data quality evaluation indicators to evaluate the consistency, completeness, accuracy, etc. of the data, reducing the impact of subjective judgment. It avoids users from making different judgments on the quality and credibility of the data due to subjective judgment, thereby improving the certainty of the system evaluation results.

[0056] 2. The present invention improves the accuracy and flexibility of the system in selecting algorithms in different scenarios by constructing a mining algorithm model and performing parameter optimization on the decision tree, and integrates multiple algorithm results to improve the accuracy and robustness of mining. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 is a system architecture diagram of the present invention;

[0058] Figure 2 This is a schematic diagram of a data source credibility evaluation module of the present invention;

[0059] Figure 3 It is a schematic diagram of a data quality evaluation module of the present invention;

[0060] Figure 4 It is a schematic diagram of the data cleaning and preprocessing module of the present invention;

[0061] Figure 5 It is a schematic diagram of the data mining algorithm module of the present invention;

[0062] Figure 6 It is a schematic diagram of a mining result verification module of the present invention;

[0063] Figure 7 is a credibility evaluation flow chart of the present invention;

[0064] Figure 8 The present invention is a system operation flow chart. DETAILED DESCRIPTION

[0065] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments 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.

[0066] Example:

[0067] Please refer to the attached Figure 1 -Attached Figure 7 The embodiment of the present invention provides a data credibility evaluation system for whole-process process mining, including:

[0068] A processor for controlling the operation of the system;

[0069] The data source credibility evaluation module is used to objectively evaluate the credibility of the data source, including:

[0070] A data collection method analysis unit, used for analyzing the data collection method;

[0071] Data source verification unit, used to verify the source and provider of data;

[0072] The data credibility assessment unit is used to comprehensively consider the evaluation results of data collection methods, data sources, data integrity and data accuracy.

[0073] Data Collection Method Analysis Unit: The main task of this unit is to conduct in-depth analysis of the data collection method. The data collection method directly affects the quality and credibility of the final data. Therefore, it is necessary to conduct detailed research and analysis on various collection methods in order to better understand the source and generation process of the data. This includes the evaluation of the performance of the data collection equipment, data sampling frequency, data transmission rate, etc.

[0074] Data source verification unit: In the data processing process, the source and provider of the data are crucial. The main responsibility of this unit is to verify whether the source of the data is legal and whether the provider has a certain reputation. The source of the data needs to be strictly reviewed to ensure the reliability and credibility of the data. In addition, the qualifications, historical data records, and data processing capabilities of the provider need to be evaluated.

[0075] Data credibility assessment unit: This unit is responsible for comprehensively considering the evaluation results of data collection methods, data sources, data integrity and data accuracy, and making an overall assessment of the credibility of the data. It is necessary to use a series of evaluation indicators and methods, such as data quality indicators, data similarity analysis, data error rate, etc., to comprehensively and objectively measure the credibility of the data.

[0076] The data quality evaluation module is used to evaluate the consistency, completeness, and accuracy of data using objective data quality evaluation indicators, including:

[0077] Data consistency assessment unit, used to assess data consistency and check whether data is consistent across different dimensions or data sources;

[0078] Data integrity assessment unit, used to assess the integrity of data and check whether the data is missing important information;

[0079] Data accuracy assessment unit, which is used to assess the accuracy of data and check whether the data contains errors, deviations or inconsistencies;

[0080] Data consistency repair unit, used to clean and repair data consistency issues;

[0081] Data integrity repair unit, used to clean and fill data for data integrity issues;

[0082] The data accuracy repair unit is used to clean and repair data for data accuracy issues.

[0083] The importance of data quality assessment module and its role in data management process

[0084] 1. Data consistency assessment unit: ensuring that data remains consistent across different dimensions or data sources

[0085] Data consistency is one of the important indicators for measuring data quality. It is necessary to ensure that data is consistent across dimensions or data sources. The data consistency assessment unit analyzes the correlation and dependency between data to check whether the data is consistent across dimensions or data sources. Once an inconsistency is found, it can be fed back to the data provider or data source in a timely manner so that appropriate measures can be taken to correct it.

