Conversational data entity creation method and system

Abstract

The invention relates to the technical field of electric digital data processing, in particular to structured data. According to the technical scheme, the method comprises the following steps: analyzing an input content calling model and segmenting data; generating a translation name text and performing matching according to keyword configuration; calling a creation interface to fill data entity information according to the output content, and performing front-end rendering; and verifying the stored result, obtaining sample data, obtaining relation information of the type and the keyword, and maintaining the relation information in the corresponding table. The system comprises a demand processing module, a type matching module and a data entity creating module, the demand processing module comprises a demand input unit and an analysis unit, the analysis unit is connected with the type matching module, and the type matching module is connected with the data entity creating module. The problems that in the prior art, efficiency is low, operation is tedious and cost is high are solved, the dialogue type data entity creating method and system are provided, and the purposes of being low in cost, easy and convenient to operate, high in automation level and high in creating efficiency are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic digital data processing, and in particular to structured data. Background Art

[0002] In the process of building digital products from 0 to 1, designers will use modeling tools to draw ER diagrams. However, the operation of these tools is currently complicated. Creating entities requires adding the Chinese name, English name and type of attributes one by one, which is extremely cumbersome and inefficient.

[0003] For example, the Chinese patent with publication number CN112380296A discloses an ER diagram generation method, system, electronic device and storage medium, and provides the following technical solutions. The present invention proposes an ER diagram generation method, system, electronic device and storage medium, and the method and technical solution thereof include an engine initialization step, creating a container tag in HTML, and introducing a canvas-based G6 engine in the container tag; an optimization configuration step, configuring ER diagram parameters in the container tag, and performing ER diagram generation optimization configuration; and an ER diagram generation step, generating an ER diagram according to the ER diagram parameters and the ER diagram generation optimization configuration. The present invention can solve the problems of low efficiency, cluttered and unsimplified interface of the ER diagram generation method. However, the above-mentioned ER diagram generation method, system, electronic device and storage medium require users to configure parameters in the canvas and improve the parameter content one by one in specific operations. Users need to perform complex operations to input parameter configuration, which greatly reduces the efficiency of entity creation. Summary of the invention

[0004] The present invention solves the problems of low efficiency, complicated operation and high cost in the prior art, and proposes a conversational data entity creation method and system, which achieves the goals of low cost, simple operation, high automation level and high creation efficiency.

[0005] To achieve the above object, the present invention adopts the following technical solutions: A conversational data entity creation method comprises the following steps: S1: Call the model to parse the input content and segment the data; S2: Generate translation name text and match it according to keyword configuration; S3: Call the creation interface to fill in data entity information based on the output content and perform front-end rendering; S4: Verify the saved results, obtain sample data and obtain the relationship information between types and keywords, and maintain them in the corresponding table.

[0006] A conversational data entity creation system includes: a demand processing module, a type matching module and a data entity creation module. The demand processing module includes a demand input unit and a parsing unit. The demand input unit is connected to the parsing unit, the parsing unit is connected to the type matching module, and the type matching module is connected to the data entity creation module.

[0007] The advantage of this design is that it automatically processes input content, reduces manual intervention through model analysis and keyword matching, improves data processing efficiency and accuracy, simplifies front-end rendering and relationship maintenance, and improves the intelligence and automation level of the overall system.

[0008] Preferably, step S2 comprises the following steps: S2.1: Take the entity name as input, call the translation interface, obtain the corresponding entity translation name, and set the entity name and the entity translation name as a group; S2.2: Take the attribute name as input, call the translation interface cyclically, obtain the attribute translation names corresponding to all attributes, and merge the entity translation name and the corresponding attribute translation name into the attribute translation name; S2.3: Set the attribute name and attribute translation name as a data group and establish the attribute type and key relationship table; S2.4: Loop through the keyword names and match them with the relational table to obtain the corresponding type names; S2.5: If no match is found, associate it with other attribute types and return to step S2.4; if a type name is obtained, and there is only one, this type shall prevail; if multiple type names are obtained, the one with the most acquisitions shall be the return value; S2.6: Output according to the matching content.

