AI teaching methods and related devices based on traditional Chinese medicine culture

Abstract

The present application discloses an AI teaching method based on TCM culture and related devices, the method comprising: obtaining a target TCM course and a target teaching plan; obtaining a target TCM database corresponding to the target TCM course; constructing a target TCM AI model based on the target TCM database and the target teaching plan; obtaining a target medical question and personal information; inputting the target medical question and personal information into the target TCM AI model to obtain a first reply content; obtaining a first feedback content corresponding to the first reply content; determining whether the target object meets a first preset condition based on the first feedback content; when the target object meets the first preset condition, generating a medical case based on the target medical question; obtaining a feedback content corresponding to the a medical case; when the a feedback content meets a second preset condition, determining that the target object has mastered the TCM knowledge corresponding to the target medical question. The embodiment of the present application improves the efficiency of TCM teaching.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to an AI teaching method and related devices based on traditional Chinese medicine culture. Background Art

[0002] As a key representative of traditional culture, Traditional Chinese Medicine (TCM) encompasses a wide range of fields, including the cultivation and processing of Chinese medicinal materials, pharmaceutical trade, and TCM diagnosis and treatment. Currently, TCM theory and experience often rely on master-apprentice inheritance and ancient texts, limiting the dissemination and inheritance of knowledge. Furthermore, TCM theory and terminology are often difficult for the average person to understand, resulting in inefficient TCM teaching.

[0003] Therefore, how to improve the efficiency of TCM teaching has become an urgent problem to be solved. Summary of the Invention

[0004] The embodiments of the present application provide an AI teaching method and related devices based on TCM culture, which improve the efficiency of TCM teaching.

[0005] In a first aspect, an embodiment of the present application provides an AI teaching method based on TCM culture, the method comprising:

[0006] Obtaining a target TCM course for a target subject and a target teaching plan for the target TCM course; the target TCM course includes one of the following: a Zhangshu TCM course, a Chinese medicine course, a TCM basic theory course, or an acupuncture and massage course;

[0007] Obtaining a target TCM database corresponding to the target TCM course;

[0008] Constructing a TCM AI model according to the target TCM database and the target teaching plan to obtain a target TCM AI model;

[0009] Obtaining target medical issues for the target TCM course and personal information of the target subject;

[0010] Inputting the target medical question and the personal information into the target TCM AI model to obtain a first response;

[0011] Obtaining first feedback content from the target object regarding the first reply content;

[0012] determining whether the target object meets a first preset condition according to the first feedback content;

[0013] When the target object meets the first preset condition, a corresponding medical case is generated based on the target medical problem to obtain a medical case; a is an integer greater than 1;

[0014] Obtaining feedback content from the target subject for each of the a medical cases to obtain a pieces of feedback content;

[0015] Detecting whether the a feedback contents meet a second preset condition;

[0016] When the a pieces of feedback content meet the second preset condition, it is determined that the target subject has mastered the traditional Chinese medicine knowledge corresponding to the target medical problem.

[0017] In a second aspect, an embodiment of the present application provides an AI teaching device based on traditional Chinese medicine culture, the device comprising: an acquisition unit, a model building unit, and a teaching unit; wherein:

[0018] The acquisition unit is configured to acquire a target TCM course for a target subject and a target teaching plan for the target TCM course; the target TCM course includes one of the following: a Zhangshu TCM course, a Chinese medicine course, a TCM basic theory course, or an acupuncture and massage course; and acquire a target TCM database corresponding to the target TCM course;

[0019] The model building unit is used to build a TCM AI model according to the target TCM database and the target teaching plan to obtain a target TCM AI model;

[0020] The acquisition unit is further configured to acquire target medical questions for the target TCM course and personal information of the target subject;

[0021] The teaching unit is used to input the target medical question and the personal information into the target traditional Chinese medicine AI model to obtain a first reply content; obtain the first feedback content of the target object for the first reply content; determine whether the target object meets the first preset condition based on the first feedback content; when the target object meets the first preset condition, generate a corresponding medical case based on the target medical question to obtain a medical case; a is an integer greater than 1; obtain the feedback content of the target object for each of the a medical cases to obtain a feedback content; detect whether the a feedback contents meet the second preset condition; when the a feedback contents meet the second preset condition, determine that the target object has mastered the traditional Chinese medicine knowledge corresponding to the target medical question.

[0022] In a third aspect, an embodiment of the present invention provides an electronic device, comprising: a processor and a memory, wherein the memory is used to store one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor, and the program includes instructions for executing the steps in the first aspect of the embodiment of the present invention.

[0023] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program for electronic data exchange, wherein the computer program enables a computer to execute some or all of the steps described in the first aspect of the embodiment of the present invention.

[0024] In a fifth aspect, embodiments of the present invention provide a computer program product, wherein the computer program product includes a non-transitory computer-readable storage medium storing a computer program, wherein the computer program is operable to cause a computer to perform some or all of the steps described in the first aspect of the embodiments of the present invention. The computer program product may be a software installation package.

[0025] The implementation of this application has the following beneficial effects:

[0026] It can be seen that the AI ​​teaching method based on traditional Chinese medicine culture described in this application obtains the first reply content by inputting the target object's personal information together with the target medical question into the model. The addition of personal information enables the reply to fully consider the characteristics of the target object, such as age, cultural background, knowledge base and other factors, thereby making personalized replies to help the target object better understand traditional Chinese medicine knowledge, reduce the time for understanding, and thus improve the efficiency of traditional Chinese medicine teaching. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the background technology, the drawings required for use in the embodiments of the present application or the background technology will be described below.

[0028] Figure 1 This is a schematic diagram of the structure of a traditional Chinese medicine teaching system provided in an embodiment of the present application;

[0029] Figure 2 This is a flowchart of an AI teaching method based on TCM culture provided in an embodiment of the present application;

[0030] Figure 3 This is a block diagram of the functional units of an AI teaching device based on TCM culture provided in an embodiment of the present application;

[0031] Figure 4 This is a structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0032] In order to enable those skilled in the art to better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0033] The terms "first," "second," and the like in the specification and claims of this application and the accompanying drawings are used to distinguish between different objects, not to describe a particular order. Furthermore, the terms "including," "having," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or elements is not limited to the listed steps or elements but may optionally include steps or elements not listed, or may optionally include other steps or elements inherent to the process, method, product, or apparatus.

[0034] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0035] The electronic devices described in the embodiments of the present application may include a traditional Chinese medicine teaching system, a smart phone (such as an Android phone, an iOS phone, a Windows Phone phone, etc.), a tablet computer, a PDA, a laptop computer, a video matrix, a monitoring platform, a mobile Internet device (MID) or a wearable device, etc. The above are only examples and not exhaustive, including but not limited to the above devices. Of course, the above electronic devices can also be servers, for example, a cloud server.

[0036] The following is an explanation of some professional terms involved in this application:

[0037] Traditional Chinese Medicine (TCM) refers to traditional medicine, characterized by a unique theoretical style and rich diagnostic and treatment experience. It uses the theory of Yin and Yang and the Five Elements as its philosophical foundation to explain the physiological functions and pathological changes of the human body and to guide the diagnosis and treatment of diseases. For example, the five internal organs (heart, liver, spleen, lungs, and kidneys) correspond to the Five Elements (fire, wood, earth, metal, and water). These elements interact with each other, such as wood generating fire and fire overcoming metal. This relationship also reflects the interconnectedness and constraints between the internal organs. In pathological conditions, for example, lesions in the liver (which belongs to the Wood element) may affect the heart (which belongs to the Fire element). This is a manifestation of TCM's explanation of the transmission of disease based on the Five Elements' mutual generation and restraint.

[0038] TCM keywords: refers to words that are of great significance in the field of TCM. These words are key elements in understanding and spreading TCM knowledge. For example, the commonly said "cold" can be expressed in TCM through TCM keywords such as wind-cold cold and wind-heat cold. TCM doctors can quickly diagnose diseases based on these TCM keywords.