[0086] 2. Data integrity assessment unit: ensuring that data contains no missing important information

[0087] Data integrity refers to whether the data contains all necessary information. The data integrity assessment unit is responsible for checking whether the data is missing important information, such as missing values, outliers, etc. For integrity issues, data cleaning and filling methods can be used to repair them to ensure the availability and credibility of the data.

[0088] 3. Data Accuracy Assessment Unit: Ensure that there are no errors, biases or inconsistencies in the data

[0089] Data accuracy is another key indicator of data quality. The data accuracy assessment unit is mainly responsible for identifying errors, deviations and inconsistencies in the data. By comparing and verifying the data, accuracy issues in the data are discovered. For accuracy issues, data cleaning, repair and re-collection can be used to ensure that the accuracy of the data meets the expected requirements.

[0090] 4. Data consistency repair unit: solve data consistency problems and improve data quality

[0091] The data consistency repair unit is mainly responsible for cleaning and repairing data consistency issues. By analyzing the correlation and dependency between data, the data with consistency issues is located. Next, through data cleaning, conversion and integration methods, data consistency issues are resolved and data quality is improved.

[0092] 5. Data integrity repair unit: solve data integrity issues and improve data quality

[0093] The data integrity repair unit is mainly responsible for data cleaning and filling of data integrity issues. Integrity issues can be discovered by checking whether the data is missing important information. For integrity issues, methods such as data filling, supplementation and collection can be used to improve data integrity, thereby ensuring that data quality is effectively guaranteed.

[0094] 6. Data Accuracy Repair Unit: Solve data accuracy issues and improve data quality

[0095] The data accuracy repair unit is mainly responsible for data cleaning and repair of data accuracy issues. Through operations such as comparison, verification and verification, accuracy issues in the data can be discovered. For accuracy issues, methods such as data cleaning, repair and re-collection can be used to improve data accuracy and ensure that data quality meets application requirements.

[0096] Summary: The important role of data quality evaluation module in data management process

[0097] The data quality assessment module detects and fixes data quality issues by evaluating the consistency, completeness and accuracy of data. This helps ensure the reliability and validity of data and provides high-quality data support for enterprises and organizations.

[0098] The data cleaning and preprocessing module is used to clean and preprocess the mining data according to the evaluation results, including:

[0099] An outlier identification unit, used to identify outliers in the data and mark these outliers;

[0100] An outlier processing unit, used for processing the identified outliers;

[0101] Missing value identification unit, used to identify missing values ​​in the data and mark these missing values;

[0102] A missing value filling unit is used to fill the identified missing values;

[0103] The data normalization unit is used to normalize the data, map the data into a unified numerical range, and eliminate the dimensional impact between different indicators;

[0104] The data conversion unit is used to convert the data.

[0105] 1. Outlier Identification Unit: Identify outliers in the data. Outliers refer to data points that differ greatly from normal data and may have a greater impact on the results of data analysis. By identifying these outliers, they are screened out from the data set for subsequent processing. Marking these outliers helps to pay more attention to them in subsequent processing, thereby reducing their impact on data analysis.

[0106] 2. Outlier processing unit: After outliers are identified, they need to be processed. Processing methods include deleting, retaining or filling out outliers. Deleting outliers may result in a reduction in the amount of data, but it is beneficial to maintain the quality of the data. Retaining outliers allows further analysis of the causes of the anomalies, thereby providing more valuable information for subsequent decision-making. Methods for filling out outliers include mean filling, median filling, and maximum filling, which ensure the integrity of the data and provide more samples for subsequent analysis.

[0107] 3. Missing value identification unit: Missing values ​​are a common problem in data, which will make the results of data analysis unreliable. The role of the missing value identification unit is to identify missing values ​​in the data and mark them. In this way, the missing value problem can be solved in a targeted manner in subsequent processing, improving the accuracy of data analysis.

[0108] 4. Missing value filling unit: After identifying the missing values, they need to be filled. Use algorithms such as interpolation and regression to fill them. The filled data better reflects the actual situation and provides a reliable basis for subsequent analysis.

[0109] 5. Data normalization unit: Different data indicators may have different dimensions, which may cause some troubles in data analysis. The function of the data normalization unit is to map the data into a unified numerical range to eliminate the dimensional impact between different indicators. Normalization methods include maximum and minimum value normalization, standard score normalization, etc. The data after normalization is comparable, which helps to analyze the data more accurately.

