Construction method of blood disease knowledge graph database

Abstract

A construction method of a blood disease knowledge graph database is characterized by comprising the following steps: determining a target user and a purpose of the database, and defining a concept and a relationship of a blood disease field; blood disease field data are collected and preprocessed; using a natural language processing technology to extract entities, relationships and attributes in the blood disease data text, and establishing relationships among the entities; establishing a blood disease ontology model and designing a database; storing the processed data in a database, designing query and retrieval interfaces, and supporting a user to query and browse a knowledge graph; and verifying the accuracy and integrity of the knowledge graph. According to the method, through the accurate entities and the relation model, the relation among the entities such as diseases, symptoms and treatment methods in the knowledge graph is clearer. The method is helpful for users to better understand related knowledge of blood diseases, and supports clinical decision and research. An effective updating mechanism is constructed, and it is ensured that information in the knowledge graph is synchronized with the latest medical research progress and clinical practice.

Description

Technical Field

[0001] This invention belongs to the field of knowledge graph technology, and in particular relates to a method for constructing a knowledge graph database for hematological diseases. Background Technology

[0002] With the development of computer science and technology, especially the rapid development of cloud computing, the Internet of Things, and big data in recent years, the level of artificial intelligence has been greatly improved. The medical field has become one of the most active and widespread areas for the application of artificial intelligence. The core and key of intelligent healthcare is intelligent diagnosis and treatment, which means making the computer a brain with medical knowledge, thereby providing auxiliary decision-making for doctors' diagnosis and treatment. Knowledge graphs are an important technology in the field of artificial intelligence. They are the technical foundation for building a computer's medical knowledge brain and one of the key technologies for realizing intelligent healthcare. With the rapid development of artificial intelligence technology and the increasing maturity of technologies such as natural language processing, image recognition, deep learning, and machine learning, a good foundation has been laid for the development of hematological disease knowledge graphs. In order to solve the contradiction between the insufficient supply of high-quality medical resources and the continuously increasing demand for medical services in my country, the development of hematological disease knowledge graphs is imperative.

[0003] The field of hematology involves complex entities such as molecules, cells, and disease subtypes, as well as intricate relationships between them. Building accurate entity and relationship models may require a deep understanding of domain knowledge, involving multi-layered ontology modeling. In the medical field, the use of terminology and standardization can vary. A lack of standardization can lead to inconsistencies in entities and relationships within the knowledge graph. Appropriate standards and specifications are needed to normalize the data. Medical data often involves patient privacy; therefore, sensitive information must be handled and compliance with relevant privacy and security regulations must be ensured when constructing the knowledge graph. User needs for knowledge graphs may constantly evolve, requiring flexible query and retrieval interfaces to meet the needs of diverse users. The construction of a knowledge graph database may face technical limitations, including storage capacity and query performance issues. Summary of the Invention

[0004] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.

[0005] In view of the problems existing in the above or existing methods for constructing hematological disease knowledge graph databases, this invention is proposed.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] This invention provides a method for constructing a hematological disease knowledge graph database, characterized by comprising: determining the target users and uses of the database, and defining concepts and relationships in the hematological disease domain; collecting and preprocessing data in the hematological disease domain; using natural language processing technology to extract entities, relationships, and attributes from the hematological disease data text and establish relationships between entities; establishing a hematological disease ontology model and designing the database; storing the processed data in the database, designing query and retrieval interfaces to support users in querying and browsing the knowledge graph; and verifying the accuracy and completeness of the knowledge graph.

[0008] As a preferred embodiment of the method for constructing the hematological disease knowledge graph database of the present invention, the definition of concepts and relationships in the field of hematological diseases includes: describing the causal relationship between diseases and other diseases, genetic factors, and environmental factors; describing the relationship between treatment methods and specific diseases; and describing the association between hematological diseases and specific genes.

