Visual modeling method and system based on process mining

Abstract

The invention belongs to the technical field of process modeling, and particularly relates to a visual modeling method and system based on process mining. The method comprises the following steps: S1, collecting process data from a log file; s2, cleaning and preprocessing the collected process data; s3, performing visual modeling by using the process data processed in the step S2; the visual modeling provides a graphical interface, and a user creates and modifies a data processing model through dragging and editing operations; and S4, the user exports the constructed data processing model as a file for subsequent use or sharing. The method has the advantages of being visual, easy to understand, flexible, efficient and high in reusability.

Description

Technical Field

[0001] The present invention belongs to the technical field of process modeling, and in particular relates to a visual modeling method and system based on process mining. Background Art

[0002] In traditional process modeling, manual input is not only time-consuming and labor-intensive, but also prone to errors. Although existing process mining tools can extract process models from log data, they often lack intuitive visualization interfaces, making it difficult for non-professional users to understand and modify models. In addition, although some visualization tools provide graphical interfaces, they fail to integrate closely with process mining technology and cannot achieve automated modeling processes. Existing solutions have deficiencies in user-friendliness, flexibility, degree of automation, and processing speed, as shown in the following: 1. Rule-based automated modeling: Although it has a high degree of automation and fast processing speed, the rule definition and maintenance are complex and lack flexibility.

[0003] 2. Machine learning-assisted modeling: It can process complex data, but requires a large amount of training data, has poor model interpretability, and a complex training process.

[0004] 3. Traditional process mining tools: The technology is mature and the processing speed is fast, but it lacks a visual interface and requires a high level of technical skills.

[0005] Based on the above problems, it is very important to design an intuitive, flexible and automated visual modeling method and system based on process mining. Summary of the invention

[0006] The present invention aims to overcome the problems of poor flexibility, poor model interpretability, complex training process and lack of visual interface in the existing process modeling methods in the prior art, and provides an intuitive, flexible and automated visual modeling method and system based on process mining.

[0007] In order to achieve the above-mentioned object of the invention, the present invention adopts the following technical solutions: The visual modeling method based on process mining includes the following steps: S1, collects process data from log files; S2, cleaning and preprocessing the collected process data; S3, using the process data processed in step S2 to perform visual modeling; the visual modeling provides a graphical interface, and the user can create and modify the data processing model through dragging and editing operations; S4, the user exports the constructed data processing model as a file for subsequent use or sharing; Step S3 includes the following steps: S31, select data source or reference existing model: the user selects the required data source or references the previously saved existing model through the graphical interface to provide the basic data set; S32, data integration: according to the data source or existing model selected in step S31, a data node is generated, and the user performs subsequent operations based on the data node to form a data set in series; the subsequent operations include field management, table association and table merging; S33, operation node: using the data set integrated in step S32, the user performs specific data processing tasks in the operation node step to obtain a customized result table, which will be used for final data analysis and decision support; the data processing tasks include data filtering, sorting and aggregation.

[0008] Preferably, in step S2, cleaning the collected process data includes missing value processing, abnormal value processing, duplicate value processing and erroneous data processing; The missing value processing refers to deleting, filling or predicting missing values; The outlier processing refers to identifying and processing outliers; the processing methods include deletion, modification and marking; The duplicate value processing refers to deleting duplicate records; The error data processing refers to correcting errors in the data set.

[0009] Preferably, in step S2, preprocessing the collected process data includes data integration, data conversion, data normalization, data dimension reduction and data enhancement; The data integration refers to merging multiple data sources; The data conversion refers to performing logarithmic transformation, normalization and discretization operations; The data standardization refers to the unification of data formats and units; The data dimensionality reduction refers to reducing the data dimension by aggregation, reduction and compression methods; The data enhancement refers to adding additional information to enrich the dataset.

[0010] Preferably, in step S32, the field management refers to the user managing the fields in the data node, including adding, deleting and renaming operations.