[0110] 6. Data conversion unit: The data conversion unit is responsible for converting the data to make it more suitable for subsequent analysis and mining. Conversion methods include data type conversion, data structure conversion, etc. For example, converting categorical data into numerical data, or converting multidimensional data into low-dimensional data. Data conversion helps improve the effect of data analysis and mine more valuable information.

[0111] The data mining algorithm module is used to mine the processed data using the data mining algorithm, including:

[0112] Business requirements analysis unit, used to analyze the user's business requirements and mining targets, determine the problems and goals that need to be solved, and provide guidance for subsequent algorithm selection;

[0113] The algorithm selection unit is used to mine data through decision trees using data mining algorithms according to business requirements and data characteristics. Its specific architecture is as follows;

[0114] 1. Tree nodes:

[0115] A decision tree consists of nodes, including internal nodes and leaf nodes. Internal nodes represent the judgment and splitting of features, and leaf nodes represent the final classification results.

[0116] 2. Splitting criteria:

[0117] In the process of building a decision tree, it is necessary to determine how to select the best features to split the nodes. The splitting criteria include information gain Gini index (CART algorithm).

[0118] Leaf node classification:

[0119] When the split of the decision tree reaches a certain condition, the node will become a leaf node and the final classification label will be assigned to the leaf node.

[0120] The basic splitting criterion is the Gini index, which has the following formula:

[0121] [\text{Gini}(t)=1-\sum_{i=1}^{c}p(i|t)^2]

[0122] Among them, (t) represents the current node, (c) represents the number of categories, and (p(i|t)) represents the proportion of samples belonging to category (i) in node (t). The closer the Gini coefficient is to 0, the higher the purity of the node; the closer it is to 1, the higher the impurity of the node.

[0123] The parameter optimization unit is used to optimize and adjust the parameters of the data mining algorithm to obtain the best algorithm performance; the specific optimization steps are:

[0124] 1. Parameter search range definition:

[0125] For the Gini index, usually no additional tuning parameters are required.

[0126] For the cost complexity pruning parameter (\alpha), it is necessary to define a parameter search range, a set of values ​​from 0 to 1.

[0127] 2. Determine the evaluation indicators:

[0128] Select model performance evaluation indicators, such as accuracy, recall, and F1 value, as the criteria for evaluating model performance.

[0129] 3. Parameter search method:

[0130] For the Gini index, no additional parameter search is required as the splitting criterion is already built into the decision tree algorithm.

[0131] For the cost complexity pruning parameter (\alpha), a grid search or random search method can be used to search for the best (\alpha) value within the defined parameter range.

[0132] 4. Cross Validation:

[0133] Each parameter combination was evaluated using a cross-validation approach to obtain a more robust estimate of the model performance.

[0134] 5. Model training and verification:

[0135] Train a decision tree model on the entire training set using the selected optimal parameter values.

[0136] The validation set is used to validate the trained model and evaluate the generalization ability of the model.

[0137] 6. Result analysis and adjustment:

[0138] Analyze the validation results and evaluate the performance of the model on unknown data.

[0139] According to the actual situation, the parameter range can be further adjusted and the above steps can be repeated until satisfactory model performance is obtained.

[0140] The model evaluation unit is used to evaluate the algorithm model after parameter optimization; in terms of the optimized algorithm model formula, the Gini index as the splitting criterion formula remains unchanged. For the cost complexity pruning parameter (\alpha), the optimized algorithm model formula is:

[0141] [C_\alpha(T)=C(T)+\alpha|T|]

[0142] In the formula, (C(T)) represents the fitting error of the subtree, and (|T|) represents the number of leaf nodes in the subtree. By adjusting the parameter (\alpha), the degree of tree pruning can be controlled to avoid overfitting.

[0143] The algorithm integration unit is used to integrate multiple algorithm results to improve the accuracy and robustness of mining. It integrates the data mining algorithm model and pruning parameter model constructed with the decision tree, thereby improving its performance in the data credibility evaluation system, further ensuring the accuracy of the mining algorithm, and improving the accuracy of the system's data credibility evaluation.

