Collaboration method and system based on data driving

Abstract

The invention relates to the technical field of artificial intelligence data processing, in particular to a collaboration method and system based on data driving. Defining a data storage mode and an affinity label according to an activity participation model, using a large model to decompose data, constructing a dynamic label according to data content, performing quantification processing on the affinity of activity participation in combination with a general label and the dynamic label, and performing quantification calculation on the affinity of activity participants and activity items; according to the method, dynamic management of activities and participating instances and functions and users is realized, association of association positioning and business collaboration of the activities and the participating instances and the functions and the users is realized, operation of triggering association functions with proximity during function operation of the users is supported, and linkage between business systems, data connection and business starting of a data driving mode are realized. The working mode that a user starts the service through menu navigation is improved, repeated input of repeated data in the system is reduced, the workload is reduced, the efficiency is improved, and the data consistency is improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence data processing technology, and in particular to a data-driven collaborative method and system. Background Art

[0002] Traditional collaborative office systems are typically planned based on the client's business needs, managed according to business processes through business analysis and functional design, and presented and manipulated in a form when completing the relevant functional interface. For storage, relational databases are used to define and operate modular functions. Ultimately, content is primarily reflected as result-based data at multiple levels, with less storage for processes, version classes, and additional data. The data volume is small, the management granularity is coarse, and the relationships between data are relatively simple. This model leads to the emergence of "data silos" between subsystems and functions, causing users to repeat operations and input between different functional systems, increasing the complexity of user experience and increasing office burdens and costs.

[0003] With the development of artificial intelligence and big model technologies, semantic recognition, data extraction, and change detection can be achieved with relatively simple methods. Big models can help accurately extract and classify information from data, understand data relationships, identify data changes, and support the construction of data versioning content. In addition, through secondary training and prompt word engineering, and through the definition and coordinated use of data standards, data standardization and business linkage can be achieved based on user characteristics.

[0004] Traditional systems primarily focus on single-business data reporting and business flow, with little consideration for horizontal cross-functional connections and data consistency. Implicitly linked data between businesses is not stored or used, making data value mining and reuse difficult. The introduction and promotion of new technologies is costly and challenging. Digital transformation presents multiple challenges in terms of data collection and governance, as well as the development of data-driven collaborative models. Summary of the Invention

[0005] To build collaborative relationships within the system's internal data, promote customer awareness of the data-driven value of intelligent models, and accelerate the upgrade and transformation of legacy systems, this solution, based on accumulated data for customers, implements intelligent transformation and application promotion of collaborative office systems. This solution, based on comprehensive data collection, conducts data governance and analysis through big data and models. Based on data collection, governance, and modeling, it constructs a label affinity system and a hierarchical ranking method to establish linkage relationships between users and functional points, achieving business linkage while also building horizontal data relationships.

[0006] This implementation plan defines the data storage method and affinity tags according to the activity participation model. After using the large model to decompose the data, dynamic tags are constructed according to the data content. The affinity of activity participation is quantitatively processed by combining general tags and dynamic tags. The affinity of activity participants and activity matters is quantitatively calculated respectively. The affinity is calculated in a hierarchical manner = person affinity * (activity classification affinity * synthesis ratio + activity dynamic tag * impact factor). This enables dynamic management of activities and their participating instances and functions with users, and realizes the association positioning and business collaboration between the two. It supports users to trigger the operation of related functions with affinity when operating functions, realizes the linkage between business systems and the connection of data, and the start of business in a data-driven mode. It improves the working method of users starting business through menu navigation, reduces the repeated entry of duplicate data in the system, reduces workload, improves efficiency, and improves data consistency.

[0007] In a first aspect, embodiments of the present application provide a data-driven collaborative method, which primarily includes the following: The first step is to obtain request and response data in a bypass manner, collect user requests and extended data, use the participation model to organize data with form fields as the basic unit, convert operation records into user participation content, annotate data versions and store them in a big data format, realize the collection of data throughout the process, and organize them according to activity participation methods and data versions.