[0039] See also Figure 1 , Figure 1 This is a schematic diagram of the structure of a TCM teaching system provided in an embodiment of the present application. Figure 1 As shown, the TCM teaching system includes: a user terminal device, a server, and a TCM teaching module. The AI ​​teaching method based on TCM culture provided in the embodiment of the present application can be applied to the TCM teaching system, wherein:

[0040] The user-end device is used to receive input from the target subject regarding the target TCM course, the target teaching plan for that course, and other related information. For example, the target subject can select on the user-end device (such as a mobile phone, tablet, or computer) whether they want to study the Zhangshu TCM course, the Chinese Pharmacy course, or another course. They can also enter their desired teaching schedule, learning focus, and other relevant teaching plan information. The target subject also uses this device to input target medical questions and personal information for the target TCM course, allowing the system to provide personalized teaching services based on these inputs.

[0041] The server is used to store the target TCM database corresponding to the target TCM course. This database contains a wealth of TCM knowledge, including ancient TCM texts, modern TCM research results, clinical cases, information on traditional Chinese medicine, and acupuncture and massage knowledge. The server also stores user personal information, learning records, feedback, and other data for comprehensive management and analysis of the user's learning process. For example, the server stores each target medical question asked by the target subject, the system's response, and the target subject's feedback, so that subsequent teaching services can be optimized based on this data.

[0042] The Traditional Chinese Medicine teaching module is used to build a Traditional Chinese Medicine AI model based on the target Traditional Chinese Medicine database and target teaching plan input by the user-end device, and answer the target object's questions through the Traditional Chinese Medicine AI model to help the target object better understand and apply Traditional Chinese Medicine knowledge.

[0043] See also Figure 2 , Figure 2 : is a flowchart of an AI teaching method based on TCM culture provided in an embodiment of the present application; the method may include the following steps:

[0044] S201. Obtain a target TCM course for a target subject and a target teaching plan for the target TCM course; the target TCM course includes one of the following: Zhangshu TCM course, Chinese medicine course, basic TCM theory course, and acupuncture and massage course.

[0045] In the embodiment of the present application, the target teaching plan may include the scope of TCM knowledge that the target subject needs to learn, as well as the teaching objectives. It should be explained that the AI ​​teaching method based on TCM culture provided in the embodiment of the present application can be applied to Figure 1 The Traditional Chinese Medicine teaching system is shown.

[0046] In a specific embodiment, a variety of different types of TCM courses can be provided for the target subject to choose from. The TCM course selected by the target subject is used as the target TCM course. Then, a target teaching plan for the target TCM course can be formulated. Specifically, the learning objective of the target subject in learning the target TCM course can be obtained. Then, a suitable teaching plan can be formulated based on the learning objective to obtain a target teaching plan. For example, assuming that the learning objective is to understand the basic knowledge of Zhangshu TCM, then it can be determined that the scope of TCM knowledge that the target subject needs to learn is Zhangshu TCM. The teaching objective is to make the target subject familiar with Zhangshu TCM knowledge and able to use the learned Zhangshu TCM knowledge to solve some simple Zhangshu TCM problems.

[0047] S202: Acquire a target TCM database corresponding to the target TCM course.

[0048] In the embodiment of the present application, the target TCM database contains all TCM knowledge that the target subject needs to learn in the target TCM course, such as TCM theoretical knowledge, TCM diagnostic methods, TCM knowledge, etc.

[0049] In a specific embodiment, a mapping relationship between preset TCM courses and TCM databases may be pre-stored, and a target TCM database corresponding to a target TCM course may be determined based on the mapping relationship.

[0050] S203: Construct a TCM AI model according to the target TCM database and the target teaching plan to obtain a target TCM AI model.

[0051] In the embodiment of the present application, the target TCM AI model may include one of the following: a deep learning model, a neural network model, a knowledge graph model, etc., which is not limited here.

[0052] In a specific embodiment, a target TCM AI model can be constructed based on a target TCM database and a target teaching plan.

[0053] Optionally, step S203, constructing a TCM AI model based on the target TCM database and the target teaching plan to obtain the target TCM AI model, may include the following steps:

[0054] S31, determining target data characteristics of the TCM data in the target TCM database;

[0055] S32. Select an AI model corresponding to the characteristics of the target data from a preset AI model library to obtain an initial AI model;

[0056] S33, selecting a first data set from the target traditional Chinese medicine database according to the target teaching plan;

[0057] S34, obtaining the target learning progress of the target subject for the target TCM course;

[0058] S35, dividing the first data set into a second data set and a third data set based on the target learning progress; the second data set is the TCM knowledge that the target subject has already learned, and the third data set is the TCM knowledge that the target subject has not yet learned;

[0059] S36. Divide the second data set and the third data set into a training set and a test set according to a preset ratio;

[0060] S37. Train the initial AI model according to the training set to obtain a first AI model;

[0061] S38. Test the first AI model using the test set to obtain a test result, and determine a first accuracy rate based on the test result;

[0062] S39. When the first accuracy rate is greater than a first preset accuracy rate, the first AI model is used as the target TCM AI model.

[0063] In an embodiment of the present application, the target data characteristics may include at least one of the following: language characteristics, rich association rules, multimodal characteristics (including data in multiple modalities such as text, images, and sounds), etc., which are not limited here; the preset ratio and the first preset accuracy rate can be preset or defaulted in advance.

[0064] In a specific embodiment, target data characteristics of the TCM data in the target TCM database are determined. Specifically, the target data characteristics may be language characteristics. Several TCM data may be randomly selected from the target TCM database, and the sentence structures of these several TCM data may be analyzed. If the analysis determines that there are more sentence structures of classical Chinese, then the language characteristics may be determined to be classical language, which is the target data characteristic. Then, an AI model corresponding to the target data characteristics may be selected from a preset AI model library to obtain an initial AI model. For example, a mapping relationship between preset data characteristics and AI models may be pre-stored, and the initial AI model corresponding to the target data characteristics is determined based on the mapping relationship. Then, the scope of TCM knowledge to be learned by the target object may be determined according to the target teaching plan, and corresponding data may be selected from the target TCM database according to the scope of TCM knowledge to obtain a first data set.

[0065] Next, the target learning progress of the target object for the target traditional Chinese medicine course can be obtained. Specifically, the learning record of the target object can be obtained, and the target learning progress can be determined based on the learning record; then, based on the target learning progress, the first data set is divided into the traditional Chinese medicine knowledge that the target object has learned to obtain a second data set, and the traditional Chinese medicine knowledge that the target object has not yet learned to obtain a third data set; then, the second data set and the third data set can be divided into a training set and a test set according to a preset ratio. For example, assuming the preset ratio is 7:3, 70% of the data in the second data set and the third data set can be used as a training set, and the remaining 30% of the data can be used as a test set; then, the initial AI model can be trained based on the training set to obtain a first AI model; then, the first AI model is tested using the test set to obtain a test result, and the number of correct results output by the first AI model in the test result is obtained. The number of correct results is divided by the total number of test results to obtain a first accuracy rate; when the first accuracy rate is greater than the first preset accuracy rate, the first AI model is used as the target traditional Chinese medicine AI model.

[0066] When the first accuracy rate is not greater than the first preset accuracy rate, it means that the accuracy rate of the first AI model is low and cannot meet the usage requirements. The first AI model needs to be further trained, or an AI model needs to be rebuilt and tested again until the first accuracy rate is greater than the first preset accuracy rate to obtain the target traditional Chinese medicine AI model.

[0067] In this way, by dividing the dataset into learned knowledge (the second dataset) and unlearned knowledge (the third dataset) based on the target subject's learning progress, and further dividing it into training and test sets, model training can be closely aligned with the target subject's actual learning situation. For previously learned knowledge, the model can provide reinforcement training; for unlearned knowledge, targeted preview or initial learning guidance can be provided. For example, in a basic theory course on Traditional Chinese Medicine, if the target subject has already learned the theory of Yin and Yang and the Five Elements, the model can use data related to this knowledge for reinforcement training during training, while also providing appropriate learning guidance for subsequent unlearned knowledge such as the theory of Qi, Blood, and Body Fluids.

[0068] S204: Obtain target medical questions for the target traditional Chinese medicine course and personal information of the target subject.