[0009] The benefit of this design is that the use of translation interfaces and data grouping methods enables the system to effectively support multi-language environments, organizes data through key relationship tables, enhances data consistency and manageability, and ensures the accuracy of attribute type matching.

[0010] Preferably, in step S2.5, if multiple type names are obtained and the maximum number of times they are obtained is the same, the one returned first is used as the return value; in the attributes that do not match the keywords, if the type is modified, the keywords and corresponding types of the split attribute name are used as sample data.

[0011] The advantage of this design is that it can obtain the best matching results, ensure the transparency and accuracy of the data processing process, and provide feedback for the continuous optimization of the system through the automatic generation of sample data.

[0012] Preferably, in step S4, the sample keywords and corresponding types are counted in the corresponding relationship table to be supplemented; the relationship table to be supplemented includes keyword information, corresponding type information and relationship values, and the relationship value is accumulated and increased by 1 each time; the same keyword is sorted according to the relationship value size of the corresponding type, and if the maximum value minus the remaining average value is greater than a certain value, this association relationship is added to the attribute type and key relationship table.

[0013] The advantage of this design is that it reduces the burden of manual management by automatically counting and updating relationship tables, and dynamically maintains attribute type associations, making the system more flexible and greatly improving adaptability.

[0014] Preferably, the data entity information includes header information, attribute information and appendix information, the header information includes entity name, entity translation name and entity ID, the appendix information includes ID information and name information, the ID information includes associated entity ID and attribute ID, the name information includes attribute name, attribute translation name and attribute type, and the entity ID is automatically generated according to the date and four-digit self-incrementing serial number rules.

[0015] The advantage of this design is that it provides clear structured information definitions, facilitates data management and query, and automatically generates various data, reducing the possibility of human errors and improving data management efficiency and security.

[0016] Preferably, in step S3, the attribute names, attribute translation names and attribute types in several data groups corresponding to the entity are cyclically filled in the name information in the appendix, and the attribute ID is automatically generated according to the entity ID and the two-digit auto-increment serial number rule. After the data entity information is filled and the result is saved, the front-end canvas renders the entity according to the data returned by the back-end interface.

[0017] The advantage of this design is that it automatically performs data filling and interface rendering, which speeds up the entity creation and display process. At the same time, by automatically generating various data, it ensures the consistency and integrity of the data and improves the user experience.

[0018] Preferably, in step S2.6, the entity name and the entity translation name are taken as one group, and the attribute name, the attribute translation name and the attribute type are divided into several data groups according to the attribute as a dimension.

[0019] Preferably, the input content includes an entity name and an attribute name, and the entity name and each attribute name are obtained after segmenting the data; if parsing is unavailable, a re-entry prompt is given.

[0020] The advantage of this design is that it provides real-time parsing and input prompt functions, improves user experience, and reduces incorrect input.

[0021] Preferably, the demand input unit passes the input content to the parsing unit, the parsing unit passes the parsed structured information to the type matching module for attribute matching and type association, the type matching module passes the matched attribute type information to the data entity creation module, the data entity creation module receives the passed matching results, creates corresponding data entities and attributes, and ensures the correct configuration of their relationships.

[0022] The benefit of this design is that it enhances the security of the system and data protection measures, ensures the correctness of data entity creation and configuration, reduces the risk of configuration errors, and improves the level of data security management.

[0023] Compared with the prior art, the invention has the following beneficial effects.

[0024] 1. The present invention makes full use of the loop to query the keyword and attribute type relationship table, and identifies the best association relationship according to the preset rules. By automatically associating attribute types with attributes, users no longer need to manually configure the association relationship one by one, which significantly improves the efficiency of data processing, makes the entire process fast and reduces human errors, and provides a reliable mechanism to ensure the accuracy and consistency of the association relationship.

[0025] 2. The automatic supplement and maintenance mechanism of the keyword and attribute type relationship table introduced in the present invention greatly reduces the cost of manual maintenance of the association relationship table. The system can adaptively adjust and improve the association relationship through dynamic updating and automatic accumulation of relationship values. This not only reduces the necessity of manual intervention, but also improves the intelligence level of the system, ensures the efficiency and accuracy of data management, and greatly reduces operating costs.

