An intelligent knowledge architecture diagram construction method and system based on large models

Abstract

The present application discloses a method and system for constructing an intelligent knowledge architecture diagram based on a large model, belonging to the field of artificial intelligence technology, including: obtaining multiple entities from multiple text blocks; classifying the entities and relationships to obtain entity types and relationship types; extracting common features of the entities to generate entity types with attributes; combining the entity types with attributes to generate relationship type triples, constructing a preliminary knowledge architecture diagram according to the relationship type triples, and adding and deleting nodes and edges to generate guiding suggestions; extracting new entities and relationships from subsequent text blocks based on the suggestions, updating the relationship type triples, and generating a final knowledge architecture diagram when the maximum number of iteration rounds is reached. The present application uses a large model to replace humans in constructing the knowledge architecture diagram, reducing the labor cost and time cost, and adopts an iterative construction method, using a small amount of modification by humans to some knowledge architecture diagrams to infer the user's demand tendency, ensuring the accuracy and comprehensiveness of the knowledge architecture diagram.

Description

Technical Field

[0001] This application belongs to the field of artificial intelligence technology, and particularly relates to a method and system for constructing an intelligent knowledge architecture diagram based on a large model. Background Art

[0002] With the advent of the large model era, the automated construction of knowledge graphs has gradually become the focus of attention. The application of large models has enabled the construction of knowledge graphs to be automated to a certain extent. However, before constructing a knowledge graph, domain experts are still required to design a comprehensive and accurate knowledge architecture diagram. The knowledge architecture diagram contains the common features of the entities and semantic relationships expected by users, namely entity types and relationship types. These entity types and relationship types will be used to guide the subsequent construction of the knowledge graph to ensure that the entities and semantic relationships included in the knowledge graph meet the needs of users. The quality of the knowledge architecture diagram will directly affect the effect of the final knowledge graph.

[0003] However, the existing technologies have the following deficiencies: First, the construction of existing knowledge architecture diagrams is usually carried out manually, resulting in low construction efficiency; Second, the knowledge architecture diagrams constructed manually often miss key entity types and relationship types, resulting in incomplete information in the knowledge architecture diagram, which in turn affects the reliability and quality of the knowledge graph, resulting in low accuracy and comprehensiveness of the constructed knowledge architecture diagram. Summary of the Invention

[0004] The purpose of the embodiments of this application is to provide a method and system for constructing an intelligent knowledge architecture diagram based on a large model, which can solve the problems of low construction efficiency, accuracy, and comprehensiveness in the process of constructing a knowledge architecture diagram in the existing technologies.

[0005] To solve the above technical problems, this application is implemented as follows:

[0006] In a first aspect, the embodiments of this application provide a method for constructing an intelligent knowledge architecture diagram based on a large model, the method comprising:

[0007] S1. Obtain a plurality of entities from within a first text block among a plurality of sequentially connected text blocks;

[0008] S2. Obtain semantic relationships between the plurality of entities and construct a preliminary knowledge architecture diagram;

[0009] S3. Add and delete nodes and edges of the preliminary knowledge architecture diagram to obtain a knowledge architecture diagram, and store the added and deleted entity types in a database;

[0010] S4. Based on the large model, fuse the added and deleted entity types in the database to generate a guidance recommendation text for the current round and update the iteration round ;

[0011] S5. When the iteration round is reached, based on the guiding advice text of the round, obtain the entities of the round and the semantic relationships of the round from the th text block, and generate the relationship instance triples of the round;

[0012] S6. Based on the large model, perform entity classification and semantic classification on the entities of the round and the semantic relationships of the round respectively, obtain the updated entity types and relationship types, and construct the updated relationship type triples;

[0013] S7. Merge the relationship type triples of the round with the updated relationship type triples to obtain the preliminary knowledge architecture diagram of the round, and return to step S3 until the current iteration round reaches the preset maximum iteration round, at which point the iteration stops and the final knowledge architecture diagram is generated.

