Medicated food knowledge graph construction method and system and medium

Abstract

The invention relates to a medicated diet knowledge map construction method and system and a medium, and the method comprises the steps: M1, a medicated diet knowledge map ontology architecture: a multi-layer traditional Chinese medicine attribute-relation-diet attribute architecture, a bottom layer comprises food materials and medicated diets, a second layer comprises primary attributes of the food materials and the medicated diets, and a third layer comprises advanced attributes of the food materials and the medicated diets; m2, data set: data set of traditional Chinese medicine and dietary elements, including food material, medicated diet and standard data set; m3, relation functions including an information transmission model, an information fusion model, a medicated diet recommendation model, a dietary nutrition model, an efficacy model and a health model; and M4, knowledge association. According to the traditional Chinese medicine attribute concept entity-relationship-diet attribute, the accurate medicated food knowledge graph is obtained, and the medicated food knowledge graph is topologically updated in combination with the medicated food recommendation model, so that the comprehensiveness and accuracy of the medicated food knowledge graph are further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of constructing a medicinal diet knowledge graph, and in particular to a method, system and medium for constructing a medicinal diet knowledge graph. Background Art

[0002] With the rapid development of information technology, especially the popularity of cloud computing, the Internet of Things and social media, the amount of global data is showing an explosive growth trend. In these fields, people are generating and sharing a large amount of information every minute, including the knowledge data of medicinal diet. Therefore, in the construction of the medicinal diet knowledge graph, how to optimize and test the medicinal diet knowledge and obtain an accurate medicinal diet knowledge graph has become a problem that we need to solve urgently. Summary of the invention

[0003] In view of the above problems, the present invention provides a method, system and medium for constructing a medicinal diet knowledge graph, which not only obtains an accurate medicinal diet knowledge graph based on the TCM attribute concept entity-relationship-diet attribute, but also combines the medicinal diet recommendation model to topology and update the medicinal diet knowledge graph, thereby further improving the comprehensiveness and accuracy of the medicinal diet knowledge graph.

[0004] In order to achieve the above-mentioned purpose and other related purposes, the technical solution provided by the present invention is as follows:

[0005] A method for constructing a medicinal diet knowledge graph, the method comprising:

[0006] M1. Medicinal diet knowledge graph ontology architecture: a multi-level TCM attribute-relationship-diet attribute architecture, with the bottom layer including ingredients and medicinal diet, the second layer including primary attributes of ingredients and medicinal diet, and the third layer including advanced attributes of ingredients and medicinal diet;

[0007] M2. Dataset: Dataset of traditional Chinese medicine and dietary elements, including food ingredients, medicinal diet, and standard data sets;

[0008] M3. Relationship function: including information transmission model, information fusion model, including medicinal diet recommendation model, dietary nutrition model, efficacy model, and health model;

[0009] M4. Knowledge association: Through relational functions, the traditional Chinese medicine properties and dietary properties of medicinal foods are associated to form a relational network, realizing the flow, transmission and integration of various information in the data set.

[0010] M5. Reasoning about medicinal diet knowledge: Based on individual basic information, health status, living habits and health management goals, calculate and derive new medicinal diet knowledge, including medicinal diet strategies and health mechanisms.

[0011] Furthermore, the method further comprises:

[0012] M6. Map the data in the data set to the ontology architecture in order from bottom to top to obtain the medicinal diet knowledge graph ontology architecture with data.

[0013] Furthermore, in step M1, establishing the medicinal diet knowledge graph ontology architecture includes:

[0014] M11. The ontology architecture is composed of a number of nodes, each of which represents a concept of medicinal diet knowledge. Further, in step M6, the establishment of the medicinal diet knowledge graph ontology architecture includes:

[0015] M61. Identify the concepts of the top-level medicinal diet knowledge set and divide it into two categories: ingredients and medicinal diet;

[0016] M62. Determine the concepts of the second layer, and then gradually refine the next layer of concepts based on the TCM meridian theory, the four properties and five flavors theory, the rise and fall theory, TCM constitution, TCM diet efficacy theory, and food nutrition theory, and establish the ontology architecture of the medicinal diet knowledge graph.

