A generation method based on domain knowledge model modeling and dynamic optimization configuration

Abstract

A generation method based on domain knowledge model modeling and dynamic optimization configuration, which relates to the field of dynamic modeling analysis, solves the problems that existing systems cannot flexibly adapt to applications in specific scenarios, have poor response capabilities and low application efficiency for different business models and model dynamic optimization work. The present invention combines model modeling with a model optimization designer, utilizes the cohesion and stability of domain knowledge, mines implicit domain knowledge, hierarchically organizes domain knowledge, establishes an attribute decomposition, quantification mechanism and transitive dependency relationship, and establishes a formal definition and inference rules for the dependency relationship between attributes and requirements. The present invention configures model forms, rules, and relationships at the management end, defines the attribute configuration of the model and the page scenario by ticking and dragging methods, and the front end defines and optimizes according to various models and application scenarios, and dynamically generates, loads and displays the page to meet different needs of users.

Description

Technical Field

[0001] The present invention relates to the technical field of dynamic modeling analysis, and particularly relates to a generation method based on domain knowledge model modeling and dynamic optimization configuration. Background Art

[0002] At present, in the process of software development, software developers usually hope that the business object model can adapt to multiple scenarios within a certain business scope, enabling the business system to have the ability to expand or adjust, adapt to different application scenarios, and thus adapt to different business industries. However, the flexibility of the existing system for establishing and optimizing models is limited. Generally speaking, the model describes a static data set, and it is required to establish a complete model before starting the analysis. Once the model is established, the cost of change will be very high. When users face data outside the predefined dimensions, data analysis cannot be effectively carried out, and multiple repetitions are required. The ability of the business system to dynamically adjust and quickly respond to changes in requirements is not strong. Through the method of the present invention, a language or tool capable of describing different data models is established. This method enables the data analysis system to be applicable to different business scenarios and has great advantages. Summary of the Invention

[0003] In order to solve the problems that the existing business system cannot flexibly adapt to the application in a specific scenario, has poor response ability and low application efficiency for different business models and model dynamic optimization, the present invention provides a generation method based on domain knowledge model modeling and dynamic optimization configuration.

[0004] A generation method based on domain knowledge model modeling and dynamic optimization configuration, which is implemented by the following steps:

[0005] Step 1: Define and manage model form information, and import and verify the data set;

[0006] Step 2: Quantitatively define the model form described in Step 1, aggregate domain knowledge, and calculate transfer dependencies to obtain a dimension model after dynamic expansion;

[0007] Step 3: Define the original model, establish the mapping from the original model requirements to entity objects and index attributes, assist in constructing the association relationship between requirements, and generate an extended dependency relationship model;

[0008] Step 4: Establish association rules and constraint relationships according to the extended dependency relationship model generated in Step 3;

[0009] Step 5: Execute the optimization program to improve the model requirement dependency network.

[0010] Further, in Step 1, define the content and classification of the model form, and map the attribute information of the model form to the form attributes;

[0011] The user manages the model form information on the management - end page. By configuring the relationship between the form stored in the model database, the object JAVA class, the form attributes, the object JAVA class, and the database - stored form are associated;

[0012] Meanwhile, configure the parameters related to the model form and the classification parameters, and set the uniqueness verification attribute set of the model form.

[0013] Furthermore, when the user maintains the model form attribute information on the management - end page, the uniqueness verification attribute set of the model form includes the ID, attribute name, type, length, data dictionary, and dimension of the model form attribute; meanwhile, configure the openness, permission scope, participation algorithm of the relevant model form, as well as the attribute information of the application scenario, query configuration, export, and loading.

[0014] Furthermore, in step two, the quantitative definition of the model form specifically refers to the definition of model indicators, which is used to complete the conversion of domain knowledge from qualitative description to quantitative description, realize the aggregation of domain knowledge and the calculation of transfer dependencies among the relationships of model data attributes, obtain the extended domain knowledge required for auxiliary model requirement modeling, and establish the dimension information of single - factor and coupled - factor influences;

[0015] Map the attribute fields in the data set to a pre - set dimension model or create a new index dimension according to the verification information in step one, forming a dynamically extended dimension model.

[0016] Furthermore, in step three, the index attributes are used for the user to configure the attribute information participating in the model definition on the management - end page; it includes model attributes, display names, component types displayed in the model form, whether it is required, static default values, dynamic default values, data dictionaries, attribute verification rules, and relevant prompt information, etc., and provides a reset function for model attributes.