[0014] As an alternative implementation manner of the first aspect of the present application, the specific steps for constructing the preliminary knowledge architecture diagram include: constructing relationship instance triples based on semantic relationships and entities; performing entity classification and semantic classification on entities and semantic relationships respectively to obtain entity types and relationship types; obtaining common features from the text blocks where multiple entities corresponding to each entity type are located based on the large model to obtain entity types with attributes; constructing multiple relationship type triples based on the entity types with attributes, relationship types, and relationship instance triples, and obtaining the preliminary knowledge architecture diagram based on the relationship type triples.

[0015] As an alternative implementation manner of the first aspect of the present application, the process of obtaining multiple sequentially connected text blocks includes:

[0016] According to the number of iteration rounds input by the user, the large model evenly divides the input text into multiple sequentially connected text blocks. Starting from the second text block, each text block retains one-tenth of the end of the previous text block as the connection part at the beginning of the current text block.

[0017] As an alternative implementation manner of the first aspect of the present application, the relationship types include: genus-species relationship, composition relationship, and location relationship.

[0018] As an alternative implementation manner of the first aspect of the present application, the preliminary knowledge architecture diagram is composed of multiple edges and multiple nodes. Among them, the nodes represent entity types, and the edges represent relationship types.

[0019] As an alternative implementation of the first aspect of the present application, the process of constructing multiple relationship type triples based on the triples of entity types with attributes, relationship types, and relationship instances includes:

[0020] );

[0021] Among them, respectively represent two different entity types, represents the entity type 's attribute list, represents the entity type 's attribute list, represents the entity type and the entity type 's semantic relationship, represents the entity type 's corresponding number of attributes.

[0022] As an alternative implementation of the first aspect of the present application, the common feature represents the static and dynamic features that are used to depict the specific instances in the same entity type, which appear in the text paragraphs corresponding to multiple entities of the same entity type and whose occurrence times are greater than or equal to the preset times according to the large model analysis.

[0023] In the second aspect, the embodiments of the present application provide an intelligent knowledge architecture diagram construction system based on a large model. The system includes:

[0024] A text chunking module, configured to evenly chunk the user input text, obtain multiple sequentially connected text chunks, and obtain multiple entities from the first text chunk among the multiple sequentially connected text chunks;

[0025] A preliminary knowledge architecture diagram construction module, configured to obtain the semantic relationships between multiple entities and construct a preliminary knowledge architecture diagram;

[0026] An artificial review and modification module, configured to add and delete nodes and edges of the preliminary knowledge architecture diagram to obtain a knowledge architecture diagram;

[0027] A storage module, configured to store the added and deleted entity types in a database;

[0028] A suggestion generation module, configured to fuse the added and deleted entity types stored in the database according to the large model and generate a guidance suggestion text;

[0029] A knowledge architecture diagram update module, configured to iteratively update the knowledge architecture diagram according to the guidance suggestion text to generate a final knowledge architecture diagram.

[0030] In a third aspect, an embodiment of the present application provides an electronic device, which includes a processor, a memory, and a program or instruction stored on the memory and executable on the processor. When the program or instruction is executed by the processor, the steps of the method according to the first aspect are implemented.

[0031] In a fourth aspect, an embodiment of the present application provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the method according to the first aspect are implemented.

[0032] Compared with the prior art, a method for constructing an intelligent knowledge architecture diagram based on a large model proposed in the present application uses the large model to replace humans to construct the knowledge architecture diagram, greatly reducing the labor cost and time cost. In addition, the present application adopts an iterative construction method, using a small number of modifications by humans to a part of the knowledge architecture diagram to infer the user's demand tendency, ensuring the accuracy and comprehensiveness of the knowledge architecture diagram during the construction process. Description of the Drawings

[0033] Figure 1 is a flowchart of a method for constructing an intelligent knowledge architecture diagram based on a large model provided by the first embodiment of the present application;

[0034] Figure 2 is a schematic diagram of the knowledge architecture of a method for constructing an intelligent knowledge architecture diagram based on a large model provided by the first embodiment of the present application;

[0035] Figure 3 is a structural diagram of a system for constructing an intelligent knowledge architecture diagram based on a large model provided by the second embodiment of the present application. Detailed Embodiments

[0036] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present application.

[0037] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, rather than to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present application can be implemented in an order other than those illustrated or described herein. In addition, "and / or" in the specification and claims means at least one of the connected objects, and the character " / " generally represents an "or" semantic relationship between the associated objects before and after.