[0017] Furthermore, the dataset is constructed by collecting information related to medicinal diet, ingredients, and standards through web crawler technology and professional dataset docking methods, establishing an ingredient dataset based on the core information of nutritional components and functional components, a medicinal diet dataset based on the core information of compatibility, nature and taste, constitution, and dosage form, and a dietary standard set based on the core information of nutritional standards, dietary standards, and efficacy standards.

[0018] Further, in step M3, constructing the medicinal diet recommendation model according to the effective components and nutrients in the medicinal diet includes:

[0019] M31. Based on the medicinal diet knowledge graph ontology architecture filled with data, obtain the data information of the number of targets and functional components of the intersection of medicinal diet and disease, the weight data information of ingredients, and the data information of nutrients contained in the ingredients;

[0020] M32. Based on the data information of the number of targets and efficacy components of the intersection of medicinal diet and disease, the weight data information of ingredients, and the data information of nutrients contained in ingredients, a medicinal diet recommendation model is established, including the efficacy objective function G and the nutritional constraints.

[0021] G=c1w1+c2w2+…+c i w i …+c n w n ,

[0022] The nutritional constraints are:

[0023]

[0024] Among them, G is the content of the effective ingredient, and the maximum value is taken according to individual needs;

[0025] i is the i-th ingredient, n is the total number of ingredients, c i is the content of the main active ingredient in the i-th food ingredient, w i is the weight of the i-th ingredient in the medicinal diet, j is the j-th nutrient, there are m nutrients in total, m is the total number of nutrient types, a ji is the content of the jth nutrient in the ith ingredient, b 1j is the lower limit of recommended intake of the jth nutrient, b 2j is the upper limit of recommended intake of nutrient j;

[0026] M33. Based on the objective function G of medicinal diet recommendation, determine the medicinal diet formula w with the maximum content of effective ingredients and basic nutrients that meet the daily recommended intake i , complete the recommendation of medicinal diet.

[0027] Furthermore, the data information of nutrients contained in the food includes data information of protein, data information of carbohydrates, data information of fat, data information of calories and data information of vitamins.

[0028] Furthermore, the relationship function extracts the data in the data set as concept-relationship-concept triples and appends them to the data set to describe and illustrate the relationship between entities.

[0029] Furthermore, the method further comprises:

[0030] M7. Based on the medicinal diet knowledge graph ontology architecture, medicinal diet information is extracted through knowledge reasoning on the original data, and the data is combined with knowledge to obtain dietary analysis, model application, and information query functions.

[0031] In order to achieve the above-mentioned purpose and other related purposes, the present invention also provides a system for constructing a medicinal diet knowledge graph, including a computer device, which is programmed or configured to execute the steps of any one of the methods for constructing a medicinal diet knowledge graph.

[0032] In order to achieve the above-mentioned purpose and other related purposes, the present invention also provides a computer-readable storage medium, which stores a computer program and data information programmed or configured to execute any one of the methods for constructing a medicinal diet knowledge graph.

[0033] The present invention has the following positive effects:

[0034] 1. The present invention optimizes the number of intersection targets between medicinal diet and diseases and the intake of basic nutrients by constructing a medicinal diet recommendation model, and determines that the number of intersection targets between medicinal diet and diseases is the largest and the basic nutrients meet the daily recommended intake. It can not only optimize the data of each component in the medicinal diet to obtain an accurate medicinal diet knowledge map, but also in this process, it can be adaptively optimized according to different medicinal diets, and has strong robustness.