[0069] In the embodiment of the present application, target medical questions for a target TCM course and personal information of a target subject are obtained. Specifically, the target medical questions and personal information may be actively input into the TCM teaching system by the target subject.

[0070] S205. Input the target medical question and the personal information into the target TCM AI model to obtain a first reply.

[0071] In an embodiment of the present application, the target medical question and personal information can be input into the target TCM AI model, and the target TCM AI model outputs the first reply content.

[0072] Optionally, step S205, inputting the target medical question and the personal information into the target TCM AI model to obtain a first response content, may include the following steps:

[0073] A1. Extract keywords from the target medical question to obtain m keywords; the m keywords are used to indicate the inquiry intent of the target medical question; m is a positive integer;

[0074] A2. Determine the synonym set corresponding to each of the m keywords in the target TCM database, and obtain m synonym sets; each synonym set includes multiple TCM keywords;

[0075] A3. replacing the m keywords based on the m synonym sets to obtain m TCM keywords;

[0076] A4. Determine the query intent of the target medical question based on the m TCM keywords to obtain the target query intent;

[0077] A5. Generate a response content corresponding to the target inquiry intent to obtain a reference first response content;

[0078] A6. Adjust the reference first reply content according to the personal information to obtain the first reply content.

[0079] In an embodiment of the present application, keywords in the target medical question can be extracted to obtain m keywords. Specifically, a Chinese word segmentation tool (such as Jieba word segmentation) can be used to segment the target medical question. For example, for the question "How does traditional Chinese medicine treat coughs caused by colds", after word segmentation, m keywords such as "traditional Chinese medicine", "how", "treatment", "cold", "cause", and "cough" are obtained. Then, the synonym set corresponding to each keyword in the target traditional Chinese medicine database can be determined to obtain m synonym sets. Specifically, for each keyword in the m keywords, its word vector similarity (such as cosine similarity) with the traditional Chinese medicine keyword in the target traditional Chinese medicine database is calculated, and a similarity threshold is set. The words with a similarity higher than the threshold are used as synonyms, thereby obtaining m synonym sets. For example, for the keyword "ginseng", through word vector calculation, it is found that words such as "Codonopsis pilosula" and "American ginseng" are close to "ginseng" in the semantic vector space, and the cosine similarity is greater than the similarity threshold. These words can be included in the synonym set of "ginseng".

[0080] Furthermore, m keywords can be replaced based on m synonym sets to obtain m TCM keywords; then, the inquiry intention of the target medical question can be determined based on the m TCM keywords to obtain the target inquiry intention. Specifically, the subject, action, object and other information of the target medical question can be determined based on the m TCM keywords, and the arrangement order and grammatical structure of the m TCM keywords can be observed. The intention can be understood under the Chinese grammatical rules to obtain the target inquiry intention. For example, "Chinese herbal prescriptions", "formulation principles", and "treatment of cough" may be "formulation principles of Chinese herbal prescriptions for treating cough" according to normal grammar and logic. The word order reflects the focus and order of the questioner's attention to a certain extent, which helps to determine the inquiry intention; the target TCM AI model generates the reply content corresponding to the target inquiry intention to obtain the reference first reply content; finally, the reference first reply content can be adjusted according to personal information to obtain the first reply content.

[0081] By identifying and replacing synonyms for each keyword, the scope of semantic understanding is broadened. Because Traditional Chinese Medicine (TCM) contains numerous synonyms and near-synonyms, such as "wei pi tong tong" (stomach pain) and "wei tong tong" (stomach pain), which often express similar symptoms, incorporating synonyms into question analysis can mitigate biased understanding caused by the questioner using different but similar terms. This allows for a more comprehensive and accurate understanding of the question's core intent, resulting in a more precise response.

[0082] Optionally, step A3, replacing the m keywords based on the m synonym sets to obtain m TCM keywords, may include the following steps:

[0083] B1. Obtain a target synonym set and target keywords corresponding to the target synonym set; the target synonym set is any one of the m synonym sets; the target synonym set includes n traditional Chinese medicine keywords; n is a positive integer;

[0084] B2. Calculate the semantic similarity between each TCM keyword in the target synonym set and the target keyword to obtain n semantic similarities;

[0085] B3, determining the semantic similarities greater than a first preset similarity threshold among the n semantic similarities, to obtain i semantic similarities; i is a positive integer less than or equal to n;

[0086] B4, determining the TCM keywords corresponding to the i semantic similarities in the target synonym set, and obtaining i TCM keywords;

[0087] B5. Determine a TCM keyword associated with the target medical problem among the i TCM keywords to obtain at least one TCM keyword;

[0088] B6. Determine the TCM keyword with the greatest semantic similarity among the at least one TCM keyword based on the i semantic similarities, and obtain the TCM keyword corresponding to the target synonym set.

[0089] In the embodiment of the present application, the first preset similarity threshold may be preset in advance or set by default.

[0090] In a specific embodiment, a target synonym set and a target keyword corresponding to the target synonym set can be obtained; then, the semantic similarity between each traditional Chinese medicine keyword in the target synonym set and the target keyword can be calculated to obtain n semantic similarities. Specifically, the traditional Chinese medicine keywords and the target keywords in the target synonym set can be converted into word vectors through a preset word vector model, and a suitable vector similarity calculation method (for example, cosine similarity) is used to calculate the semantic similarity between each traditional Chinese medicine keyword in the target synonym set and the target keyword to obtain n semantic similarities, wherein the preset word vector model can be preset or defaulted in advance; then, the semantic similarity greater than the first preset similarity threshold among the n semantic similarities can be found to obtain i semantic similarities.

[0091] Furthermore, the TCM keywords corresponding to i semantic similarities in the target synonym set can be determined to obtain i TCM keywords; the TCM keywords associated with the target medical problem among the i TCM keywords can be determined to obtain at least one TCM keyword. Specifically, the inquiry intention of the target medical problem can be determined first based on the above m keywords to obtain the initial inquiry intention, and at least one TCM keyword can be selected from the i TCM keywords based on the initial inquiry intention. For example, assuming that the i TCM keywords are: "cough and reverse", "sputum", "dry cough", "wet cough", etc., and the initial inquiry intention is "how to treat cough with phlegm", then it can be determined that "sputum" and "wet cough" among the i TCM keywords are associated with the target medical problem, and at least one TCM keyword can be obtained; then, at least one semantic similarity corresponding to at least one TCM keyword can be extracted from the i semantic similarities, the maximum value among the at least one semantic similarity can be determined, and the TCM keyword corresponding to the maximum value can be used as the TCM keyword corresponding to the target synonym set.

[0092] In this way, by calculating the semantic similarity between each TCM keyword in the target synonym set and the target keyword, and screening out the parts that are greater than the first preset similarity threshold, we can accurately focus on those TCM keywords that are semantically similar to the target keyword, thereby improving analysis efficiency.

[0093] Optionally, in step A6, the personal information includes: target educational information and target age information; and the reference first reply content is adjusted based on the personal information to obtain the first reply content, including:

[0094] C1. Extract TCM keywords from the reference first reply content to obtain j TCM keywords, where j is a positive integer;

[0095] C2. Determine a comprehension difficulty parameter for each of the j TCM keywords based on the personal information to obtain j comprehension difficulty parameters;

[0096] C3. Determine a first comprehension difficulty parameter based on the j comprehension difficulty parameters;

[0097] C4. Determining a second comprehension difficulty parameter acceptable to the target subject based on the target educational background information;

[0098] C5. When the first comprehension difficulty parameter is greater than the second comprehension difficulty parameter, determining a difference between the first comprehension difficulty parameter and the second comprehension difficulty parameter to obtain a target comprehension difficulty difference;

[0099] C6. Selecting k comprehension difficulty parameters from the j comprehension difficulty parameters based on the target comprehension difficulty difference; the sum of the k comprehension difficulty parameters is greater than or equal to the target comprehension difficulty difference; k is a positive integer less than or equal to j;

[0100] C7. Determine the TCM keywords corresponding to the k comprehension difficulty parameters among the j TCM keywords, to obtain k TCM keywords;

[0101] C8. Selecting TCM keyword annotations corresponding to the k TCM keywords from the target TCM database to obtain k TCM keyword annotations;

[0102] C9. Replace the k TCM keywords in the reference first reply content with the k TCM keyword annotations to obtain a second reply content;

[0103] C10. determining a target response format corresponding to the target age information;

[0104] C11. Adjust the second reply content according to the target reply format to obtain the first reply content.

