Traditional Chinese medicine term extraction method and knowledge graph construction

Abstract

The invention relates to the field of information processing, in particular to a traditional Chinese medicine term extraction method and knowledge graph construction. When traditional Chinese medicine academic languages have the phenomena of fuzzy concepts, randomness, disordered use and the like, traditional Chinese medicine data of corresponding entities in traditional Chinese medicine ancient books are extracted, the traditional Chinese medicine entities in a traditional Chinese medicine data set are converted into standardized texts, a knowledge graph model is presented in a visualized mode, and finally the knowledge graph model is output in a visualized mode. The traditional Chinese medicine terms are standardized, data can be conveniently recognized by a computer, the accuracy of the traditional Chinese medicine terms is improved, and traditional Chinese medicine data mining is facilitated; the visualized operation increases the intuition of the whole process, facilitates the subsequent operation such as verification or troubleshooting of the traditional Chinese medicine term set by workers, further guarantees the accuracy and depth of the whole traditional Chinese medicine term extraction, and is simple and convenient to operate.

Description

Technical Field

[0001] The present invention relates to the field of information processing, and in particular to a method for extracting traditional Chinese medicine terms and constructing a knowledge graph. Background Art

[0002] The knowledge graph is a giant, networked knowledge system built on the basis of semantic networks. It can capture and present the semantic relationships between domain concepts, connect fragmented knowledge on the Internet, and support comprehensive knowledge retrieval, question-answering, and decision support and other intelligent applications. The knowledge graph of traditional Chinese medicine is mainly a knowledge graph system built on the basis of the language system of traditional Chinese medicine, and uses existing database resources to fill the knowledge graph with content.

[0003] Traditional Chinese medicine terminology refers to the special terms of traditional Chinese medicine, which can reflect the characteristics of traditional Chinese medicine. It is a silent language, a link for academic exchanges in traditional Chinese medicine, and a bridge for communication between medical disciplines. As one of the important contents of the standardization of traditional Chinese medicine terminology, although certain achievements have been made, there is still a big gap from the actual work needs. With the development of traditional Chinese medicine, the concepts of traditional Chinese medicine terminology are vague, arbitrary, and confusing, which also brings great difficulties to the research of related algorithms. Therefore, the standardization of traditional Chinese medicine terminology is imminent.

[0004] Corresponding improvements have also been made to the above-mentioned problems, such as China's patent application number CN202311672005.4, which was published on March 1, 2024. The patent discloses a method and system for extracting TCM terms based on ancient Chinese medical literature, including a data preprocessing module, a candidate term extraction module, a TCM term determination module, and a visualization module. The data preprocessing module is used to segment the text data set of the ancient Chinese medical literature to be extracted; the candidate term extraction module is used to extract named entities from the segmented text data set as candidate terms; the TCM term determination module is used to score the candidate terms, and if they exceed the threshold, they will be included in the TCM term set; the visualization module is used to present the TCM term set. The shortcomings of this patent are: although it effectively solves the blindness of ancient books, the overall accuracy and depth still need to be improved. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a method for extracting Chinese medicine terms and constructing a knowledge graph.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A method for extracting Chinese medicine terms and constructing a knowledge graph, comprising the steps of:

[0008] S1: Extract and process Chinese medicine terms from ancient Chinese medical books;

[0009] S2: Convert the TCM entities in the TCM dataset into standardized fields so that they can be presented in a recognizable way in the knowledge graph model;

[0010] S3: Present the TCM terminology in a visual way.

[0011] Furthermore, in the above-mentioned method for extracting Chinese medicine terms and constructing a knowledge graph, the implementation process of step S1 also includes the following steps:

[0012] S11: inputting a data set of ancient Chinese medicine books; wherein the Chinese medicine text data set includes a plurality of data columns, each of which includes a plurality of data containing Chinese medicine texts;

[0013] S12: Cleaning the traditional Chinese medicine text dataset;

[0014] S13: The Chinese medicine terms extracted from the Chinese medicine include Chinese pinyin;

[0015] S14: The Chinese medicine terminology for the Chinese medicine extraction includes the Latin name;

[0016] S15: The Chinese medicine terms for the Chinese medicine extraction include nature and taste;

[0017] S16: The Chinese medicine terminology for the Chinese medicine extraction includes meridians;

[0018] S17: The Chinese medicine terms extracted from the Chinese medicine include drug classification and efficacy;

[0019] S18: The Chinese medicine terms for the Chinese medicine extraction include the medication site and alias.

[0020] Furthermore, in the above-mentioned method for extracting Chinese medicine terms and constructing a knowledge graph, the implementation process of step S12 also includes the following steps:

[0021] S121: Selecting a required data column in the traditional Chinese medicine text data set;

[0022] S122: traversing the data columns, and manually classifying the data columns with the same column name or the same meaning;

[0023] S123: Traverse the standard data in the data column, generate an association table and an attribute table respectively to construct a knowledge graph.