[0035] 2. The present invention extracts the required entities from the data set by manually annotating the data, and extracts the entities through relationships to obtain concept-relationship-concept triple relationships, characterizes the relationships between the components of the medicinal diet, and constructs a graph. It can not only construct corresponding accurate graphs for different medicinal diets, but also topologically and update the medicinal diet knowledge graph in combination with the data set, thereby further improving the comprehensiveness and accuracy of the medicinal diet knowledge graph. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 Schematic diagram of the method flow of the present invention (I);

[0037] Figure 2 It is a schematic diagram of the medicinal diet knowledge graph ontology architecture of the present invention;

[0038] Figure 3 It is a schematic diagram of the relationship extraction result of the tangerine peel, poria and jujube porridge of the present invention;

[0039] Figure 4 It is a schematic diagram of the medicinal diet-porridge-knowledge map of the present invention;

[0040] Figure 5 It is a schematic diagram of the application of the medicinal diet knowledge graph of the present invention;

[0041] Figure 6 Schematic diagram of the medicinal diet knowledge graph ontology architecture of the present invention (II). DETAILED DESCRIPTION

[0042] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0043] Example 1: Figure 1 or Figure 2 As shown, a method for constructing a medicinal diet knowledge graph comprises:

[0044] M1. Medicinal diet knowledge graph ontology architecture: a multi-level TCM attribute-relationship-diet attribute architecture, with the bottom layer including ingredients and medicinal diet, the second layer including primary attributes of ingredients and medicinal diet, and the third layer including advanced attributes of ingredients and medicinal diet;

[0045] M2. Dataset: Dataset of traditional Chinese medicine and dietary elements, including food ingredients, medicinal diet, and standard data sets;

[0046] M3. Relationship function: including information transmission model, information fusion model, including medicinal diet recommendation model, dietary nutrition model, efficacy model, and health model;

[0047] M4. Knowledge association: Through relational functions, the traditional Chinese medicine properties and dietary properties of medicinal foods are associated to form a relational network, realizing the flow, transmission and integration of various information in the data set.

[0048] M5. Reasoning about medicinal diet knowledge: Based on individual basic information, health status, living habits and health management goals, calculate and derive new medicinal diet knowledge, including medicinal diet strategies and health mechanisms.

[0049] In this embodiment, the method further includes:

[0050] M6. Map the data in the data set to the ontology architecture in order from bottom to top to obtain the medicinal diet knowledge graph ontology architecture with data.

[0051] In this embodiment, in step M1, establishing the medicinal diet knowledge graph ontology architecture includes:

[0052] M11. The ontology architecture consists of several nodes, each of which represents a concept of medicinal diet knowledge.

[0053] In this embodiment, in step M6, establishing the medicinal diet knowledge graph ontology architecture includes:

[0054] M61. Identify the concepts of the top-level medicinal diet knowledge set and divide it into two categories: ingredients and medicinal diet;

[0055] M62. Determine the concepts of the second layer, and then gradually refine the next layer of concepts based on the TCM meridian theory, the four properties and five flavors theory, the rise and fall theory, TCM constitution, TCM diet efficacy theory, and food nutrition theory, and establish the ontology architecture of the medicinal diet knowledge graph.

[0056] In this embodiment, the data set is constructed by collecting information related to medicinal diet, ingredients, and standards through web crawler technology and professional data set docking methods, establishing an ingredient data set based on the core information of nutritional components and functional components, a medicinal diet data set based on the core information of compatibility, nature and taste, constitution, and dosage form, and a dietary standard set based on the core information of nutritional standards, dietary standards, and efficacy standards.

[0057] In this embodiment, in step M3, constructing the medicinal diet recommendation model according to the effective components and nutrients in the medicinal diet includes:

[0058] M31. Based on the medicinal diet knowledge graph ontology architecture filled with data, obtain the data information of the number of targets and functional components of the intersection of medicinal diet and disease, the weight data information of ingredients, and the data information of nutrients contained in the ingredients;

[0059] M32. Based on the data information of the number of targets and efficacy components of the intersection of medicinal diet and disease, the weight data information of ingredients, and the data information of nutrients contained in ingredients, a medicinal diet recommendation model is established, including the efficacy objective function G and the nutritional constraints.

