CN118053557A

CN118053557A - Traditional Chinese medicine knowledge science popularization system utilizing artificial intelligence and big data analysis

Info

Publication number: CN118053557A
Application number: CN202410216015.5A
Authority: CN
Inventors: 郝锦杰; 赵珂慧; 石磊; 黄丝语; 何成伟; 李泽英; 朱红; 吉祖华; 高立昭; 张连发; 崔建群
Original assignee: Central China Normal University
Current assignee: Central China Normal University
Priority date: 2024-02-27
Filing date: 2024-02-27
Publication date: 2024-05-17

Abstract

The invention discloses a traditional Chinese medicine knowledge science popularization system utilizing artificial intelligence and big data analysis in the technical field of knowledge science popularization, which comprises a chaos model construction module, a knowledge granularity adjustment module, a composite knowledge reasoning module, an effect prediction module, a nonlinear learning path generation module, a dynamic graph network construction module, a time sequence analysis module and a knowledge updating mechanism module. According to the invention, the chaotic model construction module is introduced, so that the system can accurately analyze the nonlinearity and dynamic change of the traditional Chinese medicine data, and deep dig the complex relationship between the treatment effect and the physique response, thereby greatly improving the depth and breadth of data analysis. The application of the composite knowledge reasoning module combines the symbolic logic and the neural network technology, improves the capability of the system for processing complex medical concepts and unstructured data, and provides accurate knowledge analysis for users. The effect prediction module improves the accuracy of the treatment method and the drug formula effect prediction through a machine learning algorithm.

Description

Traditional Chinese medicine knowledge science popularization system utilizing artificial intelligence and big data analysis

Technical Field

The invention relates to the technical field of knowledge science popularization, in particular to a traditional Chinese medicine knowledge science popularization system utilizing artificial intelligence and big data analysis.

Background

The technical field of knowledge science popularization is a field which is focused on education and spreading professional knowledge to wide non-professional groups and aims to improve the understanding and interest of the public in specific disciplines or technical fields. Such systems are particularly important in the field of traditional Chinese medicine, which contains a rich set of theories, methods of treatment and drug knowledge, which are often difficult for people without a professional background to understand. Through science popularization, the method can help people to better know and understand the value and application of the traditional Chinese medicine and promote the inheritance and development of the traditional Chinese medicine culture.

The science popularization system of traditional Chinese medicine knowledge is a platform or application aiming at popularizing the traditional Chinese medicine knowledge, and mainly aims at transmitting information such as the basic principle of traditional Chinese medicine, a treatment method, medicine use and the like to the public in a scientific and easily understood mode. Such systems are intended to eliminate public misunderstanding and prejudice to the medicine, to improve people's health consciousness and self-health ability, and to pave the way for modernization and internationalization of the traditional Chinese medicine. By improving the public knowledge of traditional Chinese medicine, the system is expected to promote the acceptance of traditional Chinese medicine and support the popularization of the traditional Chinese medicine in the global scope.

Although the prior art has achieved a certain effect in the popularization and understanding promotion of the knowledge of traditional Chinese medicine, the prior art has significant defects in the aspect of deep analysis of nonlinear and dynamic change modes of big data of traditional Chinese medicine. The existing system can not fully utilize mathematical models such as chaos theory and the like to deeply explore complex relations between treatment effects and physique reactions, and limits deeper understanding and application of the traditional Chinese medicine treatment principle. In addition, the prior art has to be improved in the capability of integrating symbolic logic and neural network reasoning to process complex medical concepts, which affects the processing capability of the system to complex medical concepts and the depth of providing solutions. In the aspect of constructing a traditional Chinese medicine effect prediction model based on big data, the method is limited and lacks of efficient data processing and model training technology, so that the accuracy and practicability of the model prediction effect are weakened. For the generation mechanism of the nonlinear learning path, the prior art can not fully exert the potential of graph theory in the aspects of designing a knowledge structure and recommending the learning path, and influence the systematicness and logic of knowledge transfer. In the application aspect of dynamic graph network and time sequence analysis technology, the capability of the existing system in capturing knowledge development trend and predicting future change is insufficient, the timeliness and accuracy of updating science popularization information are limited, and the long-term effectiveness and the foresight of knowledge content of the traditional Chinese medicine knowledge science popularization system are further affected.