[0024] Furthermore, in the above-mentioned method for extracting Chinese medicine terms and constructing a knowledge graph, the properties and flavors of the Chinese medicine terms extracted in step S15 include sour, bitter, sweet, pungent, salty, cold, hot, warm, slightly warm, very warm, cool, flat, slightly sour, slightly bitter, slightly sweet, slightly pungent, slightly salty, slightly toxic, non-toxic, highly toxic, very cold, very hot, very cool, slightly toxic, spicy, astringent, slightly astringent, numb, and light.

[0025] Furthermore, in the above-mentioned method for extracting Chinese medicine terms and constructing a knowledge graph, the Chinese medicine terms extracted in step S16 belong to the heart, liver, spleen, lung, kidney, stomach, large intestine, small intestine, bladder, gallbladder, pericardium, and triple burner.

[0026] Further, in the above-mentioned method for extracting Chinese medicine terms and constructing a knowledge graph, the efficacy of Chinese medicine terms extracted in step S17 include antipyretic drugs, heat-clearing drugs, summer-heat-relieving drugs, purgative drugs, wind-relieving drugs, antipruritic drugs, analgesics, dampness-relieving drugs, interior-warming drugs, expectorant, antitussive and antiasthmatic drugs, qi-regulating drugs, blood-regulating drugs, tranquilizers and intelligence-enhancing drugs, tranquilizers and orifices-opening drugs, liver-pacifying and wind-extinguishing drugs, drugs for eliminating accumulation and guiding stagnation, drugs for harmonizing the middle and raising and lowering, drugs for relaxing muscles and activating collaterals, astringent and firming drugs, drugs for moisturizing and promoting body fluid, tonic drugs, drugs for improving hearing and improving eyesight, drugs for anthelmintic and antimalarial drugs, drugs for regulating menstruation and stagnating pregnancy, drugs for stopping leukorrhea, drugs for detoxifying, reducing swelling and dispersing knots, drugs for promoting tissue regeneration, healing sores, corrosive sores and removing rot, and other drugs.

[0027] Furthermore, in the above-mentioned method for extracting Chinese medicine terms and constructing a knowledge graph, in step S123, the association table and the attribute table are input into the knowledge graph, and the node similarity calculation is performed on the elements, and the calculation formula is:

[0028]

[0029] Among them, set A is the nature, taste, meridians, and drug classification standardized by Herb1, set B is the nature, taste, meridians, and drug classification standardized by Herb2, A∩B is the common element of set A and set B, and A∪B is all the different elements in A and B.

[0030] Preferably, one of the visual output forms of the knowledge graph model includes: a knowledge graph image.

[0031] Preferably, one of the visual output forms of the knowledge graph model includes: a knowledge reasoning table.

[0032] The beneficial effects of the present invention are as follows: by extracting the TCM data of the corresponding entities in the ancient TCM books, and converting each TCM entity in the TCM data set into a standardized text, the knowledge graph model is presented in a visual manner and outputted visually. At the same time, the commonly used terms of standard TCM are further extracted and the data is easy to be recognized by the computer, which improves the accuracy of TCM terms and facilitates TCM data mining; the visual operation increases the intuitiveness of the whole process, which is convenient for subsequent staff to verify or troubleshoot the TCM terminology set, further ensuring the accuracy and depth of the entire TCM term extraction, and the operation is simple. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 A flowchart of the method for extracting Chinese medicine terms and constructing a knowledge graph according to an embodiment of the present invention;

[0034] Figure 2 A partial interface diagram of the Chinese medicine terminology after the traditional Chinese medicine ancient books are standardized by the present invention;

[0035] Figure 3 It is the attribute diagram of all the properties and flavors of traditional Chinese medicines after standardization of the present invention;

[0036] Figure 4 It is the attribute diagram of all the Chinese medicine meridians after standardization of the present invention;

[0037] Figure 5 It is the attribute diagram of all the Chinese medicine effects after standardization of the present invention;

[0038] Figure 6 It is a partial interface diagram of the knowledge reasoning table of the network graph of traditional Chinese medicine terms of the present invention;

[0039] Figure 7 It is a local interface diagram of the network knowledge graph of traditional Chinese medicine terms of the present invention; DETAILED DESCRIPTION

[0040] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0041] The present invention provides a method for extracting Chinese medicine terms and constructing a knowledge graph. The specific construction method has the following steps: Figure 1 As shown,

[0042] S1: Extract and process the Chinese medicine terms from ancient Chinese medical books. The scope of Chinese medicine terms is the first-level Chinese medicine discipline, which should be able to reflect the essential characteristics of a Chinese medicine.