[0008] Furthermore, when user data changes, data is saved by inserting and marking data versions without updating them. Data versions are saved simultaneously according to time series operations, and relationships are established through activity participation classification and business. In the BS system, full-path URL marking is combined with the URL path and related parameters to identify the data field names of page forms. The large model is used to compare and extract data content, and the changed data is recorded according to the independent field participation model. This achieves overall data version management and supports the second step of building a complete data model.

[0009] The second step is to build and refine the activity participation model. The data collected in the first step is organized according to activity categories, participation content, and system functions. Extended information such as activity participation, participation time, participation method, and other versions and timeliness are incorporated into the storage model to support operational analysis around the activity theme. Then, using the large model's prompt word project and activity template data elements, related content is identified and extracted. Relationships between activities are established based on identification tags within a specific time and scope, building connections between business activities and system functions, forming elements for horizontal data integration.

[0010] The third step is to use the data connection relationship of the activity participation model to generate user dynamic operation commands based on the system function points, collect and incorporate them into closed-loop data when users perform quick operations, and after data governance, dynamically match the activity participation type labels in the participation content based on the participation content and version, and mark them with weights to form basic proximity data to support matching with subsequent user function operations.

[0011] In the fourth step, when the user enters the function interface for subsequent operations, the system uses the basic information and extended information filled in by the user as input parameters, collects and executes data according to the dynamic commands predetermined in the third step, integrates the adapted data into output content, and generates guidance auxiliary information based on weights, providing direct operation and secondary operation capabilities, supporting hierarchical decision-making methods for priority calculation, and making decisions based on this.

[0012] Furthermore, when the user performs a secondary query and operation, the original system calls the interface, takes the detailed content as a parameter with the version data, revises the proximity data, and then processes it according to the secondary revised data interface, presenting possible basis and key content synchronously to support quick operations for users.

[0013] On the second aspect, the embodiment of the present application provides a data-driven collaborative system, whose modules mainly include the following five unit modules: the first unit data acquisition module transfers functional forms, extended data and process data in an integrated manner; the second unit data governance module performs data governance according to the metadata definition and model; the third unit data analysis module uses the activity model to incorporate the governed data, constructs the activity participation proximity in a labeling manner, and realizes decision-making through sorting; the fourth unit data drive module uses the proximity index to drive the operation of the decision support mode in a threshold control manner; the fifth unit business collaboration module uses the unified activity label to support business collaboration according to activity classification and key content indicators.

[0014] 1. Data Collection Module: This module uses bypass monitoring technology to parse upstream and downstream data flows through the HTTP protocol, analyzing business system requests and form submission data. It also uses an activity participation model to store raw data, using form fields as the basic unit to collect data in a standardized manner based on participants. It integrates data through form adaptation, achieving the fusion of process data and result data to ensure data consistency and integrity. It uses a data version control mechanism to store and manage data, and leverages big data technology for efficient data extraction and verification to capture complete data.

[0015] 2. Data Governance Module: This module uses metadata to define forms and data fields and data versions through the request URL. The request URL and related parameter flags establish a correspondence between activity participation models, business functions, and forms, enabling data-to-business matching. User operation labeling improves the granularity and accuracy of data relationships. Collected data is cleansed, integrated, and standardized based on defined data standards and rules. By matching activity participation with functional modules, business operations and data collection are matched, enabling data governance. Furthermore, data operation content is labeled based on the comparison of data versions and final results, supporting process management of data changes.

[0016] 3. Data Analysis Module: Relying on the activity participation model and process data, the activity participation relationship is refined into the operational relationship between and within activities. The relationship between users associated with the activity and the corresponding relationship between labels such as user positions are used to build affinity indicators. Statistics, machine learning and data mining techniques are used to analyze and mine process data and result data. According to the affinity value, affinity vectors of function points, user interfaces, activity types and related labels are established to support users in making auxiliary decisions and providing guidance before and after operations.