[0038] The following will combine the accompanying drawings to provide a detailed description of a method and system for constructing an intelligent knowledge architecture diagram based on a large model according to the embodiments of the present application through specific embodiments and their application scenarios.

[0039] Embodiment 1

[0040] Please refer to Figure 1 , which is a flowchart of a method for constructing an intelligent knowledge architecture diagram based on a large model proposed in the embodiments of the present application. The proposed method includes steps S1 to S7.

[0041] Step S1: Obtain multiple entities from the first text block among multiple sequentially connected text blocks.

[0042] Specifically, an entity is a class object created for a preset entity class. The multiple sequentially connected text blocks are obtained after the large model evenly divides the user input text into blocks. Starting from the second text block, each text block retains one-tenth of the end of the previous text block as the connection part at the beginning of the current text block. For example: When the user provides 10,000 words of text and requests ten iterations, the system will evenly cut the text into ten blocks, each block containing 1,100 words, where 100 words are the end part of the previous text block.

[0043] Furthermore, when the system obtains multiple entities from the first text block, it filters them from the text block according to the user's needs. The user's needs are a text description and belong to the user's request. For example: The user inputs the requirement: "I need to construct a knowledge graph about the principles and functions of an operating system." After the user inputs this requirement and sends it to the system, after the system receives the user's needs, it uses the large model to filter out the entities that meet this requirement from the first text block.

[0044] In the present application, the text is cut into multiple blocks and the knowledge architecture diagram is constructed in multiple rounds. Only one text block is processed in each round to construct a part of the knowledge architecture diagram, and then the user is allowed to modify this part of the knowledge architecture diagram. The system summarizes the modification information to guide the construction of the subsequent knowledge architecture diagram. In this way, the large model can better understand the user's needs during the construction process, and thus construct a comprehensive and accurate knowledge architecture diagram with less manual participation. The present application retains one-tenth of the end of the previous text block as the connection part, which can also ensure the semantic coherence of the text blocks.

[0045] Step S2: Obtain the semantic relationships between multiple entities and construct a preliminary knowledge architecture diagram.

[0046] Specifically, the specific steps for constructing a preliminary knowledge architecture diagram include: constructing relation instance triples based on semantic relations and entities; performing entity classification and semantic classification on entities and semantic relations respectively to obtain entity types and relation types; obtaining common features from the text blocks where multiple entities corresponding to each entity type are located based on a large model to obtain entity types with attributes; constructing multiple relation type triples based on the entity types with attributes, relation types, and relation instance triples, and obtaining the preliminary knowledge architecture diagram based on the relation type triples.

[0047] Specifically, the specific process of constructing relation instance triples based on semantic relations and entities includes: the system uses a large model to find out whether there is a semantic relation between every two entities from multiple sequentially connected text blocks. If there is a semantic relation, it uses concise words to summarize the semantic relation between them, and represents it with a relation instance triple ( , , ), where and represent two different entity types, and represents the semantic relation between these two entity types. For example, when the text block shows: "ChatGPT is an artificial intelligence chatbot developed by OpenAI and can understand and generate natural language.", since this text block contains two entities, "ChatGPT" and "artificial intelligence chatbot", the large model can analyze the semantic relation "belongs to" between the two, and thus obtain the relation instance triple (ChatGPT, belongs to, artificial intelligence chatbot).

[0048] Furthermore, when performing entity classification and semantic classification on entities and semantic relations respectively to obtain entity types and relation types, entity types include but are not limited to: teacher, doctor, company, school, occupation, etc., and relation types include: genus-species relation, composition relation, and location relation. Among them, the genus-species relation means: A is B, the composition relation means: A is a component part of B, and the location relation means: A is located at B. Here, A and B represent two different entities.

[0049] Further, based on the large model, common features are obtained from the text blocks where multiple entities corresponding to each entity type are located, resulting in entity types with attributes. Among them, the common features refer to those that appear in the text paragraphs where multiple entities corresponding to the same entity type are located according to the large model analysis, and the number of occurrences is greater than or equal to the preset number, and are used to describe the static and dynamic features of specific instances in the same entity type. For example: for the entity type of students corresponding to entities such as Zhang San and Li Si, the text paragraphs where they are located are respectively: "Zhang San is 17 years old this year, is a student in Class 1, Senior Three, and his student number is 1023" and "Li Si is 18 years old this year, is a student in Class 1, Senior Three, and his student number is 1024", and the system will infer class, age, and student number as the attributes of the entity type of students. The entity type with attributes can be expressed as:

[0050] ;

[0051] where et1 represents the entity type, represents the attribute list of the entity type et1, represents each different attribute of the entity type et1.