[0009] As a preferred embodiment of the method for constructing the hematological disease knowledge graph database described in this invention, the step of collecting and preprocessing data in the field of hematology includes: using MEDLINE and academic journals to retrieve literature and review articles related to hematology; collecting medical textbooks and professional reference books related to hematology and extracting the knowledge and information therein; checking the dataset, handling duplicates, missing values ​​and error messages; performing natural language processing on the literature text to extract entity relationships and attributes; using word segmentation and named entity recognition technologies to unify the naming and representation of entities and relationships; adopting a standard ontology language to represent concepts and relationships; and extracting relationships from the text through natural language processing technology.

[0010] As a preferred embodiment of the method for constructing the hematological disease knowledge graph database described in this invention, the step of using natural language processing (NLP) technology to extract entities, relationships, and attributes from hematological disease data text and establish relationships between entities includes: using NLP tools for entity recognition to identify disease names, symptoms, treatment methods, and drug entities in the text; based on the results of entity recognition, using NLP technology to extract relationships; formulating relationship extraction rules, such as determining relationships between entities by detecting verb keywords like "treatment" and "cause"; extracting attribute information of entities from the text, such as drug side effects and disease symptoms; and using NLP technology to perform text mining to identify information related to entities.

[0011] As a preferred embodiment of the method for constructing the hematological disease knowledge graph database described in this invention, the step of establishing a hematological disease ontology model and designing the database includes: defining the hierarchical structure of the ontology, including high-level concepts and sub-concepts; defining entities in each hierarchical structure, including disease entities, symptom entities, and treatment entity entities; assigning a unique identifier to each entity; defining the relationships between different entities; defining attributes for each entity; selecting OWL language for formal representation of the ontology model; and, based on the characteristics of the data and query requirements, selecting Neo4j database for relationships between entities; using MySQL to represent structured data; if a relational database is selected, designing database tables to store entities and relationships; if a graph database is selected, designing nodes and edges to represent entities and relationships, with nodes representing entities and edges representing relationships, and nodes and edges containing attribute information.

[0012] As a preferred embodiment of the method for constructing the hematological disease knowledge graph database of the present invention, the step of storing the processed data into the database includes: selecting the graph database Neo4j, creating corresponding tables or nodes and relationships in the database according to the ontology model and database design, and importing the data processed by natural language into the database.

[0013] As a preferred embodiment of the method for constructing the hematological disease knowledge graph database described in this invention, the method for verifying the accuracy and completeness of the knowledge graph includes: checking whether there are missing entities, relationships, or attributes in the knowledge graph; supplementing missing entities and relationships based on new medical knowledge under the guidance of domain experts; cross-validating interrelated entities in the knowledge graph to ensure that the relationships and attributes between them are consistent; and considering the time information in the knowledge graph to ensure that the attributes of entities such as diseases and treatment methods are consistent over time.

[0014] The beneficial effects of this invention are as follows: By employing accurate entity and relation models, the relationships between entities such as diseases, symptoms, and treatments in the knowledge graph become clearer. This helps users better understand knowledge related to hematological diseases, supporting clinical decision-making and research. An effective update mechanism is constructed to ensure that the information in the knowledge graph is synchronized with the latest medical research progress and clinical practice. This helps maintain the timeliness of the knowledge graph, providing users with the latest medical knowledge. By adopting multi-level ontology modeling, the complexity of the hematological field is better represented, including the molecular, cellular, and clinical levels. This helps to gain a deeper understanding of the multi-level structure of hematological diseases, providing more refined information for research and treatment. A user-friendly query interface is provided, supporting keyword queries, complex queries, and graphical displays. This helps users easily obtain the information they need, improving the user experience. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0016] Figure 1 A flowchart illustrating a method for constructing a hematological disease knowledge graph database, as provided in an embodiment of the present invention.

[0017] Figure 2 This is a flowchart illustrating the steps of an embodiment 1 of the method for constructing a blood disease knowledge graph database according to the present invention. Detailed Implementation

[0018] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0019] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0020] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0021] Example 1

[0022] Reference Figures 1-2 This is one embodiment of the present invention, which provides a method for constructing a hematological disease knowledge graph database, including:

[0023] S1: Determine the target users and uses of the database, and define concepts and relationships in the field of hematology.