[0060] G=c1w1+c2w2+…+c i w i …+c n w n ,

[0061] The nutritional constraints are:

[0062]

[0063] Among them, G is the content of the effective ingredient, and the maximum value is taken according to individual needs;

[0064] i is the i-th ingredient, n is the total number of ingredients, c i is the content of the main active ingredient in the i-th food ingredient, w i is the weight of the i-th ingredient in the medicinal diet, j is the j-th nutrient, there are m nutrients in total, m is the total number of nutrient types, a ji is the content of the jth nutrient in the ith ingredient, b 1j is the lower limit of recommended intake of the jth nutrient, b 2j is the upper limit of recommended intake of nutrient j;

[0065] M33. Based on the objective function G of medicinal diet recommendation, determine the medicinal diet formula w with the maximum content of effective ingredients and basic nutrients that meet the daily recommended intake i , complete the recommendation of medicinal diet.

[0066] In this embodiment, the data information of nutrients contained in the food includes data information of protein, data information of carbohydrates, data information of fat, data information of calories and data information of vitamins.

[0067] In this embodiment, the relationship function extracts the data in the data set as concept-relationship-concept triples and appends them to the data set to describe and illustrate the relationship between entities.

[0068] In this embodiment, the method further includes:

[0069] M7. Based on the medicinal diet knowledge graph ontology architecture, medicinal diet information is extracted through knowledge reasoning on the original data, and the data is combined with knowledge to obtain dietary analysis, model application, and information query functions.

[0070] In this embodiment, the present invention provides a system for constructing a medicinal diet knowledge graph, including a computer device, which is programmed or configured to execute the steps of any one of the methods for constructing a medicinal diet knowledge graph.

[0071] In this embodiment, the present invention provides a computer-readable storage medium, on which is stored a computer program programmed or configured to execute any one of the methods for constructing a medicinal diet knowledge graph.

[0072] Any reference to memory, storage, data set or other medium used in the embodiments provided in the present application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM) and memory bus dynamic RAM (RDRAM).

[0073] Example 2: Based on the method for constructing a medicinal diet knowledge graph in Example 1, the present invention is further illustrated and described below.

[0074] like Figure 1 or Figure 2 As shown, a method for constructing a medicinal diet knowledge graph comprises:

[0075] M1. Medicinal diet knowledge graph ontology architecture: a multi-level TCM attribute-relationship-diet attribute architecture, with the bottom layer including ingredients and medicinal diet, the second layer including primary attributes of ingredients and medicinal diet, and the third layer including advanced attributes of ingredients and medicinal diet;

[0076] M2. Dataset: Dataset of traditional Chinese medicine and dietary elements, including food ingredients, medicinal diet, and standard data sets;

[0077] M3. Relationship function: including information transmission model, information fusion model, including medicinal diet recommendation model, dietary nutrition model, efficacy model, and health model;

[0078] M4. Knowledge association: Through relational functions, the traditional Chinese medicine properties and dietary properties of medicinal foods are associated to form a relational network, realizing the flow, transmission and integration of various information in the data set.

[0079] M5. Reasoning about medicinal diet knowledge: Based on individual basic information, health status, living habits and health management goals, calculate and derive new medicinal diet knowledge, including medicinal diet strategies and health mechanisms.

[0080] A method for constructing a medicinal diet knowledge graph comprises the following steps: constructing a data set, wherein the collected data set includes two parts: a set of traditional Chinese medicine ingredients and a set of non-traditional Chinese medicine ingredients, covering information such as ingredient name, Latin name, toxicity, active ingredient, five flavors, meridians, function, origin, nutritional ingredients, drug target, corresponding gene, corresponding disease, energy, minerals, water, vitamins, macronutrients, data source, etc., in order to provide a reference for users to choose appropriate ingredients.