[0011] Preferably, in step S32, the table association refers to associating multiple data nodes to form a new data set, which specifically includes merging two tables through a common field.

[0012] Preferably, in step S32, the table merging refers to merging multiple data nodes into a single data set; the merging method includes vertical merging or horizontal merging.

[0013] Preferably, in step S32, the data set has the following characteristics: Comprehensiveness: Specifically, comprehensiveness refers to merging data sets from different sources through field management, table association, and table merging operations to cover all relevant information required for user analysis; Consistency, specifically refers to ensuring the uniformity of field names, formats, and units through field management.

[0014] The present invention also provides a visual modeling system based on process mining, including: The collection module is used to collect process data from log files; The preprocessing module is used to clean and preprocess the collected process data; A visual modeling module is used to perform visual modeling using the process data processed by the preprocessing module; the visual modeling provides a graphical interface, and the user can create and modify the data processing model through dragging and editing operations; The export module is used to export the constructed data processing model as a file.

[0015] Compared with the prior art, the present invention has the following advantages: (1) intuitive and easy to understand: through the graphical interface, the present invention allows users to understand each step of data processing more intuitively; (2) flexible and efficient: the present invention allows users to easily adjust the data processing flow with low trial and error costs; (3) strong reusability: the model constructed by the present invention can be reused multiple times, thereby improving work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flowchart of a visual modeling method based on process mining in the present invention; Figure 2 A schematic diagram of an interface for selecting a data source or referencing an existing model process in the present invention; Figure 3 A schematic diagram of an interface of the data integration process in the present invention; Figure 4 A schematic diagram of an interface for operating nodes in the present invention; Figure 5 A schematic diagram of an interface for storing and exporting a data processing model in the present invention; Figure 6 A schematic diagram of a modeling process in practical application of a visual modeling method based on process mining provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to more clearly illustrate the embodiments of the present invention, the specific implementation methods of the present invention will be described below with reference to the accompanying drawings. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those of ordinary skill in the art, other accompanying drawings and other implementation methods can be obtained based on these accompanying drawings without creative work.

[0018] like Figure 1 As shown, the visual modeling method based on process mining of the present invention includes the following steps: 1. Collect process data from log files; 2. Clean and preprocess the collected process data to ensure data quality and consistency; 3. Perform visual modeling using the process data processed in step 2; the visual modeling provides a graphical interface, and the user can create and modify the data processing model through dragging and editing operations; 4. The user exports the constructed data processing model as a file for subsequent use or sharing.

[0019] Furthermore, in step 2, the collected process data is cleaned including missing value processing, abnormal value processing, duplicate value processing and erroneous data processing; Missing value processing refers to deleting, filling or predicting missing values ​​to ensure data integrity; Outlier processing refers to identifying and processing outliers to avoid adverse effects on analysis results; The processing methods include but are not limited to: Deletion: Delete abnormal data records that are seriously deviated from the normal range and have no business value; Modification: Reasonably replace or modify outliers according to business rules or algorithms (such as median interpolation, neighbor value filling, etc.) to ensure the integrity and accuracy of the data; Marking: For outliers that may have analytical value, add special marks to facilitate subsequent analysis; Duplicate value processing refers to deleting duplicate records to improve the uniqueness and accuracy of data; Error data processing refers to correcting errors in data sets and improving data reliability.

[0020] Preprocessing of collected process data includes data integration, data conversion, data normalization, data dimension reduction and data enhancement; Data integration refers to merging multiple data sources to solve entity recognition and attribute redundancy problems; Data transformation refers to performing logarithmic transformation, normalization, and discretization operations to eliminate data inconsistencies; Data normalization refers to unifying data formats and units to ensure data consistency; Data dimensionality reduction refers to reducing data dimensions through aggregation, reduction, and compression methods to improve analysis efficiency; Data augmentation refers to adding additional information to enrich the dataset, such as adding geolocation information through an external database.

[0021] Furthermore, in the visual modeling stage, the specific steps are as follows: 3-1. Select a data source or reference an existing model: Figure 2 As shown, the user first selects the required data source through a graphical user interface or references an existing model saved previously, which provides a basic data set for subsequent data integration.