[0026] 3. The present invention improves the user experience through highly automated data filling and front-end rendering processes. The system automatically generates and manages data, ensures consistency and integrity, and quickly renders data entities to the user interface, which not only speeds up the entity creation and display process, but also reduces the risk of human error. At the same time, it maintains the responsiveness of the user interface and the accuracy of data display, greatly enhancing user satisfaction in using the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 The present invention provides an overall flow chart of a conversational data entity creation method.

[0028] Figure 2 A system block diagram of a conversational data entity creation system of the present invention. DETAILED DESCRIPTION

[0029] In order to make the purpose, technical scheme and advantages of the present disclosure clearer, the following will further describe the embodiments of the present disclosure in detail with reference to the accompanying drawings. The proportions of the components are not drawn according to the actual proportions, and the proportions and sizes shown in the accompanying drawings should not limit the substantial technical scheme of the present invention. These embodiments do not describe all the details in detail, nor do they limit the invention to the specific embodiments described.

[0030] See also Figure 1-2 As shown, a conversational data entity creation method includes the following steps: S1: Call the model to parse the input content and segment the data; S2: Generate translation name text and match it according to keyword configuration; S3: Call the creation interface to fill in data entity information based on the output content and perform front-end rendering; S4: Verify the saved results, obtain sample data and obtain the relationship information between types and keywords, and maintain them in the corresponding table.

[0031] A conversational data entity creation system includes: a demand processing module, a type matching module and a data entity creation module. The demand processing module includes a demand input unit and a parsing unit. The demand input unit is connected to the parsing unit, the parsing unit is connected to the type matching module, and the type matching module is connected to the data entity creation module.

[0032] The present invention designs a conversational data entity creation method and system, which uses AI semantic analysis to obtain the name of the entity to be created and the name of the attribute, automatically obtains the attribute type through keyword and attribute type relationship table query, and obtains the English name according to the English name generation rule, and then realizes a new interactive mode of conversational entity creation, thereby comprehensively improving the efficiency of ER diagram or data modeling.

[0033] like Figure 1 In one embodiment shown, Figure 1 The overall flow chart of a conversational data entity creation method of the present invention. First, the system receives the content input by the user, including entity names and attribute names. In order to ensure the integrity and correctness of the data, the system calls the model to parse and segment the input data. The parsed data is converted into a structured form that is easier to manage. If the parsing fails, the system will prompt the user to re-enter to ensure the quality of the input. Next, the system translates and matches the entity and attribute names. This step obtains the translated names of the entities and attributes by calling the translation interface to support multi-language environments. The entity name and its translated name are set as a group, and the attribute name and its translated name are also stored in groups. Through this process, the system establishes attribute type and key relationship tables, laying the foundation for subsequent automatic matching.

[0034] On this basis, the system cyclically matches keywords and relationship tables to obtain the corresponding type names of the attributes. For attributes that cannot be matched, the system temporarily associates them with other attribute types and tries to match again. If the type name is successfully matched, the matching result is used directly; if there are multiple matching results, the type with the most occurrences is selected as the final result. In this step, if the attribute type is modified, the system saves the corresponding keywords and types as sample data for future optimization.

[0035] Afterwards, the system calls the creation interface and fills in the data entity information according to the processed output content. During the data filling process, the system automatically generates an attribute ID to ensure the uniqueness of each attribute and saves the data results. After the data is saved, the front-end canvas renders the entity through the data returned by the back-end interface to achieve a visual display of the data.

[0036] Finally, the system verifies the saved results and obtains sample data to maintain the relationship information between types and keywords. Sample keywords and their corresponding types are counted in the to-be-supplemented relationship table, which is dynamically updated and the relationship value is accumulated each time it is counted. The system sorts by relationship value. When the difference between the maximum value and the remaining average value is greater than a certain threshold, the association is added to the attribute type and key relationship table.