[0043] S2: Convert the TCM entities in the TCM dataset into standardized fields so that they can be presented in a recognizable way in the knowledge graph model;

[0044] S3: Present the TCM terminology in a visual way.

[0045] Specifically, the implementation process of step S1 also includes the following steps:

[0046] S11: Inputting the ancient Chinese medicine data set; wherein, the Chinese medicine text data set includes multiple data columns, each of which includes multiple data containing Chinese medicine texts, totaling 16668 kinds of Chinese medicines, specifically Figure 2 As shown;

[0047] S12: Cleaning the traditional Chinese medicine text dataset;

[0048] S13: The Chinese medicine terms extracted from the Chinese medicine include Chinese pinyin;

[0049] S14: The Chinese medicine terminology for the Chinese medicine extraction includes the Latin name;

[0050] S15: The Chinese medicine terms for the Chinese medicine extraction include nature and taste;

[0051] S16: The Chinese medicine terminology for the Chinese medicine extraction includes meridians;

[0052] S17: The Chinese medicine terms extracted from the Chinese medicine include drug classification and efficacy;

[0053] S18: The Chinese medicine terms for the Chinese medicine extraction include the medication site and alias.

[0054] The implementation process of step S12 also includes the following steps:

[0055] S121: Selecting a required data column in the traditional Chinese medicine text data set;

[0056] S122: traversing the data columns, and manually classifying the data columns with the same column name or the same meaning;

[0057] S123: Traverse the standard data in the data column, generate an association table and an attribute table respectively to construct a knowledge graph.

[0058] The Chinese medicine terms extracted in step S15 include sour, bitter, sweet, pungent, salty, cold, hot, warm, slightly warm, very warm, cool, neutral, slightly sour, slightly bitter, slightly sweet, slightly pungent, slightly salty, slightly toxic, non-toxic, very toxic, very cold, very hot, very cool, slightly toxic, spicy, astringent, slightly astringent, numb, and light, a total of 29 types, as shown in the following examples. Figure 3 As shown;

[0059] The Chinese medicine terms extracted in step S16 include heart, liver, spleen, lung, kidney, stomach, large intestine, small intestine, bladder, gallbladder, pericardium, and triple energizer, a total of 12 types. Figure 4 As shown;

[0060] The Chinese medicine term efficacy extracted in step S17 includes antipyretic drugs, heat-clearing drugs, heat-removing drugs, purgative drugs, wind-removing drugs, itching drugs, analgesics, dampness-removing drugs, warming drugs, expectorant, cough-relieving and asthma-relieving drugs, qi-regulating drugs, blood-regulating drugs, tranquilizing and intelligence-enhancing drugs, convulsion-inducing and orifice-opening drugs, liver-pacifying and wind-extinguishing drugs, drugs for removing accumulation and stagnation, drugs for harmonizing the middle and lifting and lowering, drugs for relaxing muscles and activating collaterals, astringent and astringent drugs, drugs for moistening dryness and promoting body fluid, tonic drugs, drugs for improving hearing and improving eyesight, drugs for anthelmintic and antimalarial drugs, drugs for regulating menstruation and stagnating pregnancy, drugs for stopping leukorrhea, drugs for detoxifying, reducing swelling and dispersing knots, drugs for promoting tissue regeneration, healing sores, corrosive sores and removing rot, and other drugs, totaling 28 types, specifically as follows Figure 5 As shown;

[0061] In step S123, the association table and the attribute table are input into the knowledge graph, and the node similarity calculation is performed on the elements, and the calculation formula is:

[0062]

[0063] Among them, set A is the nature, taste, meridians, and drug classification standardized by Herb1, set B is the nature, taste, meridians, and drug classification standardized by Herb2, A∩B is the common element of set A and set B, and A∪B is all the different elements in A and B.

[0064] In the above-mentioned method for extracting Chinese medicine terms and constructing a knowledge graph, one of the forms of visually outputting the knowledge graph model includes: a knowledge graph image, specifically, Figure 6 Specifically, Neo4j provides a visual and intuitive expression of different node colors representing different attributes, where blue nodes represent Chinese medicine, purple nodes represent efficacy classification, red nodes represent meridians, and brown nodes represent flavor. At the same time, according to the correlation of Chinese medicine attribute nodes, one of the visual output forms of the knowledge graph model includes: knowledge reasoning table, so as to evaluate the correlation between Chinese medicines, as shown in the figure. Figure 7 shown.

[0065] By extracting the TCM data of the corresponding entities in ancient TCM books and converting the TCM entities in the TCM data set into standardized text, the knowledge graph model and visual output are presented in a visual way. At the same time, the commonly used terms of standard TCM are further extracted and the data is easy to be recognized by the computer, which improves the accuracy of TCM terms and facilitates TCM data mining; the visual operation increases the intuitiveness of the whole process, making it easier for subsequent staff to verify or check the TCM terminology set, further ensuring the accuracy and depth of the entire TCM term extraction, and the operation is simple.