[0105] In the embodiment of the present application, the comprehension difficulty parameter is used to measure the difficulty of the target object in understanding a traditional Chinese medicine keyword.

[0106] In a specific embodiment, TCM keywords in the reference first reply content can be extracted to obtain j TCM keywords. Specifically, the method for obtaining j TCM keywords can be the same as the method for obtaining m keywords, which will not be repeated here. Then, the comprehension difficulty parameter of each TCM keyword in the j TCM keywords can be determined based on the personal information to obtain j comprehension difficulty parameters. Specifically, the TCM teaching textbook corresponding to the target TCM course can be first obtained, and the chapter position of each TCM keyword in the TCM teaching textbook can be determined to obtain j chapter positions. J reference comprehension difficulty parameters are determined based on the j chapter positions. A mapping relationship between a preset chapter position and a reference comprehension difficulty parameter can be pre-stored. The j reference comprehension difficulty parameters corresponding to the j chapter positions are determined based on the mapping relationship. The later the chapter position, the greater the reference comprehension difficulty parameter. Then, the target comprehension ability parameter of the target object can be determined based on the target academic qualification information in the personal information. For example, a mapping relationship between a preset academic qualification information and a comprehension ability parameter can be pre-stored. The target comprehension ability parameter corresponding to the target academic qualification information is determined based on the mapping relationship. The j reference comprehension difficulty parameters are adjusted according to the target comprehension ability parameter to obtain j comprehension difficulty parameters. The specific adjustment formula is as follows:

[0107] First comprehension difficulty parameter = first reference comprehension difficulty parameter × (1-(target comprehension parameter-preset comprehension parameter) / preset comprehension parameter);

[0108] Among them, the preset comprehension parameter can be preset or defaulted in advance, the first reference comprehension difficulty parameter is any reference comprehension difficulty parameter among j reference comprehension difficulty parameters; the first comprehension difficulty parameter is the comprehension difficulty parameter corresponding to the first reference comprehension difficulty parameter among the j comprehension difficulty parameters.

[0109] It should be explained that the comprehension parameter is a parameter that represents the target subject's ability to understand the TCM language. The larger the comprehension parameter is, the easier it is for the target subject to understand the TCM language.

[0110] Next, a first comprehension difficulty parameter can be determined based on the j comprehension difficulty parameters. Specifically, the average comprehension difficulty parameter of the j comprehension difficulty parameters, that is, the first comprehension difficulty parameter, can be calculated. Then, a second comprehension difficulty parameter acceptable to the target object can be determined based on the target educational information. Specifically, a mapping relationship between preset educational information and comprehension difficulty parameters can be pre-stored, and the second comprehension difficulty parameter corresponding to the target educational information is determined based on the mapping relationship. When the first comprehension difficulty parameter is greater than the second comprehension difficulty parameter, it means that the comprehension difficulty of the j TCM keywords is too great. At this time, the second comprehension difficulty parameter can be subtracted from the first comprehension difficulty parameter to obtain a target comprehension difficulty difference. Then, based on the target comprehension difficulty difference, the second comprehension difficulty parameter can be averaged from the j comprehension difficulty parameters. Select k comprehension difficulty parameters. Specifically, j comprehension difficulty parameters can be sorted from large to small to obtain a first comprehension difficulty parameter sequence. The first parameter sum is initialized to 0, and the comprehension difficulty parameters in the first comprehension difficulty parameter sequence are taken out in sequence. Each time an comprehension difficulty parameter is taken out, it is added to the first parameter sum. The size relationship between the latest first parameter sum and the target comprehension difficulty difference is compared. If the first parameter sum is greater than or equal to the target comprehension difficulty difference, stop taking out the comprehension difficulty parameters in the first comprehension difficulty parameter sequence to obtain k comprehension difficulty parameters. Conversely, if the first parameter sum is less than the target comprehension difficulty difference, continue taking out the comprehension difficulty parameters in the first comprehension difficulty parameter sequence until the first parameter sum is greater than or equal to the target comprehension difficulty difference.

[0111] Furthermore, k TCM keywords corresponding to the k comprehension difficulty parameters in j TCM keywords can be found to obtain k TCM keywords; then, TCM keyword annotations corresponding to the k TCM keywords can be selected from the target TCM database to obtain k TCM keyword annotations, where each TCM keyword corresponds to a TCM keyword annotation. Specifically, the index of each TCM keyword in the k TCM keywords can be obtained to obtain k indexes, which can be searched in the target TCM database based on the k indexes to obtain k TCM keyword annotations; then, the k TCM keywords in the reference first reply content can be replaced by the k TCM keyword annotations to obtain a second reply content; then, the target reply form corresponding to the target age information can be determined, for example, a mapping relationship between preset age information and reply form can be pre-stored, and the target reply form corresponding to the target age information can be determined based on the mapping relationship. Finally, the second reply content can be adjusted according to the target reply form to obtain the first reply content. Specifically, the reply form of the second reply content can be changed to the target reply form to obtain the first reply content; the target reply form can include one of the following: knowledge explanation type, operation guidance type, comparative analysis type, etc., which are not limited here.

[0112] In this way, by customizing the response content based on the target person's personal information (including comprehension ability, education level, age, etc.), we avoid providing stereotyped, general but untargeted responses, thereby enabling the target person to acquire and apply traditional Chinese medicine knowledge more efficiently.

[0113] S206: Obtain first feedback content from the target object regarding the first reply content.

[0114] In the embodiment of the present application, the first feedback content is used to indicate whether the target object understands the first reply content.

[0115] In a specific embodiment, the first reply content can be displayed on the user terminal device, allowing the target object to read the first reply content, and then obtain the feedback content input by the target object on the user terminal device, that is, the first feedback content.

[0116] S207: Determine whether the target object meets a first preset condition according to the first feedback content.

[0117] In the embodiment of the present application, the first preset condition can be preset in advance or defaulted.

[0118] In a specific embodiment, it can be determined whether the target object understands the first reply content based on the first feedback content. If the target object understands the first reply content, it is determined that the target object meets the first preset condition.

[0119] If the target object does not understand the content of the first reply and still has questions, it is determined that the target object does not meet the first preset condition, and the first reply content is further explained through the target TCM AI model until the target object understands the reply content output by the target TCM AI model, that is, the target object meets the first preset condition.

[0120] S208. When the target object meets the first preset condition, generate corresponding medical cases based on the target medical problem to obtain a medical cases; a is an integer greater than 1.

[0121] In an embodiment of the present application, when the target object meets the first preset condition, the target question type corresponding to the target medical question is determined, and a corresponding medical case is generated based on the target question type to obtain a medical case. Specifically, medical cases corresponding to the target question type can be screened out from the preset case library to obtain a medical case.

[0122] S209: Obtain feedback content from the target object for each of the a medical cases to obtain a pieces of feedback content.

[0123] In an embodiment of the present application, a medical cases may be displayed on a user-end device to obtain the target subject's response content to each medical case, that is, a feedback content.

[0124] S210: Detect whether the a pieces of feedback content meet a second preset condition.

[0125] In the embodiment of the present application, the second preset condition may be preset in advance or default.

[0126] Optionally, step S210, detecting whether the a pieces of feedback content meet a second preset condition, may include the following steps:

[0127] D1. Obtain a reference answer for each of the a medical cases to obtain a reference answer;

[0128] D2. Determine semantic similarities between the a feedback contents and corresponding reference responses in the a reference responses, to obtain a semantic similarities;

[0129] D3. Determine the semantic similarities among the a semantic similarities that are greater than a second preset similarity threshold, to obtain b semantic similarities; b is a natural number less than or equal to a;

[0130] D4. Determine a second accuracy rate based on the b semantic similarities and the a semantic similarities;

[0131] D5. Determine a first average similarity corresponding to the b semantic similarities;

[0132] D6. Determine a second average similarity corresponding to ab semantic similarities among the a semantic similarities excluding the b semantic similarities;

[0133] D7. Adjust the second accuracy based on the first average similarity and the second average similarity to obtain a third accuracy;

[0134] D8. When the third accuracy rate is greater than the second preset accuracy rate, determining that the a pieces of feedback content meet the second preset condition;

[0135] D9. When the third accuracy rate is less than or equal to the second preset accuracy rate, determine that the a pieces of feedback content do not meet the second preset condition.