[0037] A conversational data entity creation method designed by the present invention parses and translates the call of the interface through automation model, and uses the FOR loop to query the relationship between keywords and attribute types. The system automatically selects the best association relationship according to the rules, realizes the automatic association of attribute types, and eliminates the tedious steps of manual configuration one by one, thereby greatly improving the efficiency. In addition, the automatic supplementary maintenance mechanism greatly reduces the cost of manual maintenance by dynamically updating the keyword and attribute type relationship table, while ensuring the real-time and accuracy of the data. This mechanism enhances the intelligence and adaptability of the system, enabling it to flexibly respond to changes. The automatic generation and management of data reduces the possibility of human errors and improves the consistency and security of data. At the same time, the fast front-end rendering improves the user experience and ensures the accuracy of data display.

[0038] In one embodiment, the following scheme is adopted to implement, which includes entity creation requirement management, attribute type matching, entity and attribute creation as a whole: Demand management: including three core functions: demand input, AI semantic analysis, and AI English generation; 1. Demand input: mainly includes the demand input box and the creation interface call entry. The user can simply describe the name and attribute name of the entity to be created in the input box, and then click the create button to start calling the AI ​​model for semantic analysis; 2. AI semantic analysis: Call the AI ​​big model to perform semantic analysis and convert the user input characters into structured content, such as entity name = XX; attribute 1 name = XX; attribute 2 name = XX; 3. AI English generation: Based on the attribute name obtained based on semantic analysis, take it as input, call the AI ​​big model, and output the corresponding English. The type generation rule is: entity name English_attribute name English.

[0039] Type matching: includes three core functions: type management, keyword setting, and matching rules; 1. Type management: including adding, deleting and modifying type names. Each type has a unique ID and the type name cannot be repeated; 2. Keyword setting: refers to maintaining the relationship table between keywords and attribute types, including keyword ID, keyword content, corresponding attribute type ID, and corresponding attribute type name; (1) Manual addition: includes adding and deleting keywords, and associating keywords with corresponding types. Each keyword can only be associated with one type at a time, that is, when creating a keyword, a unique check is required; (2) Automatic supplement: When the attribute name does not match the keyword, the value is associated with other types. When the user changes the type, the changed type is considered the ideal corresponding type. Then the attribute name needs to be decomposed into keywords, and the relationship value between each keyword and type is increased by 1. Then the maximum value of the relationship value between the type and the keyword is selected, and then the average value of all other types is subtracted. If the value obtained is greater than 10, this keyword and the corresponding attribute type are maintained in the table; 3. Matching rules: Fuzzy matching is performed based on keywords. That is, the match is successful when the type name contains keywords, and the type of this attribute is set to the keyword-associated type. If the keyword is not matched, the type is associated with other types.

[0040] Entity creation: includes three core functions: entity name configuration, attribute name configuration, and attribute type configuration; 1. Entity name configuration: includes adding, modifying and deleting entity names. Entity names cannot be repeated, and the corresponding unique entity id is automatically generated according to the rule of st+date+four-digit auto-increment number; 2. Attribute name configuration: mainly includes the addition, deletion, and modification of the Chinese and English names of attributes. The corresponding unique attribute id is automatically generated according to the rule entity id + two auto-increment digits, and the attribute id and entity id are in a many-to-one correspondence, that is, an entity can contain multiple attributes, but the same attribute can only be associated with one id; 3. Attribute type configuration: You can obtain the type id and name from the type table and associate them with the attribute id. The relationship between attributes and attribute types is many-to-many. You can maintain the relationship by creating an intermediate table to associate attributes and types. The specific process steps are as follows: 1. Enter the creation requirements: (1) Input specification: must include entity name and attribute name; example: entity name = student; attributes include: student ID, name, gender, age; prompts will be set in the input box to guide users; (2) Content verification: Verify whether the content is empty. If it is empty, prompt the user to enter the content.