[0052] Further, the specific process of constructing multiple relationship type triples based on the entity type with attributes, relationship type, and relationship instance triples, and obtaining the preliminary knowledge architecture diagram includes: after replacing the two entity types , , in the relationship instance triple ( and and the semantic relationship with the corresponding entity type with attributes and relationship type respectively, the relationship type triple of the current round is obtained; multiple relationship type triples of the current round are combined to construct the preliminary knowledge architecture diagram. Among them, the relationship type triple of the current round is expressed as:

[0053] );

[0054] where, respectively represent two different entity types, represents the entity type 's attribute list, represents the entity type 's attribute list, represents the semantic relationship between the entity type and the entity type , represents the number of attributes corresponding to the entity type .

[0055] Step S3: Add or delete nodes and edges in the preliminary knowledge architecture diagram to obtain the knowledge architecture diagram, and store the added or deleted entity types in the database.

[0056] Specifically, the preliminary knowledge architecture diagram consists of multiple edges and multiple nodes. Among them, the nodes represent entity types, and the edges represent relationship types. The specific steps of adding or deleting nodes and edges in the preliminary knowledge architecture diagram to obtain the knowledge architecture diagram and storing the added or deleted entity types in the database include: The system displays the preliminary knowledge architecture diagram of the current round to the user in the form of a directed graph. The user reviews and modifies the preliminary knowledge architecture diagram of the current round (i.e., performs addition and deletion operations on nodes and edges) to obtain the knowledge architecture diagram of the current round. During the user's modification process, the system stores the entity types added or deleted by the user in the database.

[0057] During the review and modification process, the user will review and analyze each relationship type triple in the preliminary knowledge architecture diagram generated in each round, modify the incorrect relationship type triples. The system will record the user's modification information and infer suggestions based on these user modification information to guide the construction of the subsequent knowledge architecture diagram, making the knowledge architecture diagram generated in the subsequent rounds more accurate and comprehensive than the knowledge architecture diagram generated in the current round.

[0058] Step S4: Based on the large model, fuse the entity types added or deleted in the database to generate the guidance suggestion text for the current round and update the iteration round 。

[0059] Specifically, the process of fusing the entity types added or deleted in the database based on the large model to generate the guidance suggestion text for the current round includes: The system reads the entity types added or deleted by the user from the database and uses the large model to fuse the added or deleted entity types, and then generates suggestions to guide the construction of the knowledge architecture diagram in the subsequent several rounds. For example, if the user adds entity types such as teacher and doctor and deletes entity types such as company and school, the system will analyze the added entity types, fuse the entity type of occupation, and then generate a suggestion that "the user is interested in entity types of this occupation category". At the same time, the system will analyze the deleted entity types, fuse the entity type of organization, and then generate a suggestion that "the user is not interested in entity types of this organization category". At the same time, update the current iteration round Specifically, the process is as follows: At the end of each iteration, the current iteration round is incremented by 1, that is, it satisfies the relationship: 。

[0060] Step S5: When the iteration round When, based on the guidance suggestion text of the th round, obtain the th piece of text block from the The entity of the round and the semantic relationship of the round, generate the relationship instance triples of the round.

[0061] Specifically, the process of generating the relationship instance triples of the round includes: The system filters entities from the suggestions generated by the round according to the user's needs and text blocks to obtain the two entities of the round, and then uses a large model to judge the semantic relationship between these two entities to obtain the semantic relationship of the round, and based on the two entities of the round and the semantic relationship of the round, obtain the relationship instance triples of the round.

[0062] Step S6: Based on the large model, respectively perform entity classification and semantic classification on the entities of the round and the semantic relationship of the round to obtain the updated entity types and relationship types, and construct the updated relationship type triples.