Based on the above, the invention designs a traditional Chinese medicine knowledge science popularization system utilizing artificial intelligence and big data analysis to solve the problems.

Disclosure of Invention

The invention aims to provide a traditional Chinese medicine knowledge science popularization system utilizing artificial intelligence and big data analysis, which aims to solve the problems that the prior art has certain effects on popularization and understanding promotion of traditional Chinese medicine knowledge, but has obvious defects on nonlinear and dynamic change modes of deep analysis of big data of traditional Chinese medicine. The existing system can not fully utilize mathematical models such as chaos theory and the like to deeply explore complex relations between treatment effects and physique reactions, and limits deeper understanding and application of the traditional Chinese medicine treatment principle. In addition, the prior art has to be improved in the capability of integrating symbolic logic and neural network reasoning to process complex medical concepts, which affects the processing capability of the system to complex medical concepts and the depth of providing solutions. In the aspect of constructing a traditional Chinese medicine effect prediction model based on big data, the method is limited and lacks of efficient data processing and model training technology, so that the accuracy and practicability of the model prediction effect are weakened. For the generation mechanism of the nonlinear learning path, the prior art can not fully exert the potential of graph theory in the aspects of designing a knowledge structure and recommending the learning path, and influence the systematicness and logic of knowledge transfer. In the aspect of application of dynamic graph network and time sequence analysis technology, the capability of the existing system in capturing knowledge development trend and predicting future change is insufficient, the timeliness and accuracy of updating science popularization information are limited, and the long-term effectiveness of the traditional Chinese medicine knowledge science popularization system and the prospective problem of knowledge content are further affected.

In order to achieve the above purpose, the present invention provides the following technical solutions: a traditional Chinese medicine knowledge science popularization system utilizing artificial intelligence and big data analysis comprises a chaotic model construction module, a knowledge granularity adjustment module, a composite knowledge reasoning module, an effect prediction module, a nonlinear learning path generation module, a dynamic graph network construction module, a time sequence analysis module and a knowledge updating mechanism module;

The chaotic model construction module analyzes the dynamic property of traditional Chinese medicine data by adopting a chaotic dynamics theory based on historical case data, extracts key time characteristics by a time sequence analysis method, maps a nonlinear structure of the data by utilizing a phase space reconstruction technology, and calculates and confirms the chaotic characteristic of a system by Lyapunov indexes to generate a chaotic dynamics model;

The knowledge granularity adjustment module is used for collecting user feedback and learning behavior data by adopting a user modeling technology based on a chaotic dynamics model, analyzing learning preference and cognition level of a user by a content adaptation algorithm, evaluating difficulty of knowledge content by utilizing a cognition load theory, adjusting knowledge granularity to match user requirements, and generating dynamically adjusted knowledge content;

The compound knowledge reasoning module adopts symbol logic to process the determined logic relation and rule based on the dynamically adjusted knowledge content, learns the mode of fuzzy and unstructured data through a deep neural network, integrates the reasoning result by utilizing a hybrid reasoning mechanism, analyzes the medical concept and generates a compound reasoning result;

the effect prediction module analyzes the structured data by adopting a random forest algorithm based on the composite reasoning result, processes the nonlinear relation in the multidimensional feature space by using a support vector machine, and generates an effect prediction analysis result by cross-verifying optimized model parameters;

The nonlinear learning path generation module integrates knowledge points and relations thereof in the traditional Chinese medicine field by adopting a knowledge graph construction technology based on effect prediction analysis results, designs a learning path through a shortest path and a community mining algorithm, recommends learning contents for a user by referring to user interests and knowledge point importance, and generates a nonlinear learning path;

The dynamic graph network construction module captures the evolution relation among the knowledge points of the traditional Chinese medicine by adopting a dynamic network model based on a nonlinear learning path, tracks the variation trend of the knowledge points by time sequence analysis, simulates the development of a knowledge network by using a network evolution algorithm, and generates a dynamic graph network;

The time sequence analysis module is used for analyzing the traditional Chinese medicine knowledge and the change of the application of the traditional Chinese medicine knowledge along with time by adopting a time sequence analysis method based on a dynamic graph network, predicting the development direction of the knowledge by an autoregressive moving average model and a trend analysis technology, and generating a time sequence analysis result;

The knowledge updating mechanism module adopts a content management tool to maintain a knowledge base based on a time sequence analysis result, periodically introduces new knowledge points and updating information through an incremental updating strategy and version control technology, ensures the accuracy and reliability of updated content by utilizing a quality control flow, and generates an updated traditional Chinese medicine knowledge base.