[0144] The mining result verification module is used to verify and evaluate the mining results, including:

[0145] The data preprocessing unit is used to preprocess the mining results before verifying and evaluating the results;

[0146] A mining result comparison unit is used to compare the mining result with the actual situation to verify the accuracy of the mining result;

[0147] The result interpretation and analysis unit is used to interpret and analyze the mining results and convert the mining results into a language or form that can be understood by the business.

[0148] First of all, the data preprocessing unit is the first step in the verification of mining results. Before the result verification and evaluation, the data preprocessing unit will perform a series of processing on the mining results to ensure the accuracy and completeness of the data. Including data cleaning, data conversion and data normalization. The goal of the data preprocessing unit is to obtain a clean, accurate and consistent data set, providing a solid foundation for the subsequent verification and evaluation of mining results.

[0149] Secondly, the mining result comparison unit is responsible for comparing the mining results with the actual situation to verify the accuracy of the mining results. This process usually includes two steps: first, comparing the mining results with the actual data to check whether the mining results are consistent with the actual situation; second, comparing the mining results with existing research results or theories to evaluate the innovation and reliability of the mining results. The working goal of the mining result comparison unit is to ensure the accuracy and credibility of the mining results.

[0150] Finally, the result interpretation and analysis unit is responsible for in-depth interpretation and analysis of the mining results. This process mainly includes two aspects: first, converting the mining results into a language or form that can be understood by the business, so that business personnel can better understand and use the mining results; second, interpreting the business rules and trends behind the mining results to provide valuable reference for business decision-making. The working goal of the result interpretation and analysis unit is to make the mining results more practical and valuable.

[0151] A data credibility evaluation method for whole-process process mining includes the following steps:

[0152] S1. First, collect information about the data source, and then use objective evaluation indicators and algorithms to evaluate the credibility of the data source based on the collected information; the evaluation indicators include:

[0153] Data integrity indicators, expressed as: missing value ratio; number of missing values / total number of records.

[0154] Statistical formula: Missing Value Ratio = (Number of missing values) / (Total number of records)

[0155] Data accuracy indicators are expressed as: error value ratio; number of error values / total number of records.

[0156] Statistical formula: Error Value Ratio = (Number of error values) / (Total number of records)

[0157] Data consistency index, expressed as: data consistency score; by comparing data from different data sources, the data consistency score is calculated.

[0158] Statistical formula: Consistency Score = (number of consistent records) / (total number of records)

[0159] Data uniqueness indicators are expressed as: duplicate record ratio; number of duplicate records / total number of records.

[0160] Statistical formula: Duplicate Record Ratio=(Number of duplicate records) / (Totalnumber of records)

[0161] Data timeliness index, expressed as: data update frequency, the frequency of data update per unit time.

[0162] Statistical formula: Data Update Frequency=(Number of updates within a timeperiod) / (Total time period)

[0163] S2. Objectively evaluate the quality of the data based on objective data quality evaluation indicators and statistical methods, analyze the indicators calculated based on S1, evaluate the performance of the data in various quality aspects, comprehensively consider the results of various indicators, objectively evaluate the overall quality of the data, and form a data quality report or evaluation conclusion;

[0164] S3. Based on the data quality evaluation results, data cleaning and preprocessing are performed to convert the data into a form suitable for mining and analysis to improve the quality and credibility of the data;

[0165] S4. According to the characteristics of the data and the quality evaluation results, a data mining algorithm is used. The data mining algorithm is a decision tree, and the information gain of the splitting criterion used is the Gini index. The algorithm parameters are adjusted according to the credibility of the data to optimize the algorithm;

[0166] S5. Verify and evaluate the mining results to verify the credibility of the mining results.

[0167] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A data credibility evaluation system for whole-process process mining, characterized in that: include: A processor for controlling the operation of the system; Data source credibility evaluation module, used to objectively evaluate the credibility of data sources; Data quality evaluation module, which is used to evaluate the consistency, completeness, accuracy, etc. of data using objective data quality evaluation indicators; Data cleaning and preprocessing module, used to clean and preprocess the mining data according to the evaluation results; A data mining algorithm module is used to mine the processed data using a data mining algorithm; The mining result verification module is used to verify and evaluate the mining results.