[0081] A method for constructing a medicinal diet knowledge graph comprises the following steps: data organization, characterized in that relationship extraction is to extract data from a data set to form concept-relationship-concept and concept-attribute-concept triplets and store them in the data set, and the remaining sentences are stored in text form to describe and illustrate the relationship between concepts. For example, the sentence "Chenpi, Poria and Jujube Porridge has the effects of removing dampness and resolving phlegm, strengthening the spleen and harmonizing the middle" can be extracted and organized into: concept (medicinal diet)-relationship (effect)-concept (disease treatment); further construct the triple relationship of concept (medicinal diet: Chenpi, Poria and Jujube Porridge)-relationship (effect)-concept (disease treatment: removing dampness and resolving phlegm, strengthening the spleen and harmonizing the middle), then you can understand that the effect of the medicinal diet "Chenpi, Poria and Jujube Porridge" is "removing dampness and resolving phlegm, strengthening the spleen and harmonizing the middle"; or you can get the medicinal diet "Chenpi, Poria and Jujube Porridge" with the effect of "removing dampness and resolving phlegm, strengthening the spleen and harmonizing the middle"; the concepts "Chenpi, Poria and Jujube Porridge", "Removing dampness and resolving phlegm, strengthening the spleen and harmonizing the middle", and the relationship is "effect", and the extracted result is "Chenpi, Poria and Jujube Porridge-Efficacy-Removing dampness and resolving phlegm, strengthening the spleen and harmonizing the middle".

[0082] The first-level structure of the ontology architecture is “medicinal diet-effects-disease treatment”, and the corresponding second-level structure includes “tangerine peel, poria cocos and jujube porridge-effects-removing dampness and phlegm, strengthening the spleen and harmonizing the stomach”.

[0083] For example, the manual annotation method is used to extract the relationship between "food-disease indicators" from the established knowledge set, and each relationship is formed into a triple. By manually establishing nodes and defining the association relationship between concepts, the data relationship extraction information is shown in Figure 3 The nodes are concepts such as "Chenpi Fuling Dazao Congee", "Gastritis", "Dates, Fuling, Rice", etc., and the edges include "Efficacy", "Formula", "Cause Disease", and "Relationship Function". The logical relationship between medicinal food, ingredients, and efficacy is clarified through association, and the data is displayed more intuitively.

[0084] A method for constructing a medicinal diet knowledge graph comprises the following steps: data association, based on a knowledge graph platform, according to the logical relationship between data, and associating data in a top-down manner. Figure 4 This is a display of the data association results of porridge in medicinal food. Different node colors represent different levels, and the lines between nodes represent the relationship or attributes between entities.

[0085] A method for constructing a medicinal diet knowledge graph includes the following steps: intelligent application, such as Figure 5 As shown, information such as process formula, suitable population, and medicinal diet type can be extracted from the original data stored in the knowledge graph. Through knowledge combination, knowledge reasoning and other steps, support can be provided for users to choose medicinal diets suitable for their own conditions. From the original data stored in the knowledge graph, information such as the nutritional components, functional components, key targets, and limit standards of medicinal diet porridge can be extracted. A medicinal diet recommendation model can be constructed to provide a basis for the subsequent implementation of the model application. From the original data stored in the knowledge graph, information such as the TCM properties of ingredients, the chemical components contained, and the composition of ingredients in medicinal diets can be extracted, which can assist users in understanding medicinal diet information and the theoretical basis of TCM.

[0086] In the initial construction process, it is necessary to first improve the ontology layer of the medicinal diet knowledge graph, establish the ontology architecture of the medicinal diet knowledge graph, and then extract data from the data set and fill it into the data layer using a top-down construction method. The core concept of the ontology framework in the medicinal diet knowledge graph is the "Chinese medicinal diet knowledge set". Figure 6 It is the ontology architecture of the medicinal diet knowledge graph. It associates and fuses data based on the entity attributes, entity relationships and other information extracted from the medicinal diet set, ingredient set and standard set. It mainly includes two core parts: medicinal diet and ingredient. Using different colors to divide the node levels can more intuitively understand the relationship between the data and realize the visualization of medicinal diet data, which is useful for supplementing and improving the ontology structure later. In addition, the powerful search capability of the knowledge graph can help users quickly locate nodes and relationships.