[0017] 4. Data-driven module: According to the proximity index, the activities of related users are secondary identified before and after user operations and before and after the occurrence of related businesses. The proximity index is identified according to the hierarchical analysis method. When the user's corresponding threshold is reached, an operation event is created for the user to realize the business processing model of finding the person to do the business and realize the data-driven model.

[0018] 5. Business collaboration module: Build a unified workbench based on the activity participation model and activity event-driven approach, define various types of subject domains and operation domains through activity classification and subject tags, and perform business definitions for operation domains. Based on authorized control, achieve interface differentiation and collaboration of business operations through business data content and subject tags.

[0019] This implementation plan defines data storage methods and affinity tags based on the activity participation model. After using a large model to decompose the data, dynamic tags are constructed based on the data content. The affinity of activity participation is quantitatively processed by combining general tags and dynamic tags. The affinity of activity participants and activity items is quantitatively calculated separately. Affinity is calculated in a hierarchical manner: person affinity * (activity category affinity * composite ratio + activity dynamic tag * impact factor). This enables dynamic management of activities, their participating instances, and users, and achieves correlation and business collaboration between the two. Furthermore, based on the affinity between activities and people, triggered or automated processing is implemented through intelligent agents in a "find the person for the task" model.

[0020] The above description is an overview of the present invention. In order to more clearly understand the technical means of the present invention, you can refer to the contents of the specification and implement it. The relevant embodiments of the present invention are described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present invention or the existing technical solutions, the following introduces the drawings required for use in the embodiments or the description of the existing technology. The following drawings are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the drawings.

[0022] Figure 1 This is a data flow diagram of activity participation in an embodiment of the present invention; Figure 2 A schematic diagram of the composition of an activity participation model according to an embodiment of the present invention; Figure 3 Activity analysis flow chart of an embodiment of the present invention Figure 4 This is a diagram of the module composition of the device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0023] first Figure 1 This is a data flow diagram of activity participation in an embodiment of the present invention, the main contents of which are as follows: S110 Activity construction and initialization: Using big data storage, the client's request is used to create activity participation content according to the URL and form key information, and the activity form fields are versioned and stored according to the participation content to achieve the creation and storage of activity participation content. In addition, metadata is used to configure and manage activity types, build a mapping relationship between function points and activities, establish an association relationship between activity creators and activity participants, and create and digitize content in all aspects according to the activity theme, activity domain and content.

[0024] S120 Creation of associated activities: The operations of creating associated activities are basically the same as those of initializing activities. The difference is that associated activities are created based on the creation of predecessor activities. The predecessor activities are specified in the successor activities. If the successor activities have multiple predecessor activities, they are specified in a one-to-many manner.

[0025] S130 Versioning confirmation of activity participation content: A separate copy of data is saved for each activity participation behavior, but the specific content is saved and compared with the previous activities of the activity participant. If the data changes, the content is created; if the data does not change, the domain data is not created and saved; for data in the same activity domain, the domain name change data is saved, and the content change version is marked in the new activity participation.

[0026] S140 Activity Participation Content Data Governance: In accordance with the big data governance method, newly added data is stored and managed according to the activity participation model. For historical data, data versions are constructed through log comparison to implement version management and data standardization operations. In addition, according to the latest data format definition, relevant data is formatted and data governance work is defined to achieve data content governance.

[0027] S150 Activity Participation Proximity Analysis: Defines the relationship between activity creation and activity participants, constructs proximities based on parameters such as position, age, company length of service, number of joint activities, activity participation, and similarity of activity content, and uses a hierarchical diagnostic approach to clarify and label relevant content, converting it into activity participation proximities.