[0041] 2. Call AI analysis: (1) Interface call: call the API of AI_GLM-4-LONG, send the creation request to the model server, and obtain the interface return string; (2) Data storage: The AI ​​interface obtains the string and stores it in the entity table, where the entity main table contains: entity ID (unique identifier), which is automatically generated according to (st+year-month-day+four-digit auto-increment), and entity name, which is assigned a value according to the string; entity appendix: contains entity name and entity ID (unique identifier), which is automatically generated according to (sx+year-month-day+four-digit auto-increment).

[0042] 3. Generate English name: (1) Obtaining the English name of an entity: Take the Chinese name of the entity as input, call the AI ​​translation interface, and obtain the corresponding English name; and set the Chinese name and English name of the entity as a group; (2) Obtaining the English name of an entity: Taking the Chinese name of an attribute as input, the AI ​​translation interface can be called cyclically to obtain the English name corresponding to the attribute; and the English name of the attribute is generated in the format of entity English name_attribute English name, such as student_name; and the Chinese name of the attribute and the English name of the attribute are set as a data group; The specific method is as follows: Interface call: call the API of AI_GLM-4-0520, send the name of the entity to the end of the model service, and obtain the result returned by the interface, obtain the corresponding English name, and store it in the English name field of the entity table; Attribute English name concatenation: call the API of AI_GLM-4-0520, send the name of the entity to the model service end, obtain the result returned by the interface, obtain the corresponding English, and then generate the attribute English name according to the format of entity English name_attribute English name, such as student_name; and store it in the attribute English name field of the entity appendix.

[0043] 4. Get the attribute type: (1) Keyword configuration: Type and key relationship table, divided into keyword id, keyword name, corresponding attribute type id, corresponding attribute type name. The same keyword can only be associated with one attribute type, but one attribute type can be associated with multiple keywords. For example, for photo type, the keywords that can be configured are: photo, picture; for number type, the keywords that can be configured are: quantity, amount, number; for enumeration type, the keywords that can be configured are: status, gender, type, region, unit, etc.; for character type, the keywords that can be configured are others; (2) Keyword matching: Loop through the keyword name and the relationship table to obtain the corresponding type name; if no type is obtained, return the associated attribute type; if a type name is obtained, and there is only one, then this type will prevail; if multiple type names are obtained, the one with the most acquisitions will be returned first, followed by the one that has been returned first; (3) Output matching content: Entity name and entity English name are grouped as one, and attribute Chinese name, attribute English name, and attribute type are divided into multiple data groups according to the attribute as the dimension.

[0044] 5. Call the creation interface: (1) Entity information description: The entity table is divided into a header and an appendix. The header contains three fields: Chinese name, English name, and entity id; the appendix contains five fields: attribute id, attribute name, attribute English name, attribute type, and associated entity id; (2) Fill in the header information: The entity name and English name are filled in the Chinese name and English name fields in the corresponding header; the entity id is automatically generated according to the rule of st+date+four-digit auto-increment serial number; (3) Filling in attribute information: Loop through the Chinese names, English names and types in the multiple data groups corresponding to the entity and fill them in the attribute names, attribute English names and attribute types in the attached table respectively; the attribute id is automatically generated based on the entity id + two digits of auto-increment; the associated attribute id = the corresponding entity id; after filling in, save the result.

[0045] 6. Front-end rendering results: The front-end canvas renders the entity according to the data returned by the back-end interface.

[0046] 7. Verify and save content: Operators can verify data based on the front-end rendering results, and can modify, delete and add operations; and save the final results after confirmation.

[0047] 8. Get sample data: For attributes that do not match keywords, if the type is manually modified, the keywords and corresponding types of the split attribute names are used as sample data.

[0048] 9. Obtain the type and keyword relationship: Maintain the type and keyword relationship in the corresponding table for subsequent matching: (1) Sample data processing: Count the sample keywords and corresponding types in the corresponding relationship table to be supplemented; the fields in the table include: keyword id, keyword name, corresponding type id, corresponding type name, relationship value; the relationship value is accumulated and increased by 1 each time; (2) Association confirmation rule: Sort the same keyword according to the relationship value of the corresponding type. If the value obtained by subtracting the remaining average from the maximum value is greater than 10 (10 is used in this embodiment, which can be flexibly configured in practice), then this association is added to the keyword and type relationship table.