[0087] Relationship determination

[0088] Relationship extraction is to obtain the relationship between entities. When there is a relationship between entities, it is necessary to extract the relationship type and parameters between them, which respectively represent the relationship type between entities and the entities involved in this relationship. To extract relationships, the triple of TCM attribute concept entity-relationship-dietary feature entity is stored in the data set, and the remaining sentences are saved in text form to describe and illustrate the relationship between entities. For example, the sentence "Chenpi Fuling Dajujube Congee has the effects of removing dampness and resolving phlegm, strengthening the spleen and harmonizing the middle" can be extracted as follows: the entities are "Chenpi Fuling Dajujube Congee", "Removing dampness and resolving phlegm, strengthening the spleen and harmonizing the middle", and the relationship is "effect". The result of relationship extraction is "Chenpi Fuling Dajujube Congee-effect-Removing dampness and resolving phlegm, strengthening the spleen and harmonizing the middle". This paper adopts the method of manual annotation to extract the relationship between "food-disease indicator" from the established knowledge set, and forms triples of each relationship. The nodes are manually established and the association relationship between entities is defined. The data relationship extraction information can be found in Figure 3 The nodes are entities such as "Chenpi Fuling Dazao Congee", "Gastritis", "Dates, Fuling, Rice", etc., and the edges are relationships such as "Efficacy", "Formula", "Cause Disease", etc. The logical relationship between medicinal food, ingredients and efficacy is clarified through association, and the data is displayed more intuitively.

[0089] Knowledge Fusion

[0090] Formula determination

[0091] Based on the medicinal diet data set, the information of each ingredient in the recipe is extracted according to the user's needs, and the medicinal diet recipe is recommended to the user. The recipe matrix is ​​shown in expression (2). The extracted pork liver and wolfberry porridge recipe contains 5 ingredients, m=5, with a total weight of 2219g, consisting of 22.19% pork liver, 0.67% wolfberry, 22.19% polished rice, 90.13% water, and 0.02% salt. From expression (2-1), the porridge contains 100g pork liver, 15g wolfberry, 100g polished rice, 2000g water, and 4g salt. Similarly, the mustard greens porridge contains 250g of mustard greens, 100g of polished rice, 2000g of water, and 4g of salt; the Codonopsis, Astragalus and Lean Meat Porridge contains 20g of Codonopsis, 20g of Astragalus, 5g of dried tangerine peel, 200g of lean meat, 100g of polished rice, 2000g of water, and 4g of salt; the Gastrodia elata and chrysanthemum porridge contains 10g of Gastrodia elata, 6g of chrysanthemum, 100g of polished rice, 2000g of water, and 4g of salt.

[0092] F=W0[w1,w2,w3,...,wi,...,wn](2)

[0093] Where:

[0094] W0——total weight of medicinal food,

[0095] wi——is the proportion of ingredients in the medicinal diet. i=1,2,3,...m, there are m ingredients in total.

[0096] F1 (pork liver and wolfberry porridge) = w total weight * [w pork liver, w wolfberry, w rice, w water, w salt] (2-1)

[0097] =[w total weight*w pig liver, w total weight*w wolfberry, w total weight*w polished rice, w total weight*w water, w total weight*w salt](2-1)

[0098] F1 (pork liver and wolfberry porridge) = [1001510020004] (2-1)

[0099] Pork liver, wolfberry, rice water, salt

[0100] F1 (mustard porridge) = [25010020004] (2-2)

[0101] Mustard Rice Water Salt

[0102] F1 (Dangshen, Astragalus and Lean Meat Porridge) = [2020520010020004] (2-3)

[0103] Codonopsis, Astragalus, Tangerine Peel, Lean Meat, Rice, Water and Salt F1 (Gastrodia and Chrysanthemum Porridge) = [106100200015] (2-4)

[0104] Gastrodia elata, chrysanthemum, rice water and salt

[0105] Users analyze their health status, physique, age, gender, taste preferences and other information, and perform knowledge extraction and dietary analysis based on the existing data in the knowledge graph, as shown in the following table.