[0028] S160 proximity content construction and function triggering: The segmented indicators of the proximity of related activity participation are converted uniformly according to the proximity adjustment mode defined by the current user. When the activity content and participation status change, the proximity of each user participating in the activity is converted to achieve the proximity relationship between each user and the activity. The driving relationship between the activity and the person is reconstructed according to the latest proximity, forming the starting point for data-driven activities, and supporting secondary processing based on intelligent agents and human participation.

[0029] Secondly Figure 2 This is a schematic diagram of the activity participation model composition of this embodiment, which is constructed and triggered from four levels, and the storage and analysis model content is formed in a proximity manner.

[0030] 1. Activity Classification: Different from traditional system function definitions, this solution uses an activity participation model to build a unified data model for data version and content management. Activity classification establishes a correspondence between classification and business functions through data standard specifications. It supports design based on data model definitions, uses metadata to associate function definitions, and determines the relevance of new and old systems based on activity classification and function through activity scenarios. It digitally defines activity operation classifications, operator roles, function associations, front-end and back-end functional interfaces, and system functional interfaces according to the digitalization of scenarios, and dynamically adjusts them through tags.

[0031] 2. Proximity index: The proximity index is designed in a layered manner and superimposed layer by layer. It is defined around users and activities. Users are defined by labels such as age and responsibilities. The participation of users and activities is defined according to the matching of content tags, summaries and content. The layers are superimposed to finally generate the proximity index of users and activities around user relationships. The indicators are then sorted by proximity and matters are processed based on detailed indicators to realize the "find the person for the matter" driving model.

[0032] 3. Activity Participation: Activity participation is defined around the relationships between participants and their participation in activities. Personnel relationships are defined based on similar habits, shared participation, and similar responsibilities. Content analysis ranks these similarities, prioritizing stakeholders involved in the same activity. This supports the acquisition of functional data related to the participants and enables digital processing of related activities. Personnel-activity participation relationships are linked based on business data content, defined digitally using similar relationships and related data. After the activity is processed, a final label is generated based on the results to confirm the primary ranking indicator.

[0033] 4. Activity Content: Based on the associated functional system, activity content analyzes user participation habits through activity operation frequency, organizes activity operation trends according to the interval classification, and identifies user characteristics through the association between functions and activity categories and the horizontal and vertical association relationships. Activity associations are further identified based on the distinction between the response relationships of upper and lower functions, and the context of different categorized activities is established. Data versioning subdivides data granularity from form to form field control, implementing version management based on data changes.

[0034] Secondly Figure 3 This is a flow chart of activity analysis according to an embodiment of the present invention, which implements the integration of segmentation indicators and event triggering through model definition based on proximity.

[0035] S310 The previous user performs business function operations: After the previous user performs the operation, the system creates an instance according to the activity participation model, establishes a matching relationship between the function node and the activity type, performs data storage and initial version management, and builds the initial participant proximity relationship according to the activity participants, forming the label relationship and weight of the initial activity and the creator and participant.

[0036] S320 Subsequent user participation business processing: According to the functional flow, after the subsequent user operates, on the basis of building a participation model, AI technology is used to compare the content of the previous activities to form a data version, and unchanged data is excluded. At the same time, on the basis of establishing the relationship between the participation of the previous and subsequent activities, data governance is carried out according to the activity participation model to achieve standardization and unification of data formats and methods.

[0037] S330 decomposes activity participation content based on the model: it compares the activity content, extracts labels from the activity content based on parameters such as user operations, user habits, and user proximity, and assigns values ​​to relevant indicators based on user relationships and participation weights to form detailed indicator content. The final results of the activity are processed in a dynamic labeling manner to form indicators with higher weights.

[0038] S340 determines the proximity of segmented indicators in a hierarchical manner: the main labels formed in each activity are dynamically matched using a quantitative method, and the adaptation of the label quantitative data is matched to the personnel relationships at different levels, thereby forming the proximity of the label indicator. Then, the superposition calculation is performed based on the proximity ranking to form the final proximity. The algorithm converts according to [hierarchical calculation of proximity = personnel proximity * (activity classification proximity * synthesis ratio + activity dynamic label * impact factor)] and converts in a weighted superposition manner.