[0063] Specifically, the specific process of obtaining the updated entity types and relationship types includes: The system extracts the existing entity types and relationship types from the knowledge architecture diagram generated by the round, and then uses the large model to classify the entities of the round and the semantic relationship of the round. First, classify them into the existing entity types and relationship types, and filter out the entities and relationships that are not classified into the existing entity types and relationship types (for example: there is an entity type of "teacher" in the existing entity types. When obtaining entities such as "Teacher Zhang" and "Teacher Yang", these entities will be preferentially classified into the entity type of "teacher". Otherwise, in this round, "Teacher Zhang" and "Teacher Yang" may be classified into the entity type of "teacher" again. Since the meanings of "teacher" and "teacher" are essentially the same, redundant entity types will appear). Then, classify these entities and relationships respectively to obtain the updated entity types and relationship types.

[0064] Step S7: Merge the relationship type triples of the round with the updated relationship type triples to obtain the preliminary knowledge architecture diagram of the round, and return to Step S3 until the current iteration round reaches the preset maximum iteration round and stops iterating to generate the final knowledge architecture diagram.

[0065] Specifically, the preset maximum number of iterations of the present invention is , if the current iteration number is less than , steps S3 to S7 will be repeated until the current iteration number reaches . This indicates that the construction of the knowledge architecture diagram has reached the last round. At this time, the knowledge architecture diagram generated in the th round is the final knowledge architecture diagram.

[0066] Please refer to Figure 2 . Figure 2 FIG. Figure 2 is a schematic diagram of the knowledge architecture provided by the first embodiment of the present application, which is composed of multiple relationship type triples. Under the guidance of the pattern (operating system, compliance, protocol), the large model can extract relationship type triples such as (Windows operating system, compliance, SMB protocol), (Linux operating system, compliance, NFS protocol) from relevant text blocks.

[0067] The beneficial effect of an intelligent knowledge architecture diagram construction method based on a large model provided by the present application is that, firstly, the construction of the knowledge architecture diagram in the present application is divided into two processes, namely the current round construction and the th round construction. The construction of each round includes extraction and classification, and manual review and modification. The construction of the th round is carried out under the guidance of the knowledge architecture diagram and suggestions generated in the previous round. In this way, the large model can better understand the user's needs during the construction process, so as to construct a comprehensive and accurate knowledge architecture diagram with less manual participation, greatly reducing the labor cost and time cost. Secondly, when the knowledge architecture diagram is generated in each round, the user can make a small number of modifications to the edges and nodes of part of the knowledge architecture diagram to infer the user's demand tendency. Modifying the edge means modifying the relationship type, and modifying the node means modifying the entity type. Through modification, the accuracy and comprehensiveness of the knowledge architecture diagram during the construction process can be ensured.

[0068] Embodiment 2

[0069] Please refer to Figure 3 . FIG. Figure 3 shows a structural diagram of an intelligent knowledge architecture diagram construction system based on a large model provided by the second embodiment of the present application, including:

[0070] A text chunking module 100, configured to evenly chunk the user input text to obtain a plurality of sequentially connected text blocks, and obtain a plurality of entities from the first text block among the plurality of sequentially connected text blocks;

[0071] A preliminary knowledge architecture diagram construction module 200, configured to obtain semantic relationships between a plurality of entities and construct a preliminary knowledge architecture diagram;

[0072] The manual review and modification module 300 is used to add or delete nodes and edges of the preliminary knowledge architecture diagram to obtain the knowledge architecture diagram;

[0073] The storage module 400 is used to store the added or deleted entity types in the database;

[0074] The suggestion generation module 500 is used to fuse the added or deleted entity types stored in the database according to the large model to generate a guidance suggestion text;

[0075] The knowledge architecture diagram update module 600 is used to iteratively update the knowledge architecture diagram according to the guidance suggestion text to generate the final knowledge architecture diagram.