[0106] The knowledge graph extracts and fuses various types of knowledge and then displays them visually, thereby realizing the effectiveness analysis of knowledge. By inputting keywords, the system extracts knowledge and fuses data based on the input keyword information to output query results. The query results are displayed in two ways: "graph" and "text description". The "graph" is the association between the target knowledge and knowledge related to the target, and the "text description" is a general overview of the target knowledge.

[0107] As various knowledge is constantly updated, the underlying knowledge of the medicinal diet knowledge graph should be continuously improved. On the one hand, it is necessary to continuously supplement entity, relationship and attribute triples; on the other hand, it is necessary to update or delete existing knowledge to improve the entire knowledge graph system. This article explores a variety of knowledge extraction methods, which can extract and update knowledge in real time, manually establish or remove the association between various knowledge, and delete, edit, connect, and other operations on a single entity. When deleting entities, relationships, or attributes, delete them from the bottom up in a hierarchical manner to ensure that the relationship between other entities is not affected, and to avoid accidental deletions that lead to incomplete knowledge graph structures.

[0108] The medicinal diet data is identified and extracted, and the structured data is directly entered into the system after integration, and the semi-structured and unstructured data is entered into the system after manual annotation. Collect information related to medicinal diet, ingredients, and standards, and establish an ingredient database based on core information such as nutrients and functional ingredients, a medicinal diet database based on core information such as compatibility, nature and taste, constitution, and dosage form, and a dietary standard database based on core information such as nutritional standards, dietary standards, and efficacy standards. The massive data in the knowledge base is processed by entity recognition, entity extraction and alignment, and relationship extraction. The medicinal diet knowledge graph is constructed through knowledge association and knowledge fusion processing data. By using the knowledge graph, users can quickly query medicinal diet formula information, medicinal diet processing methods, ingredient information of medicinal materials contained in medicinal diet, active compounds, etc., and through the constructed medicinal diet data relationship network, intuitively understand the relationship between each node, assist users in choosing more suitable medicinal diets, and realize the intelligent application of knowledge graphs.

[0109] In summary, the present invention not only obtains an accurate medicinal diet knowledge graph based on the TCM attribute concept entity-relationship-diet attribute, but also combines the medicinal diet recommendation model to topologically update the medicinal diet knowledge graph, thereby further improving the comprehensiveness and accuracy of the medicinal diet knowledge graph.

[0110] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for constructing a medicinal diet knowledge graph, characterized in that: The method comprises: M1. Medicinal diet knowledge graph ontology architecture: a multi-level TCM attribute-relationship-diet attribute architecture, with the bottom layer including ingredients and medicinal diet, the second layer including primary attributes of ingredients and medicinal diet, and the third layer including advanced attributes of ingredients and medicinal diet; M2. Dataset: Dataset of traditional Chinese medicine and dietary elements, including food ingredients, medicinal diet, and standard data sets; M3. Relationship function: including information transmission model, information fusion model, including medicinal diet recommendation model, dietary nutrition model, efficacy model, and health model; M4. Knowledge association: Through the relational function, the TCM attributes and dietary attributes of medicinal food are associated to form a relational network, realizing the flow, transmission and integration of various information in the data set; M5. Reasoning about medicinal diet knowledge: Based on individual basic information, health status, living habits and health management goals, calculate and derive new medicinal diet knowledge, including medicinal diet strategies and health mechanisms.