[0039] S350 generates the proximity of activity stakeholders: After calculating the proximity between activity participants and the proximity between activity participants, the proximity between the final activity and the participants is calculated through superposition, supporting business flow operations based on proximity.

[0040] S360 triggers secondary participation events to stakeholders: After each user completes an operation, it first generates the final version of the stakeholder proximity, then sorts the proximity data superimposed with the final proximity, and then selects and sorts the proximity of relevant personnel. For any changed proximity, it re-sorts the activities or sends subscription notifications according to the business definition or person-defined threshold of the participants to inform the stakeholders to handle the matters.

[0041] Secondly Figure 4 This is a module composition diagram of the system according to an embodiment of the present invention. The diagram describes the main modules of the system. The modules have been explained in the text and will not be described here: A data-driven collaborative method and system builds a data storage model for activity participation, defines proximity indicators, and forms the proximity between people and activity content by merging sub-indicators. The final activity proximity is converted through the definition of [Proximity = Person Proximity * (Activity Category Proximity * Composite Ratio + Activity Dynamic Label * Influence Factor)], realizing the technical elements of data-driven activity linkage, realizing business linkage in the "find person for something" model, simplifying user operations, and effectively improving customers' office efficiency.

[0042] Through the above description of the embodiments, those skilled in the art will be able to understand how the described example embodiments can be implemented, and that they can also be implemented through software combined with necessary hardware. The technical solutions of the disclosed embodiments are embodied in software, whose primary storage medium relies on a cloud-native network. This software product includes methods for enabling computing devices to function properly, employing large models and artificial intelligence technologies, and utilizing agent technology to simplify business operations.

[0043] A person skilled in the art will easily think of other implementation plans after referring to the invention disclosed in the specification and related practices. The present application is intended to cover relevant modifications, uses or adaptive changes of the present disclosure. The relevant adaptive changes follow the general principles of the present disclosure and include known attempts or customary technical means in the technical field that are not disclosed in the present disclosure. The description and examples are to be regarded as exemplary only, and the present disclosure indicates relevant rights in accordance with the claims.

[0044] Main technical points: This solution uses an activity participation model to integrate extended data associated with business systems. Through data tags and association affinities, it identifies cross-system connections between activity participation and business functions, enabling business linkages based on changes in data elements and supporting a data-driven model. During data processing, it enables business operation forecasting and supports reporting and decision-making with business data, effectively improving the efficiency and quality of business operations while enhancing the consistency and transparency of system processing.

[0045] 1. Based on the expansion of collected data, a unified activity participation model is adopted for data storage and management. In combination with activity classification and custom tags, artificial intelligence technology is used to identify and process the correlation between activity content and its version data, forming a close relationship between data content, activities and users, and relying on closeness to achieve synergy between activity content and business functions.

[0046] 2. Relying on the activity participation model and data association relationship, on the basis of collecting and extracting activity content, the activity classification and specific operations of activity participation are calculated using a hierarchical model through the closeness between users and between activities and users, and auxiliary tools are used to support the closed-loop reminders and operations of the user system.

[0047] Glossary Activity Participation Model: This model defines a unified data storage model, defining activity storage methods based on four levels: activity content, activity participation, and the relationships between participants. Technically, it utilizes a big data storage approach. Unlike typical functional designs, activity participation uniformly defines business operations, facilitating the establishment of horizontal relationships between business functions and content. By establishing unified activity relationships, it achieves the integration of business, functions, and data, supporting the linkage of business functions based on activity associations and enabling business collaboration.