[0017] Furthermore, in step three, the user defines and sets the attributes of the original model on the management - end page; when creating the original model definition, the function of the model designer will be provided. On the original model definition page, drag in the relevant attribute list corresponding to the model, drag it to the model page designer area, and support the configuration of the attribute controls, model indicators, and personalized information of the attributes in the design page; and support loading external model information;

[0018] Create a form based on the dynamically extended dimensional model obtained in Step 2, configure the relationship between the attribute fields and the data elements corresponding to the attribute fields, and the relationship table between the form attribute fields and the hierarchy within the dimensional model; establish an influence factor model, collect the underlying nodes or basic metrics, perform coupling calculations, define the matching fitness of relevant influence factors and the rules for relevant path branches, and form a dynamic modeling framework; at the same time, provide the functions of resetting and copying and creating new models defined.

[0019] Further, in Step 4, based on the generated extended dependency relationship model, establish association rules and constraint relationships, specifically:

[0020] Establish an initial requirement model according to the requirements, generate an extended requirement model in combination with domain knowledge, and summarize or discover new requirements based on the corresponding association rules on the basis of the original requirements, including the mapping relationship between the original model and the application scenario, the comparison relationship between the original models, the decomposition between the application scenarios, and the dependency association relationship.

[0021] Further, in Step 5, execute an optimization program to improve the model requirement dependency relationship network, specifically:

[0022] Execute an auxiliary optimization program, configure the model training parameters and rules, automatically load the data and dynamically execute the optimization model to determine which models or rules inherit the dependency relationship of the parent node; through continuous dynamic adjustment and model optimization, obtain a fine extended dependency relationship to meet the requirements of knowledge conversion, relationship reasoning, and obtaining ideal conclusions.

[0023] Advantages of the present invention:

[0024] The generation method described in the present invention combines model modeling with a model optimization designer, utilizes the cohesion and stability of domain knowledge, and proposes a business requirement modeling and dynamic optimization configuration assistance method based on a domain knowledge model, mines implicit domain knowledge, hierarchically organizes domain knowledge, establishes an attribute decomposition, quantification mechanism, and transfer dependency relationship calculation method, and establishes a formal definition and inference rule for the dependency relationship between attributes and requirements. By configuring the model form, rules, and relationships at the management end, defining the attribute configuration of the model and the page scenario in a ticking and dragging manner, and the front end generates and loads the display dynamically according to various model and application scenario definitions and optimizations, which is more flexible, so as to meet different user requirements without the intervention of developers and without republishing the program package. Description of the Drawings

[0025] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0026] Figure 1 It is a flowchart of a generation method based on domain knowledge model modeling and dynamic optimization configuration according to the present invention. Specific implementation manners

[0027] To make the objectives, technical solutions and advantages of the present application clearer, the following will describe the technical solutions of the present invention in detail with reference to the drawings and embodiments. Obviously, the described embodiments are only some embodiments of the present application, rather than all embodiments. Based on the embodiments in the present application, all other implementation manners obtained by those of ordinary skill in the art without creative efforts belong to the scope protected by the present application.

[0028] A generation method based on domain knowledge model modeling and dynamic optimization configuration described in this implementation manner is realized by the following steps:

[0029] S1: Define and manage model form information, import and verify the data set; that is, import the data set information that conforms to the model form definition, and perform inspections in terms of data type, content, range, etc., including: inspecting the data attribute fields representation, data element representation, and related consistency degree of the data set;

[0030] In the preparation stage, first determine the content and classification of the model form. The model form is used to accurately map information requirements to form attributes, that is: based on the sorting and design of the relevant attributes, definitions, and relationships of the model form, configure and map them to the database form and the associated relationship. Provide a basis for realizing the mapping and organization of the original information requirements; the user manages the model form information on the management end page. By configuring the relationship between the model database storage form and the object JAVA class, establish an association between the form attributes, the object JAVA processing class, and the database storage form, providing a basis for the implementation of the model application program. By configuring the parameters related to the model form and the classification parameters, prepare for the subsequent definition and maintenance of the model and business scenarios. In order to achieve the accuracy of model form attribute management, a series of configurations at the model form level such as setting a relevant set of uniqueness verification attribute IDs and anti-tampering are also required.