[0136] In the embodiment of the present application, the second preset accuracy rate can be preset or defaulted in advance.

[0137] In a specific embodiment, a reference answer for each of a medical cases can be obtained to obtain a reference answers. Specifically, a preset case library may include reference answers, and a reference answers corresponding to a medical cases are obtained from the preset case library. Then, the semantic similarities between a feedback content and the corresponding reference answers in the a reference answers can be calculated to obtain a semantic similarities. Specifically, the method for calculating a semantic similarities can be the same as the method for calculating the n semantic similarities mentioned above, which will not be repeated here. Then, a semantic similarity greater than a second preset similarity threshold can be found among the a semantic similarities to obtain b semantic similarities. Then, a second accuracy rate can be determined based on the b semantic similarities and the a semantic similarities, as follows:

[0138] Second accuracy = b / a × 100%;

[0139] According to the above formula, the second accuracy rate can be obtained. Then, the average value of the b semantic similarities can be calculated, that is, the first average similarity. Further, ab semantic similarities excluding the b semantic similarities among the a semantic similarities can be determined first, and then the average value of the ab semantic similarities can be calculated, that is, the second average similarity. Then, the second accuracy rate can be adjusted based on the first average similarity and the second average similarity to obtain a third accuracy rate. When the third accuracy rate is greater than the second preset accuracy rate, it is determined that the a feedback contents meet the second preset condition.

[0140] When the third accuracy rate is less than or equal to the second preset accuracy rate, it is determined that the a pieces of feedback content do not meet the second preset condition.

[0141] In this way, by adjusting the second accuracy rate based on the two average similarities to obtain the third accuracy rate, the limitations of relying solely on simple comparison or single dimension to judge the accuracy rate are avoided, thereby making the final third accuracy rate more realistic and accurate.

[0142] Optionally, step D7, adjusting the second accuracy based on the first average similarity and the second average similarity to obtain a third accuracy, may include the following steps:

[0143] E1. Determine a first similarity sequence based on the a semantic similarities;

[0144] E2. Determine the similarity difference between every two adjacent similarities in the first similarity sequence to obtain a-1 similarity differences;

[0145] E3. Determine the similarity difference values ​​greater than the first preset difference value among the a-1 similarity differences, to obtain c similarity differences; where c is a natural number less than or equal to a-1;

[0146] E4. Determine a ratio between the c similarity differences and ac-1 similarity differences that are not greater than a first preset difference value among the a-1 similarity differences to obtain a first ratio;

[0147] E5. When the first ratio is less than a preset ratio, determine a ratio between the first average similarity and the second average similarity to obtain a second ratio;

[0148] E6. Determine a difference between the second ratio and the first ratio to obtain a first difference;

[0149] E7. Determine a target adjustment factor corresponding to the first difference;

[0150] E8. Adjust the second accuracy rate according to the target adjustment factor to obtain the third accuracy rate.

[0151] In the embodiment of the present application, the preset ratio can be preset in advance or defaulted.

[0152] In a specific embodiment, the a semantic similarities can be sorted according to the generation time of the corresponding feedback content to obtain a first similarity sequence, where the earlier the generation time, the higher the order; then, the similarity difference between each two adjacent similarities in the first similarity sequence can be calculated to obtain a-1 similarity difference values; further, a similarity difference greater than a first preset difference value can be found among the a-1 similarity difference values ​​to obtain c similarity difference values; then, ac-1 similarity difference values ​​not greater than the first preset difference value can be obtained among the a-1 similarity difference values, and the ratio between the c similarity difference values ​​and the ac-1 similarity difference values ​​can be calculated. The specific calculation formula is as follows:

[0153] First ratio = c / (ac-1);

[0154] According to the above formula, the first ratio can be obtained; when the first ratio is less than the preset ratio, the ratio between the first average similarity and the second average similarity is determined. The specific calculation formula is as follows:

[0155] Second ratio = first average similarity / second average similarity;

[0156] According to the above formula, the second ratio can be obtained; the first ratio is subtracted from the second ratio to obtain the first difference; then, the target adjustment factor corresponding to the first difference can be determined. For example, a mapping relationship between a preset difference and an adjustment factor can be pre-set, and the target adjustment factor corresponding to the first difference can be determined based on the mapping relationship. The value range of the target adjustment factor can be -0.2 to 0.2; finally, the second accuracy can be adjusted according to the target adjustment factor. The specific calculation formula is as follows:

[0157] Third accuracy = second accuracy × (1 + target adjustment factor);

[0158] The third accuracy can be obtained according to the above formula.

[0159] In this way, by determining the c similarity differences among the a-1 similarity differences that are greater than the first preset difference, it is possible to accurately screen out the key points that cause significant fluctuations in semantic similarity. These significant change points often contain important information and may indicate that there are some issues that require special attention in the collection of feedback content, the setting of reference responses, or the understanding of corresponding medical knowledge. For example, if a large similarity difference suddenly appears in several consecutive medical cases, it may mean that the explanation of the knowledge points corresponding to these cases is not clear enough, resulting in a large deviation in the understanding of the feedback recipient, thus causing significant fluctuations in the feedback quality. By screening out these key differences, it is possible to quickly locate these "abnormal areas" where problems may exist, and thus re-teach the target subjects in these "abnormal areas" to improve the efficiency of traditional Chinese medicine teaching.

[0160] Optionally, when the first ratio is greater than the preset ratio, the method may include the following steps:

[0161] F1. Determine a difference between the first ratio and the preset ratio to obtain a second difference;

[0162] F2. When the second difference is greater than a second preset difference, determining that the a pieces of feedback content do not meet the second preset condition;

[0163] F3. When the second difference is not greater than a second preset difference, determine the feedback contents corresponding to the a feedback contents by the c similarity differences, to obtain d feedback contents; d is an integer greater than c and less than or equal to a;

[0164] F4. Generate new medical cases based on the target medical problem, obtaining e medical cases; e equals d;

[0165] F5. Obtain feedback content from the target subject for each of the e medical cases to obtain e pieces of feedback content;

[0166] F6. Replace the d pieces of feedback content in the a pieces of feedback content with the e pieces of feedback content to obtain updated a pieces of feedback content;

[0167] F7. For the updated a pieces of feedback content, executing the step of detecting whether the a pieces of feedback content meet a second preset condition.

[0168] In the embodiment of the present application, the second preset difference can be preset in advance or defaulted.

[0169] In a specific embodiment, the first ratio can be subtracted from the preset ratio to obtain a second difference; when the second difference is greater than the second preset difference, it means that the semantic similarity of the a feedback contents generated by the target object fluctuates too much, is not stable, and has an abnormality. Therefore, it can be determined that the a feedback contents do not meet the second preset condition.

[0170] When the second difference is not greater than the second preset difference, the feedback content corresponding to the a feedback content with c similarity differences can be determined to obtain d feedback content; then, new medical cases can be generated according to the target medical problem to obtain e medical cases. Specifically, the method for generating e medical cases can be the same as the above-mentioned method for generating a medical case, which will not be repeated here; it needs to be explained that each medical case in the e medical cases is different from the medical case in the a medical case.

[0171] Furthermore, the feedback content of the target object for each of the e medical cases can be obtained to obtain e feedback contents; the e feedback contents are used to replace d feedback contents in the a feedback contents to obtain updated a feedback contents; then, the step of detecting whether the a feedback contents meet the second preset condition can be executed on the updated a feedback contents.