[0024] Preferably, it describes the causal relationship between the disease and other diseases, genetic factors, and environmental factors; the relationship between treatment methods and specific diseases; and the association between blood diseases and specific genes.

[0025] S2: Collect and preprocess data in the field of hematology.

[0026] The preferred approach is to use MEDLINE and academic journals to retrieve literature and review articles related to hematological diseases, collect medical textbooks and professional reference books related to hematological diseases, and extract the knowledge and information from them; examine the dataset, handle duplicates, missing values ​​and error messages, perform natural language processing on the literature text, extract entity relationships and attributes, use word segmentation and named entity recognition technology, unify the naming and representation of entities and relationships, adopt standard ontology language to represent concepts and relationships, and extract relationships from the text through natural language processing technology.

[0027] S3: Use natural language processing technology to extract entities, relationships, and attributes from blood disease data text and establish relationships between entities.

[0028] Preferably, NLP tools are used for entity recognition to identify disease names, symptoms, treatment methods, and drug entities in the text. Based on the results of entity recognition, NLP technology is used for relation extraction. Relationship extraction rules are formulated, such as determining the relationship between entities by detecting verb keywords such as "treatment" and "cause". Attribute information of entities is extracted from the text, such as the side effects of drugs and the symptoms of diseases. NLP technology is used for text mining to identify information related to the entities.

[0029] Furthermore, identify key verbs that represent the relationships between entities, such as "treat," "cause," and "influence." Use these key verbs to formulate relationship extraction rules. For example, the rules could be formulated as follows:

[0030] If a sentence contains "treat," a treatment relationship exists; if a sentence contains "cause," a cause relationship exists; if a sentence contains "influence," an influence relationship exists. Using parsing techniques, such as dependency parsing, the subject and object of sentences containing key verbs are extracted as relationships between entities. For example:

[0031] For the sentence "Drug A treats leukemia," the extraction relation is: there is a treatment relationship between "Drug A" (subject) and "leukemia" (object). For the sentence "Gene mutation causes anemia," the extraction relation is: there is a cause relationship between "gene mutation" (subject) and "anemia" (object).

[0032] S4: Establish a hematological ontology model and design a database.

[0033] Preferably, a hierarchical structure of the ontology is defined, including high-level concepts and sub-concepts. Entities are defined in each hierarchy, including disease entities, symptom entities, and treatment entity entities. A unique identifier is assigned to each entity, relationships between different entities are defined, and attributes are defined for each entity. OWL is selected for formal representation of the ontology model. Based on the characteristics of the data and query requirements, Neo4j is selected as the database for relationships between entities. MySQL is used to represent structured data. If a relational database is selected, database tables are designed to store entities and relationships. If a graph database is selected, nodes and edges are designed to represent entities and relationships, with nodes representing entities and edges representing relationships. Nodes and edges contain attribute information.

[0034] S5: Store the processed data in a database, design query and retrieval interfaces, and support users in querying and browsing the knowledge graph.

[0035] Preferably, the graph database Neo4j is selected. Based on the ontology model and database design, corresponding tables or nodes and relationships are created in the database, and the data processed by natural language is imported into the database.

[0036] S6: Verify the accuracy and completeness of the knowledge graph.

[0037] Preferably, the knowledge graph is checked for missing entities, relationships, or attributes. Under the guidance of domain experts, missing entities and relationships are supplemented based on new medical knowledge. Interrelated entities in the knowledge graph are cross-validated to ensure that the relationships and attributes between them are consistent. The temporal information in the knowledge graph is considered to ensure that the attributes of entities such as diseases and treatment methods are consistent over time.

[0038] Example 2

[0039] Referring to Table 1, another embodiment of the present invention is provided. Based on the above method, a scientific comparative experiment is provided to verify its beneficial effects.