Claims

1. A data-driven collaborative approach characterized by The steps include: S110 activity construction and initialization: Using big data storage, we create activity participation content based on the client's request URL and form key information, and version the activity form fields according to the participation content, thus realizing the creation and storage of activity participation content. In addition, we use metadata to configure and manage activity types, build a mapping relationship between function points and activities, and establish a relationship between activity creators and activity participants. We also create and digitize content in all aspects according to the activity theme, activity field, and content. Creation of S120 associated activities The creation of an associated activity is basically the same as the initialization of an activity. The difference is that the associated activity is created based on the creation of the predecessor activity. The predecessor activity is specified in the subsequent activity. If the subsequent activity has multiple predecessor activities, they are specified in a one-to-many manner. S130 Version confirmation of activity participation content Each activity participation behavior saves a separate copy of the data. However, the specific content is compared with the previous activities of the activity participant. If the data changes, the content is created; if the data does not change, the domain data is not created and saved. For data in the same activity domain, the domain name change data is saved, and the content change version is marked in the new activity participation. S140 Activity Participation Content Data Governance In accordance with big data governance methods, newly added data is stored and managed according to the activity participation model. For historical data, data versions are constructed through log comparison to implement version management and data standardization operations. In addition, relevant data is formatted and data governance work is defined according to the latest data format definition to achieve data content governance; S150 Activity Participation Proximity Analysis Define the relationship between event creators and event participants, and build proximity based on parameters such as position, age, company length of service, number of joint events, amount of event participation, and similarity of event content. A hierarchical diagnostic approach is used to identify and label relevant content, and convert it into activity participation proximity. S160 Proximity content construction and function triggering The detailed indicators of the proximity of relevant activity participation are converted uniformly according to the proximity adjustment mode defined by the current user. When the activity content and participation status change, the proximity of each user participating in the activity is converted to realize the proximity relationship between each user and the activity. The driving relationship between the activity and the person is reconstructed according to the latest proximity, forming the starting point for data-driven activities, and supporting secondary processing based on intelligent agents and human participation.

2. The data-driven collaborative method according to claim 1, characterized in that The activities in S110 are divided into activity classification, proximity index, activity participation and activity content; Activity classification: Use the activity participation model to build a unified data model for data version and content management. Activity classification establishes a correspondence between classification and business functions through the specification of data standards. It supports design based on data model definition and the use of metadata to associate function definitions. Activity classification and function are determined by activity scenarios to determine the relevance of new and old systems. Activity operation classification, operator role, function association, front-end and back-end function interfaces, and system function interfaces are digitally defined according to the digitalization of the scenario, and dynamically adjusted through tags. Proximity Indicators: Affinity indicators are designed in layers and stacked layer by layer. They are defined around users and activities. Users are defined by age, role, and other labels. User and activity participation is defined by the matching of content labels, summaries, and content. These layers are stacked to ultimately generate a proximity indicator around user relationships. These indicators are then sorted by proximity, and detailed indicators are used to trigger event processing. Activity Participation: Activity participation is defined around the relationships between participants and the relationships between participants in the activity. Personnel relationships are defined based on similar habits, degree of joint participation, and similarity of responsibilities. Content analysis is used to sort the similarities and prioritize stakeholders in the same activity. This supports the acquisition of functional data related to the sorted personnel, enabling digital processing of related activities. The relationship between personnel activities is associated with the data content of the business and is defined in a digital way using similar relationships and related data. After the activities are processed, the final label is formed based on the results to confirm the main ranking indicators. Activity content: Based on the associated function system, activity content analyzes user participation habits through activity operation frequency, sorts out activity operation trends according to activity operation interval classification, and identifies user characteristics through the association between functions and activity classifications and according to horizontal and vertical association relationships; On the basis of distinguishing the response relationship between superior and subordinate functions, the activity association is re-identified through content to construct the contextual relationship between different classification activities; Data versioning breaks down data granularity from forms to form fields, and implements version management based on data changes.