[0066] The examples described in the present invention are merely descriptions of the preferred implementation modes of the present invention, and are not intended to limit the concept and scope of the present invention. Without departing from the design concept of the present invention, various modifications and improvements made to the technical solutions of the present invention by engineers and technicians in this field should all fall within the protection scope of the present invention.

Claims

1. A method for extracting Chinese medicine terms and constructing a knowledge graph, characterized in that: The steps include: S1: Extract and process Chinese medicine terms from ancient Chinese medical books; S2: Convert the TCM entities in the TCM dataset into standardized fields so that they can be presented in a recognizable way in the knowledge graph model; S3: Present the TCM terminology in a visual way.

2. According to claim 1, a method for extracting Chinese medicine terms and constructing a knowledge graph is characterized in that: The step S1 includes the following fields: S11: inputting a data set of ancient Chinese medicine books; wherein the Chinese medicine text data set includes a plurality of data columns, each of which includes a plurality of data containing Chinese medicine texts; S12: Cleaning the traditional Chinese medicine text dataset; S13: The Chinese medicine terms extracted from the Chinese medicine include Chinese pinyin; S14: The Chinese medicine terminology for the Chinese medicine extraction includes the Latin name; S15: The Chinese medicine terms for the Chinese medicine extraction include nature and taste; S16: The Chinese medicine terminology for the Chinese medicine extraction includes meridians; S17: The Chinese medicine terms extracted from the Chinese medicine include drug classification and efficacy; S18: The Chinese medicine terms for the Chinese medicine extraction include the medication site and alias.

3. According to claim 2, a method for extracting Chinese medicine terms and constructing a knowledge graph is characterized in that: The step S12 comprises: S121: Selecting a required data column in the traditional Chinese medicine text data set; S122: traversing the data columns, and manually classifying the data columns with the same column name or the same meaning; S123: Traverse the standard data in the data column, generate an association table and an attribute table respectively to construct a knowledge graph.

4. The method for extracting Chinese medicine terms and constructing a knowledge graph according to claim 2, characterized in that: The Chinese medicine terms for properties and flavors extracted in step S15 include sour, bitter, sweet, pungent, salty, cold, hot, warm, slightly warm, very warm, cool, neutral, slightly sour, slightly bitter, slightly sweet, slightly pungent, slightly salty, slightly toxic, non-toxic, very toxic, very cold, very hot, very cool, slightly toxic, spicy, astringent, slightly astringent, numb, and bland.

5. The method for extracting Chinese medicine terms and constructing a knowledge graph according to claim 2, characterized in that: The Chinese medicine terms extracted in step S16 include the heart, liver, spleen, lung, kidney, stomach, large intestine, small intestine, bladder, gallbladder, pericardium, and triple energizer.

6. The method for extracting Chinese medicine terms and constructing a knowledge graph according to claim 2, characterized in that: The Chinese medicine terms extracted in step S17 include antipyretic drugs, heat-clearing drugs, summer-removing drugs, purgative drugs, wind-removing drugs, antipruritic drugs, analgesics, dampness-removing drugs, interior-warming drugs, expectorant, antitussive and antiasthmatic drugs, qi-regulating drugs, blood-regulating drugs, tranquilizing and intelligence-enhancing drugs, twitching and invigorating drugs, liver-pacifying and wind-extinguishing drugs, drugs for removing accumulation and guiding stagnation, drugs for harmonizing the middle and lifting and lowering, drugs for relaxing muscles and activating collaterals, astringent and firming drugs, drugs for moisturizing and promoting body fluid, tonic drugs, drugs for improving hearing and improving eyesight, drugs for anthelmintic and antimalarial drugs, drugs for regulating menstruation and stagnating pregnancy, drugs for stopping leukorrhea, drugs for detoxifying, reducing swelling and dispersing knots, drugs for promoting tissue regeneration, healing sores, corrosive sores and removing rot, and other drugs.

7. The method for extracting Chinese medicine terms and constructing a knowledge graph according to claim 3, characterized in that: In step S123, the association table and the attribute table are input into the knowledge graph, and the node similarity calculation is performed on the elements, and the calculation formula is: Among them, set A is the nature, taste, meridians, and drug classification standardized by Herb1, set B is the nature, taste, meridians, and drug classification standardized by Herb2, A∩B is the common element of set A and set B, and A∪B is all the different elements in A and B.

8. The method for extracting Chinese medicine terms and constructing a knowledge graph according to claim 1, characterized in that: One of the visual output forms of the knowledge graph model includes: a knowledge graph image.

9. The method for extracting Chinese medicine terms and constructing a knowledge graph according to claim 1, characterized in that: One of the visual output forms of the knowledge graph model includes: a knowledge reasoning table.

Images