[0049] Among them, data entities refer to some very important people, things, objects, places and other information in the business field, which is the mapping of actual business in the information world, such as products, suppliers, expense reports, etc.; entity attributes are part of the entity, which are used to describe the characteristics or attributes of the entity. For example, students include: student number, name, gender, age, photo and other attributes; attribute type means that each attribute has a value domain, and its type can be character, integer, picture, enumeration, etc.; for example, in students, student number is a character, gender is an enumeration, photo is a picture, age is an integer, etc.

[0050] like Figure 2 In one embodiment shown, Figure 2 The system block diagram of a conversational data entity creation system of the present invention. A conversational data entity creation system designed by the present invention consists of a demand processing module, a type matching module and a data entity creation module. The demand processing module includes a demand input unit and a parsing unit. The demand input unit is responsible for receiving user input and passing its content to the parsing unit for data parsing. After the parsing unit converts the input data into structured information, it passes it to the type matching module. The type matching module undertakes the task of attribute matching and type association, and transmits the result to the data entity creation module by matching the attributes and types. The data entity creation module is responsible for creating corresponding data entities and attributes based on the received matching results, ensuring the correctness and consistency of their relationship configuration. Through the close connection and efficient collaboration between modules, the entire system realizes the rapid and accurate creation of data entities, and improves the user experience and the intelligent level of data management.

[0051] In one embodiment, a conversational data entity creation system designed by the present invention mainly includes the following three modules: a demand processing module, a type matching module and a data entity creation module. Among them, the demand processing module includes three sub-management functions: demand input management, semantic parsing management and English generation management. The demand input management is responsible for receiving the user's input and verifying the input format to ensure the integrity and correctness of the data. Next, the semantic parsing management calls the AI ​​model to perform semantic parsing on the input, and processes the parsing results to ensure its validity and integrity. Subsequently, the English generation management converts the parsed attribute names into English format and handles any abnormal situations that may arise. After the structured information is processed, it is passed to the type matching module for further processing.

[0052] The type matching module includes three parts: type management, keyword management, and matching rule management. Type management is responsible for adding, deleting, and modifying attribute types to ensure their consistency. Keyword management is responsible for maintaining the relationship between keywords and attribute types, and handling the automatic supplement and update of keywords. Matching rule management confirms attribute types by fuzzy matching based on keywords, and handles situations where keywords are not matched. By receiving structured information from the demand processing module, the type matching module performs keyword matching and type confirmation, and then passes the matching results to the data entity creation module.

[0053] In the data entity creation module, entity configuration management is responsible for adding, modifying and deleting entity names to ensure the uniqueness of entities. Attribute configuration management handles the addition, deletion and modification of attributes to ensure the uniqueness of attributes and their correct association with entities. Type association management establishes and maintains many-to-many relationships between attributes and attribute types to ensure the accuracy and consistency of these relationships. This module uses the matching results received from the type matching module to create corresponding entities and attributes and maintain the relationships between them.

[0054] The entire system achieves collaboration through the following connections: user input is initially managed and parsed by the demand processing module, and the results are passed to the type matching module for keyword matching and type confirmation. Finally, the data entity creation module creates data entities and attributes based on the matching results to ensure the correct configuration of all relationships. This modular design ensures the flexibility and scalability of the system, making the data processing process efficient and accurate.

[0055] In summary, the present invention provides a conversational data entity creation method and system, which calls the AI ​​semantic parsing function to standardize the input content, intercepts key information, and uses AI to generate English names according to rules, and then realizes automatic binding of attribute types based on association relationship queries and automatic supplementary maintenance, thereby realizing conversational creation of entities, quickly improving the efficiency of ER diagram drawing and modeling in digital application design, and thus reducing the design cost of enterprise digital products.

[0056] The present invention is not limited to the above-mentioned embodiments. No matter any changes are made in the shape or material composition, any structural design provided by the present invention is a variation of the present invention and should be considered to be within the protection scope of the present invention.