[0076] The beneficial effect of an intelligent knowledge architecture diagram construction system based on a large model provided by this application is that, first, this application uses the text chunking module 100 to evenly chunk the input text, which can ensure the coherence of the semantics of each text block. Second, the manual review and modification module 300 is used to add or delete nodes and edges of the preliminary knowledge architecture diagram to obtain the knowledge architecture diagram, which can infer the user's demand tendency by using a small amount of modification of part of the knowledge architecture diagram by humans, ensuring the accuracy and comprehensiveness of the knowledge architecture diagram during the construction process. Finally, the suggestion generation module 500 is used to fuse the added or deleted entity types in the database to generate a guidance suggestion text, which can further enable the large model to better understand the user's needs during the construction process, so as to construct a comprehensive and accurate knowledge architecture diagram with a small amount of human participation.

[0077] An intelligent knowledge architecture diagram construction system based on a large model in an embodiment of this application can be a device, or a component, an integrated circuit, or a chip in a terminal. The device can be a mobile electronic device or a non-mobile electronic device. Exemplarily, the mobile electronic device can be a mobile phone, a tablet computer, a laptop computer, a handheld computer, a vehicle-mounted electronic device, a wearable device, an ultra-mobile personal computer (UMPC), a netbook, or a personal digital assistant (PDA), etc., and the non-mobile electronic device can be a server, a Network Attached Storage (NAS), a personal computer (PC), a television (TV), a teller machine, or a self-service machine, etc. The embodiments of this application do not make specific limitations.

[0078] An intelligent knowledge architecture diagram construction system based on a large model in an embodiment of the present application can be a device with an operating system. The operating system can be the Android operating system, the iOS operating system, or other possible operating systems, which are not specifically limited in the embodiments of the present application.

[0079] An intelligent knowledge architecture diagram construction system based on a large model provided in an embodiment of the present application can implement Figures 1 to 2 each process implemented in a method embodiment of an intelligent knowledge architecture diagram construction method based on a large model. To avoid repetition, it will not be elaborated here.

[0080] Optionally, an embodiment of the present application further provides an electronic device, including a processor, a memory, and a program or instruction stored on the memory and executable on the processor. When the program or instruction is executed by the processor, it implements each process of the above method embodiment of an intelligent knowledge architecture diagram construction method based on a large model and can achieve the same technical effect. To avoid repetition, it will not be elaborated here.

[0081] An embodiment of the present application further provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, it implements each process of the above method embodiment of an intelligent knowledge architecture diagram construction method based on a large model and can achieve the same technical effect. To avoid repetition, it will not be elaborated here.

[0082] Wherein, the processor is the processor in the electronic device described in the above embodiment. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs, etc.

[0083] It should be noted that in this text, the term "comprise", "include" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or device comprising that element. In addition, it should be pointed out that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, but may also include performing functions in a substantially simultaneous manner or in the reverse order according to the functions involved. For example, the described methods may be performed in an order different from that described, and various steps may be added, omitted, or combined. Additionally, the features described with reference to certain examples may be combined in other examples.

[0084] From the description of the above embodiments, those skilled in the art can clearly understand that the above-described example methods can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions for causing a terminal (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of the present application.

[0085] The embodiments of the present application have been described above in conjunction with the accompanying drawings. However, the present application is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Those of ordinary skill in the art, under the inspiration of the present application and without departing from the spirit and scope protected by the claims of the present application, can also make many forms, all of which fall within the protection scope of the present application.

Claims

1. An intelligent knowledge architecture diagram construction method based on a large model, characterized in that, Including: S1. Obtain multiple entities from the first text block among multiple sequentially connected text blocks; S2. Obtain the semantic relationships among the multiple entities and construct a preliminary knowledge architecture graph. The specific steps for constructing the preliminary knowledge architecture graph include: Construct relationship instance triples based on the semantic relationships and the entities; Perform entity classification and semantic classification on the entities and semantic relationships respectively to obtain entity types and relationship types; Based on a large model, obtain common features from the text blocks where multiple entities corresponding to each entity type are located, and obtain entity types with attributes; Construct multiple relationship type triples based on the entity types with attributes, the relationship types, and the relationship instance triples, and obtain a preliminary knowledge architecture graph based on the relationship type triples. Among them, the process of constructing multiple relationship type triples based on the entity types with attributes, the relationship types, and the relationship instance triples includes: )； Among them, respectively represent two different entity types, represents the entity type 's attribute list, represents the entity type 's attribute list, represents the entity type and the entity type 's semantic relationship, represents the number of attributes corresponding to the entity type ; S3. Add or delete nodes and edges of the preliminary knowledge architecture graph to obtain a knowledge architecture graph, and store the added or deleted entity types in a database; S4. Integrate the entity types added or deleted in the database based on the large model, generate the guidance recommendation text for the current round, and update the iteration round ; S5. When the iteration round is reached, based on the guidance suggestion text of the -th round, obtain the entities of the -th block text block and the semantic relationships of the -th round, and generate the relationship instance triples of the -th round; ​ S6. Based on the large model, entity classification and semantic classification are respectively carried out on the entities in the round and the semantic relationships in the round to obtain the updated entity types and relationship types, and construct the updated relationship type triples; S7. Merge the relationship type triples of the round with the updated relationship type triples to obtain the preliminary knowledge architecture diagram of the round, and return to step S3 until the current iteration round reaches the preset maximum iteration round, then stop the iteration and generate the final knowledge architecture diagram.