2. The method for constructing a medicinal diet knowledge graph according to claim 1, characterized in that: The method further comprises: M6. Map the data in the data set to the ontology architecture in order from bottom to top to obtain the medicinal diet knowledge graph ontology architecture with data.

3. The method for constructing a medicinal diet knowledge graph according to claim 1, characterized in that: In step M1, the establishment of the medicinal diet knowledge graph ontology architecture includes: M11. The ontology architecture consists of several nodes, each of which represents a concept of medicinal diet knowledge.

4. The method for constructing a medicinal diet knowledge graph according to claim 2, characterized in that: In step M6, the establishment of the medicinal diet knowledge graph ontology architecture includes: M61. Identify the concepts of the top-level medicinal diet knowledge set and divide it into two categories: ingredients and medicinal diet; M62. Determine the concepts of the second layer, and then gradually refine the next layer of concepts based on the TCM meridian theory, the four properties and five flavors theory, the rise and fall theory, TCM constitution, TCM diet efficacy theory, and food nutrition theory, and establish the ontology architecture of the medicinal diet knowledge graph.

5. The method for constructing a medicinal diet knowledge graph according to claim 1, characterized in that: The dataset is constructed by collecting information related to medicinal diet, ingredients, and standards through web crawler technology and professional dataset docking methods, establishing an ingredient dataset based on the core information of nutritional components and functional components, a medicinal diet dataset based on the core information of compatibility, nature and taste, constitution, and dosage form, and a dietary standard set based on the core information of nutritional standards, dietary standards, and efficacy standards.

6. The method for constructing a medicinal diet knowledge graph according to claim 1, characterized in that: In step M3, constructing the medicinal diet recommendation model according to the effective ingredients and nutrients in the medicinal diet includes: M31. Based on the medicinal diet knowledge graph ontology architecture filled with data, obtain the data information of the number of targets and functional components of the intersection of medicinal diet and disease, the weight data information of ingredients, and the data information of nutrients contained in the ingredients; M32. Based on the data information of the number of targets and efficacy components of the intersection of medicinal diet and disease, the weight data information of ingredients, and the data information of nutrients contained in ingredients, a medicinal diet recommendation model is established, including the efficacy objective function G and the nutritional constraints. G=c1w1+c2w2+…+c i w i …+c n w n , The nutritional constraints are: Among them, G is the content of the effective ingredient, and the maximum value is taken according to individual needs; i is the i-th ingredient, n is the total number of ingredients, c i is the content of the main active ingredient in the i-th food ingredient, w i is the weight of the i-th ingredient in the medicinal diet, j is the j-th nutrient, there are m nutrients in total, m is the total number of nutrient types, a ji is the content of the jth nutrient in the ith ingredient, b 1j is the lower limit of recommended intake of the jth nutrient, b 2j is the upper limit of recommended intake of nutrient j; M33. Based on the objective function G of medicinal diet recommendation, determine the medicinal diet formula w with the maximum content of effective ingredients and basic nutrients that meet the daily recommended intake i , complete the recommendation of medicinal diet.

7. The method for constructing a medicinal diet knowledge graph according to claim 6, characterized in that: The data information of nutrients contained in the food material includes data information of protein, data information of carbohydrates, data information of fat, data information of calories and data information of vitamins.

8. The method for constructing a medicinal diet knowledge graph according to claim 1, characterized in that: The relationship function extracts the data in the data set into concept-relationship-concept triples and appends them to the data set to describe and illustrate the relationship between entities. The method also includes: M7. Based on the medicinal diet knowledge graph ontology architecture, medicinal diet information is extracted through knowledge reasoning on the original data, and the data is combined with knowledge to obtain dietary analysis, model application, and information query functions.

9. A system for constructing a medicinal diet knowledge graph, comprising a computer device, characterized in that: The computer device is programmed or configured to execute the steps of the method for constructing a medicinal diet knowledge graph as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program and data information programmed or configured to execute the method for constructing a medicinal diet knowledge graph as described in any one of claims 1 to 8.

Images