[0172] In this way, by generating new medical cases and updating feedback content around the target medical problem, the characteristics of different medical problems and the specific situation of the target subjects in mastering the corresponding knowledge are fully considered. For example, for questions about TCM disease diagnosis, if it is found that the similarity of the feedback content in certain disease cases fluctuates greatly, more new cases with similar diseases but different symptom details or varying levels of complexity will be generated to guide the target subjects to rethink and provide feedback. In this way, the weak links in the target subjects' knowledge of disease diagnosis can be more accurately identified, and the explanation and training of relevant knowledge can be strengthened in a targeted manner, thereby achieving personalized medical knowledge transfer and ability training, and improving the overall teaching and learning effect.

[0173] S211. When the a pieces of feedback content meet the second preset condition, determine that the target subject has mastered the traditional Chinese medicine knowledge corresponding to the target medical problem.

[0174] In the embodiment of the present application, when a pieces of feedback content meet the second preset condition, it indicates that the target object has mastered the traditional Chinese medicine knowledge related to the target medical problem.

[0175] The implementation of this application has the following beneficial effects:

[0176] It can be seen that the AI ​​teaching method based on traditional Chinese medicine culture described in this application obtains the first reply content by inputting the target object's personal information together with the target medical question into the model. The addition of personal information enables the reply to fully consider the characteristics of the target object, such as age, cultural background, knowledge base and other factors, thereby making personalized replies to help the target object better understand traditional Chinese medicine knowledge, reduce the time for understanding, and thus improve the efficiency of traditional Chinese medicine teaching.

[0177] See also Figure 3 , Figure 3 This is a functional unit block diagram of an AI teaching device 300 based on TCM culture provided in an embodiment of the present application. The AI ​​teaching device 300 based on TCM culture includes: an acquisition unit 301, a model building unit 302, and a teaching unit 303; wherein:

[0178] The acquisition unit 301 is configured to acquire a target TCM course for a target subject and a target teaching plan for the target TCM course; the target TCM course includes one of the following: a Zhangshu TCM course, a Chinese medicine course, a TCM basic theory course, or an acupuncture and massage course; and acquire a target TCM database corresponding to the target TCM course.

[0179] The model building unit 302 is used to build a TCM AI model based on the target TCM database and the target teaching plan to obtain a target TCM AI model;

[0180] The acquisition unit 301 is further configured to acquire target medical questions for the target TCM course and personal information of the target subject;

[0181] The teaching unit 303 is used to input the target medical question and the personal information into the target traditional Chinese medicine AI model to obtain a first reply content; obtain the first feedback content of the target object for the first reply content; determine whether the target object meets the first preset condition based on the first feedback content; when the target object meets the first preset condition, generate a corresponding medical case based on the target medical question to obtain a medical case; a is an integer greater than 1; obtain the feedback content of the target object for each of the a medical cases to obtain a feedback content; detect whether the a feedback contents meet the second preset condition; when the a feedback contents meet the second preset condition, determine that the target object has mastered the traditional Chinese medicine knowledge corresponding to the target medical question.

[0182] Optionally, in the aspect of constructing a TCM AI model according to the target TCM database and the target teaching plan to obtain a target TCM AI model, the model construction unit 302 is specifically configured to:

[0183] determining target data characteristics of the TCM data in the target TCM database;

[0184] Select an AI model corresponding to the characteristics of the target data from a preset AI model library to obtain an initial AI model;

[0185] selecting a first data set from the target traditional Chinese medicine database according to the target teaching plan;

[0186] Obtaining the target learning progress of the target subject for the target TCM course;

[0187] Dividing the first data set into a second data set and a third data set based on the target learning progress; the second data set is the TCM knowledge that the target subject has already learned, and the third data set is the TCM knowledge that the target subject has not yet learned;

[0188] Dividing the second data set and the third data set into a training set and a test set according to a preset ratio;

[0189] Training the initial AI model according to the training set to obtain a first AI model;

[0190] Testing the first AI model using the test set to obtain a test result, and determining a first accuracy rate based on the test result;

[0191] When the first accuracy rate is greater than a first preset accuracy rate, the first AI model is used as the target TCM AI model.

[0192] Optionally, in inputting the target medical question and the personal information into the target TCM AI model to obtain the first answer content, the teaching unit 303 is specifically configured to:

[0193] Extracting keywords from the target medical question to obtain m keywords; the m keywords are used to indicate the inquiry intention of the target medical question; m is a positive integer;

[0194] Determine a synonym set corresponding to each of the m keywords in the target TCM database to obtain m synonym sets; each synonym set includes a plurality of TCM keywords;

[0195] Replacing the m keywords based on the m synonym sets to obtain m TCM keywords;

[0196] Determining the inquiry intention of the target medical question based on the m TCM keywords to obtain the target inquiry intention;

[0197] Generate a response content corresponding to the target inquiry intention to obtain a reference first response content;

[0198] The reference first reply content is adjusted according to the personal information to obtain the first reply content.

[0199] Optionally, in the aspect of replacing the m keywords based on the m synonym sets to obtain m TCM keywords, the teaching unit 303 is specifically configured to:

[0200] Obtain a target synonym set and a target keyword corresponding to the target synonym set; the target synonym set is any one of the m synonym sets; the target synonym set includes n traditional Chinese medicine keywords; n is a positive integer;

[0201] Calculating the semantic similarity between each TCM keyword in the target synonym set and the target keyword to obtain n semantic similarities;

[0202] Determine the semantic similarities greater than a first preset similarity threshold among the n semantic similarities to obtain i semantic similarities, where i is a positive integer less than or equal to n;

[0203] Determine the TCM keywords corresponding to the i semantic similarities in the target synonym set to obtain i TCM keywords;

[0204] Determining a TCM keyword associated with the target medical problem among the i TCM keywords to obtain at least one TCM keyword;

[0205] A TCM keyword with the greatest semantic similarity among the at least one TCM keyword is determined according to the i semantic similarities, and the TCM keyword corresponding to the target synonym set is obtained.

[0206] Optionally, the personal information includes target educational background information and target age information. In adjusting the reference first reply content according to the personal information to obtain the first reply content, the teaching unit 303 is specifically configured to:

[0207] Extracting TCM keywords from the reference first reply content to obtain j TCM keywords; j is a positive integer;

[0208] determining a comprehension difficulty parameter for each of the j TCM keywords based on the personal information, to obtain j comprehension difficulty parameters;

[0209] determining a first comprehension difficulty parameter according to the j comprehension difficulty parameters;

[0210] Determining a second comprehension difficulty parameter acceptable to the target subject based on the target educational background information;

[0211] When the first comprehension difficulty parameter is greater than the second comprehension difficulty parameter, determining a difference between the first comprehension difficulty parameter and the second comprehension difficulty parameter to obtain a target comprehension difficulty difference;

[0212] Selecting k comprehension difficulty parameters from the j comprehension difficulty parameters based on the target comprehension difficulty difference; the sum of the k comprehension difficulty parameters is greater than or equal to the target comprehension difficulty difference; k is a positive integer less than or equal to j;

[0213] Determine the TCM keywords corresponding to the k comprehension difficulty parameters among the j TCM keywords to obtain k TCM keywords;

[0214] Selecting TCM keyword annotations corresponding to the k TCM keywords from the target TCM database to obtain k TCM keyword annotations;

[0215] Replacing the k TCM keywords in the reference first reply content with the k TCM keyword annotations to obtain a second reply content;

[0216] determining a target response form corresponding to the target age information;

[0217] The second reply content is adjusted according to the target reply form to obtain the first reply content.

[0218] Optionally, in terms of detecting whether the a pieces of feedback content meet a second preset condition, the teaching unit 303 is specifically configured to:

[0219] Obtain a reference answer for each of the a medical cases to obtain a reference answers;

[0220] Determining semantic similarities between the a feedback contents and corresponding reference responses in the a reference responses to obtain a semantic similarities;

[0221] Determine the semantic similarities greater than a second preset similarity threshold among the a semantic similarities, to obtain b semantic similarities; b is a natural number less than or equal to a;

[0222] Determining a second accuracy rate based on the b semantic similarities and the a semantic similarities;

[0223] Determining a first average similarity corresponding to the b semantic similarities;

[0224] Determine a second average similarity corresponding to ab semantic similarities excluding the b semantic similarities among the a semantic similarities;

[0225] Adjusting the second accuracy rate based on the first average similarity and the second average similarity to obtain a third accuracy rate;

[0226] When the third accuracy rate is greater than the second preset accuracy rate, determining that the a pieces of feedback content meet the second preset condition;

[0227] When the third accuracy rate is less than or equal to the second preset accuracy rate, it is determined that the a pieces of feedback content do not meet the second preset condition.