2. The method for constructing an intelligent knowledge architecture diagram based on a large model according to claim 1, wherein The process of obtaining the multiple sequentially connected text blocks includes: According to the number of iteration rounds input by the user, the large model evenly divides the input text into multiple sequentially connected text blocks. Starting from the second text block, each text block will retain one-tenth of the end of the previous text block as the connection part at the beginning of the current text block.

3. A method for constructing an intelligent knowledge architecture diagram based on a large model according to claim 1, characterized in that, The relationship types include: genus-species relationship, composition relationship, and location relationship.

4. A method for constructing an intelligent knowledge architecture diagram based on a large model according to claim 1, characterized in that, The preliminary knowledge architecture graph consists of multiple edges and multiple nodes. Among them, the nodes represent entity types, and the edges represent relationship types.

5. A method for constructing an intelligent knowledge architecture diagram based on a large model according to claim 1, characterized in that, The common features refer to those that appear in the text paragraphs where multiple entities corresponding to the same entity type are located and whose occurrence times are greater than or equal to a preset number according to the analysis of the large model, and are used to describe the static and dynamic features of specific instances in the same entity type.

6. An intelligent knowledge architecture diagram construction system based on a large model, characterized in that, The system includes: A text chunking module for evenly chunking the user input text to obtain multiple sequentially connected text blocks, and obtaining multiple entities from the first text block among the multiple sequentially connected text blocks; A preliminary knowledge architecture graph construction module for obtaining the semantic relationships among the multiple entities and constructing a preliminary knowledge architecture graph. The specific steps for constructing the preliminary knowledge architecture graph include: constructing relationship instance triples based on the semantic relationships and the entities; performing entity classification and semantic classification on the entities and semantic relationships respectively to obtain entity types and relationship types; obtaining common features from the text blocks where multiple entities corresponding to each entity type are located based on a large model to obtain entity types with attributes; constructing multiple relationship type triples based on the entity types with attributes, the relationship types, and the relationship instance triples, and obtaining a preliminary knowledge architecture graph based on the relationship type triples. Among them, the process of constructing multiple relationship type triples based on the entity types with attributes, the relationship types, and the relationship instance triples includes: )； Among them, respectively represent two different entity types, represents the entity type 's list of attributes, represents the entity type 's list of attributes, represents the entity type and the entity type 's semantic relationship, represents the number of attributes corresponding to the entity type ; An artificial review and modification module for adding or deleting nodes and edges of the preliminary knowledge architecture graph to obtain a knowledge architecture graph; A storage module for storing the added and deleted entity types in a database; A suggestion generation module for fusing the added and deleted entity types stored in the database according to a large model to generate a guidance suggestion text; A knowledge architecture diagram update module for iteratively updating the knowledge architecture diagram according to the guidance suggestion text to generate a final knowledge architecture diagram.

7. An electronic device, characterized in that, It includes a processor, a memory, and programs or instructions stored on the memory and executable on the processor. When the programs or instructions are executed by the processor, the steps of an intelligent knowledge architecture diagram construction method based on a large model as described in any one of claims 1-5 are implemented.

8. A readable storage medium, characterized in that, Programs or instructions are stored on the readable storage medium. When the programs or instructions are executed by the processor, the steps of an intelligent knowledge architecture diagram construction method based on a large model as described in any one of claims 1-5 are implemented.

Images