3. The data-driven collaborative method according to claim 1, characterized in that The S160 proximity content construction and function triggering are divided into the following steps: S310 The preceding user performs business function operations After the previous user operation, the system creates an instance according to the activity participation model, establishes a matching relationship between functional nodes and activity types, performs data storage and initial version management, and builds the initial participant affinity relationship according to the activity participants, forming the label relationship and weight of the initial activity and the creator and participant; S320 Subsequent user participation in business processing Following the functional flow, after subsequent users operate, AI technology is used to compare the content of previous activities based on the participation model to form a data version, excluding unchanged data. At the same time, based on the establishment of the participation relationship between previous and subsequent activities, data governance is carried out according to the activity participation model to achieve standardization and unification of data formats and methods; S330 Model-based decomposition activity participation content: Compare activity content, extract tags from activity content based on user actions, user habits, user affinity, and other parameters, and assign values ​​to relevant indicators based on user relationships and participation weights to form detailed indicator content. The final results of the activity are processed using dynamic tags to form indicators with higher weights. S340 determines the proximity of segment indicators in a hierarchical manner: The main tags formed in each activity are dynamically matched using a quantitative method. According to the adaptation of the tag quantitative data, they are mapped to the personnel relationships at different levels, thus forming the proximity of the tag indicators. Then, based on the proximity ranking, the superposition calculation is performed to form the final proximity. The algorithm calculates proximity in a hierarchical manner = personnel proximity * (activity classification proximity * synthesis ratio + activity dynamic tag * impact factor) and converts it in a weighted superposition manner. S350 generates activity stakeholder proximity: After calculating the closeness between activity participants and the closeness between activity participants, the closeness between the final activity and participants is calculated through superposition, supporting business flow operations based on closeness; S360 triggers a secondary engagement event to stakeholders: After each user completes an operation, the final version of the stakeholder proximity is generated first, and then the proximity data superimposed with the final proximity is sorted. Then, the proximity of the relevant personnel is selected and sorted. For any changed proximity, the activities are re-sorted or subscription notifications are sent according to the business definition or human definition threshold of the participants to inform the stakeholders to handle the matters.

4. A system based on a data-driven collaborative method according to any one of claims 1 to 3, characterized in that include: Data collection module, data governance module, data analysis module, data-driven module and business collaboration module; Data collection module: This module uses bypass monitoring technology to analyze upstream and downstream data flows through the HTTP protocol, parsing business system requests and form submission data. It also uses an activity participation model to store raw data, using form fields as the basic unit and performing standardized data collection based on participants. Integrate data by adapting forms to achieve the fusion of process data and result data; use data version control mechanism to store and manage data, and use big data technology to extract and verify data; Data Governance Module: This module uses metadata to define forms and data fields and data versions through the request URL. The request URL and related parameter flags establish a correspondence between activity participation models, business functions, and forms, enabling data and business matching. Furthermore, through user operation labeling, the module cleanses, integrates, and normalizes collected data according to defined data standards and rules. Furthermore, by matching activity participation with functional modules, the module matches business operations with data collection, achieving data governance. Furthermore, by comparing data versions with final results, the module adds labels to data operation content, supporting process management of data changes. Data Analysis Module: Relying on activity participation models and process data, this module breaks down activity participation relationships into operational relationships between and within activities. It also constructs affinity indicators based on the relationships between users associated with activities and the corresponding relationships between labels such as user positions. It uses statistics, machine learning, and data mining techniques to analyze and mine process and result data. Based on affinity values, it establishes affinity vectors for function points, user interfaces, activity types, and related labels, supporting users in making auxiliary decisions and providing guidance before and after operations. Data-driven module: Based on the proximity index, secondary identification is performed on the activities of related users before and after user operations and related business operations. The proximity index is identified using a hierarchical analysis method. When the user reaches the corresponding threshold, an operation event is created for the user, realizing a data-driven model. Business collaboration module: Build a unified workbench based on the activity participation model and activity event-driven approach, define various types of subject domains and operation domains through activity classification and subject tags, and perform business definitions on the operation domains. Based on authorized control, achieve interface differentiation and collaboration of business operations through business data content and subject tags.

Images