[0022] 3-2. Data integration: Figure 3 As shown in the figure, according to the data source or existing model selected in step 3-1, a data node will be generated. Users can perform subsequent operations based on this node, including field management, table association, and table merging, so as to form a comprehensive and consistent data set in series, laying a solid foundation for further analysis and modeling. The "comprehensiveness" of the data set is reflected in the merger of data sets from different sources through field management, table association, and table merging operations to cover all relevant information required for user analysis. For example, table association can connect multiple tables through common fields to ensure the integrity of data content; table merging expands the dimension and depth of the data set through vertical or horizontal merging. The "consistency" of the data set is reflected in the standardization and uniformity of the data. Through field management, the consistency of field names, formats, and units is ensured to avoid the occurrence of duplicate and redundant data, thereby improving data quality and ensuring the accuracy and availability of the data set.

[0023] 3-3. Operation nodes: Figure 4 As shown, using the data set integrated in step 3-2, users can perform specific data processing tasks in the operation node, such as data filtering, sorting, aggregation and other operations, to obtain a customized result table, which will be used for final data analysis and decision support.

[0024] Further, such as Figure 5 As shown, for step 4, the user can export the constructed data processing model into a file, such as JSON or XML format, for subsequent use or sharing.

[0025] In addition, the present invention also provides a visual modeling system based on process mining, including: The collection module is used to collect process data from log files; The preprocessing module is used to clean and preprocess the collected process data; A visual modeling module is used to perform visual modeling using the process data processed by the preprocessing module; the visual modeling provides a graphical interface, and the user can create and modify the data processing model through dragging and editing operations; The export module is used to export the constructed data processing model as a file.

[0026] The following is a specific example scenario. Figure 6 The process shown in the figure details the specific implementation steps of the present invention in practical application: Assume there is an e-commerce dataset that includes an order table and a user table. Use visual modeling to analyze user purchasing behavior.

[0027] Step 1: Select a data source or reference an existing model (corresponding to Figure 6 Data node 1) In this step, the order table and user table in the e-commerce dataset are selected as data sources. The order table contains detailed order information, such as order ID, user ID, order amount, purchased items, etc.; the user table contains basic user information, such as user ID, name, age, gender, region, etc.

[0028] Step 2: Data integration (corresponding to Figure 6 Data node 2) In the data integration step, the order table and the user table are associated through the "user ID" field to generate a new data set. This new data set specifically includes the following actual content: Order details: including the order ID, order amount, purchased product details, order date, etc. of each order.

[0029] User basic information: including the name, age, gender, region, etc. of the user associated with the order.

[0030] Related information: Displays the relationship between each user and his / her orders, such as the types of goods purchased by the user, frequency, and consumption habits.

[0031] This step provides a comprehensive data view that directly links users' personal characteristics to their purchasing behavior, providing a rich information basis for further analysis.

[0032] Step 3: Operation nodes (corresponding to Figure 6 Data node 3) In the operation node step, the generated new data set is further processed. For example, you can add aggregation operations to calculate the total consumption amount, average consumption amount, number of purchases, etc. for each user. In addition, you can group users according to their age and region to analyze the differences in purchasing behavior among users of different age groups and regions.

[0033] Step 4: Set as the result table (corresponding to Figure 6 The final model in In the Set as Result Table step, the merged data set is set as the result table. This result table specifically includes the following actual contents: User purchasing behavior statistics: including each user’s total spending amount, average spending amount, number of purchases, and other statistical data.

[0034] User grouping analysis: Group users by age and region to display the purchasing behavior characteristics of different groups.

[0035] Product sales analysis: Analyze which products are most popular, which products have the highest sales, and the relationship between product sales and user characteristics.

[0036] Trend forecasting: Based on historical purchase data, predict future sales trends and changes in user purchasing behavior.