Claims

1. A conversational data entity creation method, characterized in that: The following steps are involved: S1: Call the model to parse the input content and segment the data; S2: Generate translation name text and match it according to keyword configuration; S3: Call the creation interface to fill in data entity information based on the output content and perform front-end rendering; S4: Verify the saved results, obtain sample data and obtain the relationship information between types and keywords, and maintain them in the corresponding table.

2. A conversational data entity creation method according to claim 1, characterized in that: The step S2 comprises the following steps: S2.1: Take the entity name as input, call the translation interface, obtain the corresponding entity translation name, and set the entity name and the entity translation name as a group; S2.2: Take the attribute name as input, call the translation interface cyclically, obtain the attribute translation names corresponding to all attributes, and merge the entity translation name and the corresponding attribute translation name into the attribute translation name; S2.3: Set the attribute name and attribute translation name as a data group and establish the attribute type and key relationship table; S2.4: Loop through the keyword names and match them with the relational table to obtain the corresponding type names; S2.5: If no match is found, associate it with other attribute types and return to step S2.4; if a type name is obtained, and there is only one, this type shall prevail; if multiple type names are obtained, the one with the most acquisitions shall be the return value; S2.6: Output according to the matching content.

3. A conversational data entity creation method according to claim 2, characterized in that: In step S2.5, if multiple type names are obtained and the maximum number of times they are obtained is the same, the first one returned is used as the return value; in the attribute that does not match the keyword, if the type is modified, the keyword and corresponding type split from the attribute name are used as sample data.

4. A conversational data entity creation method according to claim 2, characterized in that: In the step S4, the sample keywords and corresponding types are counted in the corresponding relationship table to be supplemented; the relationship table to be supplemented includes keyword information, corresponding type information and relationship values, and the relationship value is accumulated and increased by 1 each time; the same keyword is sorted according to the relationship value size of the corresponding type, and if the maximum value minus the remaining average value is greater than a certain value, this association relationship is added to the attribute type and key relationship table.

5. A conversational data entity creation method according to any one of claims 1 to 4, characterized in that: The data entity information includes header information, attribute information and appendix information, the header information includes entity name, entity translation name and entity ID, the appendix information includes ID information and name information, the ID information includes associated entity ID and attribute ID, the name information includes attribute name, attribute translation name and attribute type, and the entity ID is automatically generated according to the date and four-digit self-increment serial number rules.

6. A conversational data entity creation method according to claim 5, characterized in that: In the step S3, the attribute names, attribute translation names and attribute types in several data groups corresponding to the entity are cyclically filled in the name information in the appendix, and the attribute ID is automatically generated according to the entity ID and the two-digit auto-increment serial number rule. After the data entity information is filled and the result is saved, the front-end canvas renders the entity according to the data returned by the back-end interface.

7. A conversational data entity creation method according to claim 2 or 3, characterized in that: In the step S2.6, the entity name and the entity translation name are taken as a group, and the attribute name, the attribute translation name and the attribute type are divided into several data groups according to the attribute as a dimension.

8. A conversational data entity creation method according to claim 1, characterized in that: The input content includes entity name and attribute name. After segmenting the data, the entity name and each attribute name are obtained; if the parsing fails, a re-entry prompt is given.

9. A conversational data entity creation system, using a conversational data entity creation method according to any one of claims 1 to 8, characterized in that: include: A demand processing module, a type matching module and a data entity creation module, wherein the demand processing module includes a demand input unit and a parsing unit, wherein the demand input unit is connected to the parsing unit, wherein the parsing unit is connected to the type matching module, and wherein the type matching module is connected to the data entity creation module.

10. A conversational data entity creation system according to claim 9, characterized in that: The demand input unit passes the input content to the parsing unit, and the parsing unit passes the parsed structured information to the type matching module for attribute matching and type association. The type matching module passes the matched attribute type information to the data entity creation module. The data entity creation module receives the passed matching results, creates corresponding data entities and attributes, and ensures the correct configuration of their relationships.

Images