2. According to the data credibility evaluation system of the whole process mining according to claim 1, it is characterized in that: The data source credibility evaluation module includes: A data collection method analysis unit, used for analyzing the data collection method; Data source verification unit, used to verify the source and provider of data; The data credibility assessment unit is used to comprehensively consider the evaluation results of data collection methods, data sources, data integrity and data accuracy.

3. According to the data credibility evaluation system of the whole process mining of claim 1, it is characterized in that: The data quality evaluation module includes: Data consistency assessment unit, used to assess data consistency and check whether data is consistent across different dimensions or data sources; Data integrity assessment unit, used to assess the integrity of data and check whether the data is missing important information; Data accuracy assessment unit, which is used to assess the accuracy of data and check whether the data contains errors, deviations or inconsistencies; Data consistency repair unit, used to clean and repair data consistency issues; Data integrity repair unit, used to clean and fill data for data integrity issues; The data accuracy repair unit is used to clean and repair data for data accuracy issues.

4. According to the data credibility evaluation system of the whole process mining of claim 1, it is characterized in that: The data cleaning and preprocessing module includes: An outlier identification unit, used to identify outliers in the data and mark these outliers; An outlier processing unit, used for processing the identified outliers; Missing value identification unit, used to identify missing values ​​in the data and mark these missing values; A missing value filling unit is used to fill the identified missing values; The data normalization unit is used to normalize the data, map the data into a unified numerical range, and eliminate the dimensional impact between different indicators; The data conversion unit is used to convert the data.

5. The data credibility evaluation system for full-process process mining according to claim 1 is characterized in that: The data mining algorithm module includes: Business requirements analysis unit, used to analyze users' business requirements and explore targets, and determine the problems and targets that need to be solved; Algorithm selection unit, used to use data mining algorithms according to business requirements and data characteristics; Parameter optimization unit, used to optimize and adjust the parameters of the data mining algorithm to obtain the best algorithm performance; A model evaluation unit, used to evaluate the algorithm model after parameter optimization; The algorithm integration unit is used to integrate multiple algorithm results to improve the accuracy and robustness of mining.

6. The data credibility evaluation system for full-process process mining according to claim 1 is characterized in that: The mining result verification module includes: The data preprocessing unit is used to preprocess the mining results before verifying and evaluating the results; A mining result comparison unit is used to compare the mining result with the actual situation to verify the accuracy of the mining result; The result interpretation and analysis unit is used to interpret and analyze the mining results and convert the mining results into a language or form that can be understood by the business.

7. A data credibility evaluation method for whole-process process mining, characterized in that: A data credibility evaluation system for full-process process mining according to any one of claims 1 to 6, comprising the following steps: S1. First, collect information about the data source, and then apply objective evaluation indicators and algorithms to evaluate the credibility of the data source based on the collected information; S2. Objectively evaluate the quality of data based on objective data quality evaluation indicators and statistical methods; S3. Based on the data quality evaluation results, data cleaning and preprocessing are performed to convert the data into a form suitable for mining and analysis to improve the quality and credibility of the data; S4. Use data mining algorithms based on the characteristics of the data and the results of quality evaluation, and adjust the algorithm parameters according to the credibility of the data to optimize the algorithm; S5. Verify and evaluate the mining results to verify the credibility of the mining results.

8. The data credibility evaluation method for full-process flow mining according to claim 7 is characterized in that: The evaluation indicators in the SI include: Data completeness indicators,expressed as: missing value proportion; Data accuracy indicators,expressed as: error value ratio; Data consistency index,expressed as: data consistency score; Data uniqueness indicators,expressed as: the proportion of duplicate records; Data timeliness indicator, expressed as: data update frequency.

9. The data credibility evaluation method for full-process flow mining according to claim 7 is characterized in that: The objective evaluation in S2 is: to analyze the indicators calculated in S1, evaluate the performance of the data in various quality aspects, comprehensively consider the results of various indicators, objectively evaluate the overall quality of the data, and form a data quality report or evaluation conclusion.

10. The data credibility evaluation method for full-process flow mining according to claim 7 is characterized in that: The data mining algorithm in S4 is a decision tree, and the information gain of the splitting criterion used is the Gini index.

Images