[0228] Optionally, in the aspect of adjusting the second accuracy rate based on the first average similarity and the second average similarity to obtain a third accuracy rate, the teaching unit 303 is specifically configured to:

[0229] Determine a first similarity sequence according to the a semantic similarities;

[0230] Determine a similarity difference between every two adjacent similarities in the first similarity sequence to obtain a-1 similarity differences;

[0231] Determine a similarity difference value greater than a first preset difference value among the a-1 similarity differences, to obtain c similarity differences; c is a natural number less than or equal to a-1;

[0232] Determine a ratio between the c similarity differences and ac-1 similarity differences that are not greater than a first preset difference value among the a-1 similarity differences to obtain a first ratio;

[0233] When the first ratio is less than a preset ratio, determining a ratio between the first average similarity and the second average similarity to obtain a second ratio;

[0234] determining a difference between the second ratio and the first ratio to obtain a first difference;

[0235] determining a target adjustment factor corresponding to the first difference;

[0236] The second accuracy rate is adjusted according to the target adjustment factor to obtain the third accuracy rate.

[0237] Optionally, when the first ratio is greater than the preset ratio, the AI ​​teaching device 300 based on TCM culture is specifically configured to:

[0238] determining a difference between the first ratio and the preset ratio to obtain a second difference;

[0239] When the second difference is greater than a second preset difference, determining that the a feedback contents do not meet the second preset condition;

[0240] When the second difference is not greater than the second preset difference, determining the feedback contents corresponding to the a feedback contents by the c similarity differences, to obtain d feedback contents; d is an integer greater than c and less than or equal to a;

[0241] Generate new medical cases based on the target medical problem, and obtain e medical cases; e equals d;

[0242] Obtaining feedback content from the target subject for each of the e medical cases to obtain e pieces of feedback content;

[0243] Replacing the d pieces of feedback content in the a pieces of feedback content with the e pieces of feedback content to obtain updated a pieces of feedback content;

[0244] For the updated a pieces of feedback content, the step of detecting whether the a pieces of feedback content meet a second preset condition is performed.

[0245] In a specific implementation, the AI ​​teaching device 300 based on TCM culture described in the embodiment of the present invention can also execute other implementation methods described in the AI ​​teaching method based on TCM culture provided in the above embodiment of the present invention, which will not be repeated here.

[0246] See Figure 4 , Figure 4This is a structural diagram of an electronic device provided in an embodiment of the present application, which includes a processor, a memory, a communication interface and one or more programs. The processor, memory and communication interface are interconnected through a bus. The one or more programs are stored in the memory and are configured to be executed by the processor. The one or more programs include instructions for executing other implementation methods described in the AI ​​teaching method based on traditional Chinese medicine culture provided in the above embodiment of the present invention, which will not be repeated here.

[0247] An embodiment of the present invention also provides a computer storage medium, wherein the computer storage medium stores a computer program for electronic data exchange, and the computer program enables a computer to execute part or all of the steps of any method described in the above method embodiment, and the above computer includes an electronic device.

[0248] The present application also provides a computer program product comprising a non-transitory computer-readable storage medium storing a computer program, wherein the computer program is operable to cause a computer to perform some or all of the steps of any of the methods described in the above method embodiments. The computer program product may be a software installation package, and the computer may comprise an electronic device.

[0249] It should be noted that for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that this application is not limited by the order of the actions described, because according to this application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by this application.

[0250] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0251] In the several embodiments provided in this application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the above-mentioned units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0252] The units described above as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0253] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0254] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the existing technology, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a memory and includes a number of instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the above-mentioned methods in each embodiment of the present application. The aforementioned memory includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, etc. Various media that can store program code.

[0255] The above is a detailed introduction to the embodiments of the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core idea of ​​the present application. At the same time, for those skilled in the art, according to the idea of ​​the present application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. An AI teaching method based on Chinese medicine culture, characterized by: The method comprises: Obtaining a target TCM course for a target subject and a target teaching plan for the target TCM course; the target TCM course includes one of the following: a Zhangshu TCM course, a Chinese medicine course, a TCM basic theory course, or an acupuncture and massage course; Obtaining a target TCM database corresponding to the target TCM course; Constructing a TCM AI model according to the target TCM database and the target teaching plan to obtain a target TCM AI model; Obtaining target medical issues for the target TCM course and personal information of the target subject; Inputting the target medical question and the personal information into the target TCM AI model to obtain a first response; Obtaining first feedback content from the target object regarding the first reply content; determining whether the target object meets a first preset condition according to the first feedback content; When the target object meets the first preset condition, a corresponding medical case is generated based on the target medical problem to obtain a medical case; a is an integer greater than 1; Obtaining feedback content from the target subject for each of the a medical cases to obtain a pieces of feedback content; Detecting whether the a feedback contents meet a second preset condition; When the a pieces of feedback content meet the second preset condition, determining that the target subject has mastered the traditional Chinese medicine knowledge corresponding to the target medical problem; The step of inputting the target medical question and the personal information into the target TCM AI model to obtain a first response includes: Extracting keywords from the target medical question to obtain m keywords; the m keywords are used to indicate the inquiry intention of the target medical question; m is a positive integer; Determine a synonym set corresponding to each of the m keywords in the target TCM database to obtain m synonym sets; each synonym set includes a plurality of TCM keywords; Replacing the m keywords based on the m synonym sets to obtain m TCM keywords; Determining the inquiry intention of the target medical question based on the m TCM keywords to obtain the target inquiry intention; Generate a response content corresponding to the target inquiry intention to obtain a reference first response content; Adjusting the reference first reply content according to the personal information to obtain the first reply content; The personal information includes: target educational information and target age information; and the reference first reply content is adjusted according to the personal information to obtain the first reply content, including: Extracting TCM keywords from the reference first reply content to obtain j TCM keywords; j is a positive integer; determining a comprehension difficulty parameter for each of the j TCM keywords based on the personal information, to obtain j comprehension difficulty parameters; determining a first comprehension difficulty parameter according to the j comprehension difficulty parameters; Determine a second comprehension difficulty parameter acceptable to the target subject based on the target educational background information; specifically, pre-store a mapping relationship between preset educational background information and comprehension difficulty parameters, and determine the second comprehension difficulty parameter corresponding to the target educational background information based on the mapping relationship; When the first comprehension difficulty parameter is greater than the second comprehension difficulty parameter, determining a difference between the first comprehension difficulty parameter and the second comprehension difficulty parameter to obtain a target comprehension difficulty difference; Selecting k comprehension difficulty parameters from the j comprehension difficulty parameters based on the target comprehension difficulty difference; the sum of the k comprehension difficulty parameters is greater than or equal to the target comprehension difficulty difference; k is a positive integer less than or equal to j; Determine the TCM keywords corresponding to the k comprehension difficulty parameters among the j TCM keywords to obtain k TCM keywords; Selecting TCM keyword annotations corresponding to the k TCM keywords from the target TCM database to obtain k TCM keyword annotations; Replacing the k TCM keywords in the reference first reply content with the k TCM keyword annotations to obtain a second reply content; Determining a target response format corresponding to the target age information; the target response format includes one of the following: knowledge explanation type, operation guidance type, and comparative analysis type; The second reply content is adjusted according to the target reply form to obtain the first reply content.