[0037] This result table provides e-commerce companies with in-depth analysis of user purchasing behavior, helping them optimize marketing strategies and improve sales efficiency. For example, by analyzing user purchasing behavior, companies can find that users in certain regions are more inclined to buy high-value goods, so they can promote high-end products in these regions in a targeted manner.

[0038] Through this process-based processing, the present invention not only improves the efficiency and accuracy of modeling, but also enhances the reusability of the model and the shareability of the results. Each step is a logical continuation of the previous step, ensuring the coherence and systematicness of the entire modeling process.

[0039] The above description is only a detailed description of the preferred embodiments and principles of the present invention. For ordinary technicians in this field, according to the ideas provided by the present invention, there will be changes in the specific implementation methods, and these changes should also be regarded as the protection scope of the present invention.

Claims

1. A visual modeling method based on process mining, characterized in that: The method comprises the following steps: S1, collects process data from log files; S2, cleaning and preprocessing the collected process data; S3, using the process data processed in step S2 to perform visual modeling; the visual modeling provides a graphical interface, and the user can create and modify the data processing model through dragging and editing operations; S4, the user exports the constructed data processing model as a file for subsequent use or sharing; Step S3 includes the following steps: S31, select data source or reference existing model: the user selects the required data source or references the previously saved existing model through the graphical interface to provide the basic data set; S32, data integration: according to the data source or existing model selected in step S31, a data node is generated, and the user performs subsequent operations based on the data node to form a data set in series; the subsequent operations include field management, table association and table merging; S33, operation node: using the data set integrated in step S32, the user performs specific data processing tasks in the operation node step to obtain a customized result table, which will be used for final data analysis and decision support; the data processing tasks include data filtering, sorting and aggregation.

2. The visual modeling method based on process mining according to claim 1 is characterized in that: In step S2, the collected process data is cleaned including missing value processing, abnormal value processing, duplicate value processing and erroneous data processing; The missing value processing refers to deleting, filling or predicting missing values; The outlier processing refers to identifying and processing outliers; the processing methods include deletion, modification and marking; The duplicate value processing refers to deleting duplicate records; The error data processing refers to correcting errors in the data set.

3. The visual modeling method based on process mining according to claim 2 is characterized in that: In step S2, the collected process data is preprocessed including data integration, data conversion, data normalization, data dimension reduction and data enhancement; The data integration refers to merging multiple data sources; The data conversion refers to performing logarithmic transformation, normalization and discretization operations; The data standardization refers to the unification of data formats and units; The data dimensionality reduction refers to reducing the data dimension by aggregation, reduction and compression methods; The data enhancement refers to adding additional information to enrich the dataset.

4. The visual modeling method based on process mining according to claim 1 is characterized in that: In step S32, the field management refers to the user managing the fields in the data node, including adding, deleting and renaming operations.

5. The visual modeling method based on process mining according to claim 1 is characterized in that: In step S32, the table association refers to associating multiple data nodes to form a new data set, which specifically includes merging two tables through a common field.

6. The visual modeling method based on process mining according to claim 1 is characterized in that: In step S32, the table merging refers to merging multiple data nodes into a single data set; the merging method includes vertical merging or horizontal merging.

7. The visual modeling method based on process mining according to claim 1 is characterized in that: In step S32, the data set has the following characteristics: Comprehensiveness: Specifically, comprehensiveness refers to merging data sets from different sources through field management, table association, and table merging operations to cover all relevant information required for user analysis; Consistency, specifically refers to ensuring the uniformity of field names, formats, and units through field management.

8. A visual modeling system based on process mining, used to implement the visual modeling method based on process mining according to any one of claims 1 to 7, characterized in that: The process mining-based visual modeling system includes: The collection module is used to collect process data from log files; The preprocessing module is used to clean and preprocess the collected process data; A visual modeling module is used to perform visual modeling using the process data processed by the preprocessing module; the visual modeling provides a graphical interface, and the user can create and modify the data processing model through dragging and editing operations; The export module is used to export the constructed data processing model as a file.

Images