[0040] Table 1

[0041]

[0042] This invention clarifies the relationships between entities such as diseases, symptoms, and treatments in a knowledge graph through accurate entity and relation models. This helps users better understand knowledge related to hematological diseases, supporting clinical decision-making and research. An effective update mechanism is also constructed to ensure that the information in the knowledge graph is synchronized with the latest medical research progress and clinical practice.

[0043] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for constructing a knowledge graph database for hematological diseases, characterized in that, include: Identify the target users and uses of the database, and define the concepts and relationships in the field of hematology; Collect and preprocess data in the field of hematology; Natural language processing technology is used to extract entities, relationships, and attributes from hematological disease data text and establish relationships between entities. Establish a hematological ontology model and design a database; The processed data is stored in a database, and query and retrieval interfaces are designed to support users in querying and browsing the knowledge graph. Verify the accuracy and completeness of the knowledge graph.

2. The method for constructing a hematological disease knowledge graph database as described in claim 1, characterized in that, The definitions of concepts and relationships in the field of hematology include: Describe the causal relationships between diseases and other diseases, genetic factors, and environmental factors; describe the relationship between treatment methods and specific diseases; and describe the association between blood diseases and specific genes.

3. The method for constructing a hematological disease knowledge graph database as described in claim 1, characterized in that, The collection and preprocessing of data in the field of hematology includes: Using MEDLINE and academic journals, we retrieved literature and review articles related to hematological diseases, collected medical textbooks and professional reference books related to hematological diseases, and extracted the knowledge and information from them; we examined the dataset, handled duplicates, missing values ​​and error messages, performed natural language processing on the literature text, extracted entity relationships and attributes, used word segmentation and named entity recognition technologies, standardized the naming and representation of entities and relationships, adopted standard ontology language to represent concepts and relationships, and extracted relationships from the text through natural language processing techniques.

4. The method for constructing a hematological disease knowledge graph database as described in claim 1, characterized in that, The process of using natural language processing technology to extract entities, relationships, and attributes from blood disease data text and establishing relationships between entities includes: Entity recognition is performed using NLP tools to identify disease names, symptoms, treatment methods, and drug entities in the text. Based on the results of entity recognition, NLP techniques are used for relation extraction. Relationship extraction rules are formulated, such as determining the relationships between entities by detecting verb keywords like "treatment" and "cause". Attribute information of entities is extracted from the text, such as drug side effects and disease symptoms. NLP techniques are then used for text mining to identify information related to the entities.

5. The method for constructing a hematological disease knowledge graph database as described in claim 1, characterized in that, The establishment of the hematological ontology model and the design of the database include: Define the hierarchical structure of the ontology, including high-level concepts and sub-concepts. Define entities in each hierarchy, including disease entities, symptom entities, and treatment entity entities. Assign a unique identifier to each entity, define the relationships between different entities, and define attributes for each entity. Choose OWL language for formal representation of the ontology model. Based on the characteristics of the data and query requirements, choose Neo4j database for entity relationships. Use MySQL to represent structured data. If a relational database is chosen, design database tables to store entities and relationships; if a graph database is chosen, design nodes and edges to represent entities and relationships. Nodes represent entities, edges represent relationships, and nodes and edges contain attribute information.

6. The method for constructing a hematological disease knowledge graph database as described in claim 1, characterized in that, The process of storing the processed data in the database includes: Choose the graph database Neo4j, and based on the ontology model and database design, create corresponding tables or nodes and relationships in the database, and import the data that has undergone natural language processing into the database.

7. The method for constructing a hematological disease knowledge graph database as described in claim 1, characterized in that, The verification of the accuracy and completeness of the knowledge graph includes: Check the knowledge graph for missing entities, relationships, or attributes. Under the guidance of domain experts, supplement the missing entities and relationships based on new medical knowledge. Cross-validate the interconnected entities in the knowledge graph to ensure that the relationships and attributes between them are consistent. Consider the time information in the knowledge graph to ensure that the attributes of entities such as diseases and treatment methods are consistent over time.

Images