2. The method according to claim 1, wherein The step of constructing a TCM AI model based on the target TCM database and the target teaching plan to obtain the target TCM AI model includes: determining target data characteristics of the TCM data in the target TCM database; Select an AI model corresponding to the characteristics of the target data from a preset AI model library to obtain an initial AI model; selecting a first data set from the target traditional Chinese medicine database according to the target teaching plan; Obtaining the target learning progress of the target subject for the target TCM course; Dividing the first data set into a second data set and a third data set based on the target learning progress; the second data set is the TCM knowledge that the target subject has already learned, and the third data set is the TCM knowledge that the target subject has not yet learned; Dividing the second data set and the third data set into a training set and a test set according to a preset ratio; Training the initial AI model according to the training set to obtain a first AI model; Testing the first AI model using the test set to obtain a test result, and determining a first accuracy rate based on the test result; When the first accuracy rate is greater than a first preset accuracy rate, the first AI model is used as the target TCM AI model.

3. The method according to claim 1, wherein The m keywords are replaced based on the m synonym sets to obtain m TCM keywords, including: Obtain a target synonym set and a target keyword corresponding to the target synonym set; the target synonym set is any one of the m synonym sets; the target synonym set includes n traditional Chinese medicine keywords; n is a positive integer; Calculating the semantic similarity between each TCM keyword in the target synonym set and the target keyword to obtain n semantic similarities; Determine the semantic similarities greater than a first preset similarity threshold among the n semantic similarities to obtain i semantic similarities, where i is a positive integer less than or equal to n; Determine the TCM keywords corresponding to the i semantic similarities in the target synonym set to obtain i TCM keywords; Determining a TCM keyword associated with the target medical problem among the i TCM keywords to obtain at least one TCM keyword; A TCM keyword with the greatest semantic similarity among the at least one TCM keyword is determined according to the i semantic similarities, and the TCM keyword corresponding to the target synonym set is obtained.

4. The method according to any one of claims 1 to 3, wherein The detecting whether the a feedback contents meet the second preset condition includes: Obtain a reference answer for each of the a medical cases to obtain a reference answers; Determining semantic similarities between the a feedback contents and corresponding reference responses in the a reference responses to obtain a semantic similarities; Determine the semantic similarities greater than a second preset similarity threshold among the a semantic similarities, to obtain b semantic similarities; b is a natural number less than or equal to a; Determining a second accuracy rate based on the b semantic similarities and the a semantic similarities; Determining a first average similarity corresponding to the b semantic similarities; Determine a second average similarity corresponding to ab semantic similarities excluding the b semantic similarities among the a semantic similarities; Adjusting the second accuracy rate based on the first average similarity and the second average similarity to obtain a third accuracy rate; When the third accuracy rate is greater than the second preset accuracy rate, determining that the a pieces of feedback content meet the second preset condition; When the third accuracy rate is less than or equal to the second preset accuracy rate, it is determined that the a pieces of feedback content do not meet the second preset condition.

5. The method according to claim 4, wherein The adjusting the second accuracy rate based on the first average similarity and the second average similarity to obtain a third accuracy rate includes: Determine a first similarity sequence according to the a semantic similarities; Determine a similarity difference between every two adjacent similarities in the first similarity sequence to obtain a-1 similarity differences; Determine a similarity difference value greater than a first preset difference value among the a-1 similarity differences, to obtain c similarity differences; c is a natural number less than or equal to a-1; Determine a ratio between the c similarity differences and ac-1 similarity differences that are not greater than a first preset difference value among the a-1 similarity differences to obtain a first ratio; When the first ratio is less than a preset ratio, determining a ratio between the first average similarity and the second average similarity to obtain a second ratio; determining a difference between the second ratio and the first ratio to obtain a first difference; determining a target adjustment factor corresponding to the first difference; The second accuracy rate is adjusted according to the target adjustment factor to obtain the third accuracy rate.

6. The method according to claim 5, wherein When the first ratio is greater than the preset ratio, the method includes: determining a difference between the first ratio and the preset ratio to obtain a second difference; When the second difference is greater than a second preset difference, determining that the a feedback contents do not meet the second preset condition; When the second difference is not greater than the second preset difference, determining the feedback contents corresponding to the a feedback contents by the c similarity differences, to obtain d feedback contents; d is an integer greater than c and less than or equal to a; Generate new medical cases based on the target medical problem, and obtain e medical cases; e equals d; Obtaining feedback content from the target subject for each of the e medical cases to obtain e pieces of feedback content; Replacing the d pieces of feedback content in the a pieces of feedback content with the e pieces of feedback content to obtain updated a pieces of feedback content; For the updated a pieces of feedback content, the step of detecting whether the a pieces of feedback content meet a second preset condition is performed.

7. An AI teaching device based on Chinese medicine culture, characterized in that: The device comprises: an acquisition unit, a model building unit, and a teaching unit; wherein: The acquisition unit is configured to acquire a target TCM course for a target subject and a target teaching plan for the target TCM course; the target TCM course includes one of the following: a Zhangshu TCM course, a Chinese medicine course, a TCM basic theory course, or an acupuncture and massage course; and acquire a target TCM database corresponding to the target TCM course; The model building unit is used to build a TCM AI model according to the target TCM database and the target teaching plan to obtain a target TCM AI model; The acquisition unit is further configured to acquire target medical questions for the target TCM course and personal information of the target subject; The teaching unit is used to input the target medical question and the personal information into the target TCM AI model to obtain a first reply content; obtain the first feedback content of the target subject for the first reply content; determine whether the target subject meets the first preset condition based on the first feedback content; when the target subject meets the first preset condition, generate a corresponding medical case based on the target medical question to obtain a medical case; a is an integer greater than 1; obtain the feedback content of the target subject for each of the a medical cases to obtain a feedback content; detect whether the a feedback content meets the second preset condition; when the a feedback content meets the second preset condition, determine that the target subject has mastered the TCM knowledge corresponding to the target medical question; In the aspect of inputting the target medical question and the personal information into the target TCM AI model to obtain the first answer content, the teaching unit is specifically used to: Extracting keywords from the target medical question to obtain m keywords; the m keywords are used to indicate the inquiry intention of the target medical question; m is a positive integer; Determine a synonym set corresponding to each of the m keywords in the target TCM database to obtain m synonym sets; each synonym set includes a plurality of TCM keywords; Replacing the m keywords based on the m synonym sets to obtain m TCM keywords; Determining the inquiry intention of the target medical question based on the m TCM keywords to obtain the target inquiry intention; Generate a response content corresponding to the target inquiry intention to obtain a reference first response content; Adjusting the reference first reply content according to the personal information to obtain the first reply content; The personal information includes target educational background information and target age information. In adjusting the reference first reply content based on the personal information to obtain the first reply content, the teaching unit is specifically configured to: Extracting TCM keywords from the reference first reply content to obtain j TCM keywords; j is a positive integer; determining a comprehension difficulty parameter for each of the j TCM keywords based on the personal information, to obtain j comprehension difficulty parameters; determining a first comprehension difficulty parameter according to the j comprehension difficulty parameters; Determine a second comprehension difficulty parameter acceptable to the target subject based on the target educational background information; specifically, pre-store a mapping relationship between preset educational background information and comprehension difficulty parameters, and determine the second comprehension difficulty parameter corresponding to the target educational background information based on the mapping relationship; When the first comprehension difficulty parameter is greater than the second comprehension difficulty parameter, determining a difference between the first comprehension difficulty parameter and the second comprehension difficulty parameter to obtain a target comprehension difficulty difference; Selecting k comprehension difficulty parameters from the j comprehension difficulty parameters based on the target comprehension difficulty difference; the sum of the k comprehension difficulty parameters is greater than or equal to the target comprehension difficulty difference; k is a positive integer less than or equal to j; Determine the TCM keywords corresponding to the k comprehension difficulty parameters among the j TCM keywords to obtain k TCM keywords; Selecting TCM keyword annotations corresponding to the k TCM keywords from the target TCM database to obtain k TCM keyword annotations; Replacing the k TCM keywords in the reference first reply content with the k TCM keyword annotations to obtain a second reply content; Determining a target response format corresponding to the target age information; the target response format includes one of the following: knowledge explanation type, operation guidance type, and comparative analysis type; The second reply content is adjusted according to the target reply form to obtain the first reply content.

8. A computer-readable storage medium, characterized in that A computer program for electronic data exchange is stored, wherein the computer program enables a computer to execute the method according to any one of claims 1 to 6.

Images