[0031] When users maintain model form attributes on the management page, they need to set the model form's attribute information, including its ID, attribute name, type, length, data dictionary, and dimensions. They can also configure the model form's openness, permission scope, participating algorithms, and other attributes, including application scenarios, query configuration, export, and load.

[0032] S2: Quantitative definition of model forms, aggregation of domain knowledge, and calculation of transitive dependencies. The definition of model indicators is a quantification technology for domain knowledge, completing the transformation of domain knowledge from qualitative description to quantitative description. According to the specific business model requirements, multi-dimensional model indicators such as trend, structure, comparison, and link are constructed to support the aggregation of domain knowledge and the calculation of transitive dependencies between model data attribute relationships, and obtain the extended domain knowledge required for auxiliary model requirement modeling. Dimensional information of single influencing factors and coupled influencing factors is established, such as analysis dimension factors such as time, space, and organization. In addition, according to the requirements of different business models, multi-dimensional influencing factor information such as production, sales, consumption, weather, planting yield, age, behavior results, etc. is constructed to better support the construction of quantitative analysis models of multi-factor coupling or mutual influence.

[0033] Based on relevant inspection information, the attribute fields in the data set are mapped to a pre-set dimensional model or a new indicator dimension is established to form a dynamically expanded dimensional model.

[0034] S3: Define the original model, establish a mapping from the original model requirements to entity objects and indicator attributes, and assist in building the relationship between requirements; ultimately generate an extended dependency model.

[0035] On the management page, users define the original model of domain knowledge and set its properties. When creating the original model definition, a model designer is provided. Within the model definition page, users can drag and drop attributes from the corresponding model attribute list into the model page designer area. This allows for customization of the relevant attribute controls, model indicators, and scenario-specific information for the attributes, such as display name, attribute data dictionary, required or not, default value, attribute validation rules, and related prompts. External model information can also be loaded. Forms are created based on the dynamically expanded dimensional model. Relationship tables are configured between attribute fields and the data elements they correspond to, as well as between form attribute fields and hierarchies within the dimensional model. An influencing factor model is established, collecting underlying nodes or basic indicators, performing coupling calculations, defining relevant influencing factors (such as metric calculations that may be used in multidimensional cube analysis, highly complex aggregation operations, and multi-table joins), matching fitness, and related path branching rules to form a dynamic modeling framework. Model definitions can also be reset and copied to create new ones.

[0036] In this embodiment, the index attributes are used for the user to configure relevant attribute information that can participate in model definition on the management end page. It includes the model attributes of the model, display names, component types displayed in the model form (such as: single-line text, drop-down selection, cascading selection, and custom components, etc.), whether it is required, static default values, dynamic default values (such as: time values, user sets, permission sets, confidentiality levels, etc.), data dictionaries, attribute verification rules, and relevant prompt information, etc., and provides a reset function for model attributes.

[0037] S4: According to the generated extended dependency relationship model, implement association rules and constraint relationships.

[0038] In order to utilize the model attribute decomposition tree and its dependency relationship matrix to improve the requirements modeling of the domain knowledge model, it is necessary to define a business application model table, so as to discover the constraint relationships between potential business application scenarios and model data and the constraint relationships between business application scenarios. Establish an initial requirements model according to the requirements, and generate an extended requirements model in combination with domain knowledge. Summarize or discover new requirements based on the corresponding association rules on the basis of the original requirements, mainly including the mapping relationship between the original model and the application scenario, and the decomposition and dependency association relationships between models and between application scenarios.

[0039] S5: Execute an optimization program to improve the model requirements dependency network.

[0040] According to the above steps, the model requirements dependency network can be improved to the greatest extent with the least amount of work, providing more sufficient and powerful support for subsequent requirements analysis and solution decision-making. However, the accuracy of the generated extended dependency relationship depends on the correctness of domain knowledge and the existing knowledge of the target system. Since domain knowledge is only a summary of past experience and has certain timeliness, abstraction, and subjectivity, as business application scenarios become more refined, problems may occur where the current model analysis results deviate from the latest reality. Therefore, during this process, it is necessary to execute a certain auxiliary optimization program, configure model training parameters and rules, automatically load data and dynamically execute the optimization model to determine which models or rules can inherit the dependency relationships of the parent nodes. Through continuous dynamic adjustment and model optimization, a refined extended dependency relationship can be obtained to meet the requirements of knowledge conversion, relationship reasoning, and obtaining ideal conclusions.

[0041] Only some exemplary embodiments of the present invention have been described above by way of illustration. Without doubt, for those of ordinary skill in the art, the described embodiments can be modified in various different ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for generating domain knowledge modeling and dynamic optimization configuration, characterized by: The method is implemented by the following steps: Step 1: Define and manage model form information, and import and verify data sets; Step 2: Quantify the model form, aggregate domain knowledge, and calculate transitive dependencies as described in Step 1 to obtain a dynamically expanded dimensional model. Step 3: Define the original model, establish a mapping from the original model requirements to entity objects and indicator attributes, assist in building the relationship between requirements, and generate an extended dependency model; Step 4: Establish association rules and constraint relationships based on the extended dependency model generated in step 3; Step 5: Execute the optimization program to improve the model demand dependency network; In step 3, the user defines and sets properties for the original model on the management page. When creating the original model definition, a model designer function is provided. On the original model definition page, the user can drag and drop the model from the list of related properties corresponding to the model to the model page designer area. The function supports the configuration of personalized scene information for property controls, model indicators, and properties on the design page, and supports the loading of external model information. Establish a form based on the dynamically expanded dimensional model obtained in step 2, configure the relationship between attribute fields and the data elements corresponding to the attribute fields, and the relationship table between the form attribute fields and the levels within the dimensional model to which they belong; establish an influencing factor model, collect underlying nodes or basic indicators, perform coupling calculations, define the matching fitness of relevant influencing factors and related path branching rules, and form a dynamic modeling framework; at the same time, provide the functions of resetting model definitions and copying and creating new ones.

2. The method for generating domain knowledge modeling and dynamic optimization configuration according to claim 1, characterized in that: In step 1, define the content and classification of the model form and map the attribute information of the model form to the form attributes; Users manage model form information on the management page. By configuring the relationship between the model database storage form and the object JAVA class, they establish associations between form attributes, object JAVA classes, and database storage forms. At the same time, configure the model form related parameters and classification parameters, and set the uniqueness verification attribute set of the model form.

3. The method for generating domain knowledge modeling and dynamic optimization configuration according to claim 2, characterized in that: When the user maintains the model form attribute information on the management page, the user sets the uniqueness verification attribute set of the model form, including the ID, attribute name, type, length, data dictionary and dimension of the model form attribute; at the same time, the openness, authority scope, participation algorithm, application scenario, query configuration, export and loading attribute information of the relevant model form are configured.

4. The method for generating domain knowledge modeling and dynamic optimization configuration according to claim 1, characterized in that: In step 2, the model form quantitative definition specifically refers to the definition of model indicators, which is used to complete the transformation of domain knowledge from qualitative description to quantitative description, realize the aggregation of domain knowledge and calculation of transitive dependencies between model data attribute relationships, obtain the extended domain knowledge required for auxiliary model requirement modeling, and establish dimensional information of single influencing factors and coupled influencing factors; Based on the verification information in step one, the attribute fields in the data set are mapped to a pre-set dimensional model or a new indicator dimension is established to form a dynamically expanded dimensional model.

5. The method for generating domain knowledge modeling and dynamic optimization configuration according to claim 4, characterized in that: In step three, the indicator attributes are used by users to configure the attribute information involved in the model definition on the management page; this includes model attributes, display name, component type displayed in the model form, whether it is required, static default value, dynamic default value, data dictionary, attribute verification rules and prompt information, and provides a reset function for model attributes.

6. The method for generating domain knowledge modeling and dynamic optimization configuration according to claim 1, characterized in that: In step 4, association rules and constraint relationships are established based on the generated extended dependency model, specifically: Establish an initial demand model based on the needs, and generate an extended demand model based on domain knowledge. Summarize or discover new needs based on the original needs according to the corresponding association rules, including the mapping relationship between the original model and the application scenario, the comparison relationship between the original models, the decomposition between application scenarios, and the dependent association relationship.

7. The method for generating domain knowledge modeling and dynamic optimization configuration according to claim 1, characterized in that: In step 5, the optimization procedure is executed to improve the model requirement dependency network, specifically: Execute auxiliary optimization programs, configure model training parameters and rules, automatically load data and dynamically execute optimization models to determine which models or rules inherit the dependencies of the parent node; through continuous dynamic adjustment and model optimization, obtain refined extended dependencies.

Images