Preferably, the chaotic dynamics model comprises stability parameters, characteristic quantities of chaotic attractors and key dynamic behavior indexes, the dynamically adjusted knowledge content comprises knowledge units with multiple difficulty levels, learning materials adjusted according to cognitive load and knowledge points matched with user preferences, the composite reasoning result comprises a logic expression of a traditional Chinese medicine rule, mode characteristics of deep learning identification and knowledge conclusion obtained through comprehensive reasoning, the effect prediction analysis result comprises treatment effect scoring of symptoms, confidence intervals of a prediction model and priority ordering of treatment suggestions, the nonlinear learning path comprises learning target setting, links of recommended learning resources and learning sequences of the key knowledge points, the dynamic graph network comprises a dependency relation graph among the knowledge points, knowledge evolution tracks in time dimension and evolution analysis of key nodes, the time sequence analysis result comprises time sequence prediction of change of the knowledge points, key time points of trend change and potential and new knowledge mining areas, and the updated traditional Chinese medicine knowledge base comprises new treatment method items, revised medicine information and verified knowledge update lists.

Preferably, the chaotic model construction module comprises a chaotic theory analysis sub-module, a dynamics model construction sub-module and an effect change rule analysis sub-module;

The chaos theory analysis submodule carries out preliminary analysis of chaos theory based on historical case data, calculates standard deviation and average value of data by using numpy and scipy libraries in Python, estimates initial chaos indexes, draws a time sequence chart by using matplotlib, monitors data volatility, determines whether chaos analysis is carried out or not through the volatility, and generates a chaos analysis result;

the dynamics model construction submodule adopts a dynamics modeling method based on chaos analysis results, uses numpy libraries in Python to reconstruct phase space, sets embedding dimension as 3 and delay time as 2, constructs a phase space image of dynamic behavior, analyzes dynamic behavior and stability of data, and generates a dynamics model;

The effect change rule analysis submodule is based on a dynamics model, adopts time sequence analysis, uses pandas and statsmodels libraries of Python to execute an autoregressive moving average model, automatically determines p values and q values through AIC criteria in parameter selection, analyzes dynamic relations between traditional Chinese medicine treatment effects and physique reactions, and generates a chaotic dynamics model.

Preferably, the knowledge granularity adjustment module comprises a user model construction sub-module, a knowledge base structural design sub-module and a dynamic granularity adjustment machine sub-module;

The user model building submodule is used for carrying out execution through scikit-learn library of Python by adopting a decision tree algorithm based on user feedback and learning behavior data collected by the chaotic dynamics model, setting parameters of a decision tree, including maximum depth of 5 and minimum segmentation sample number of 10, classifying user data, analyzing learning preference and cognition level of a user, and generating a user characteristic model;

The knowledge base structural design submodule applies a K-mean value clustering algorithm based on a user characteristic model, and operates by utilizing scikit-learn library of Python again, wherein the core parameters of the clustering algorithm are set to be the optimal clustering number selected automatically, and the knowledge base is reorganized based on the user characteristic to generate a knowledge base reorganization scheme;

The dynamic granularity adjustment machine submodule adopts a cognitive load evaluation technology based on a knowledge base recombination scheme, calculates the matching degree of task complexity and user learning ability through Python, adjusts parameters including task type weight and user learning history efficiency, dynamically adjusts the difficulty and format of knowledge content, matches the learning ability of a user, and generates dynamically adjusted knowledge content.

Preferably, the composite knowledge reasoning module comprises a symbol logic construction sub-module, a neural network construction sub-module and a composite reasoning application sub-module;

The symbol logic building submodule adopts a symbol logic method based on dynamically adjusted knowledge content to define the logic structure of the traditional Chinese medicine knowledge through Prolog, and comprises the steps of defining the dependency rules and logic relations among knowledge points to generate a knowledge logic framework;

The neural network construction submodule constructs a neural network based on a knowledge logic framework by using deep learning frameworks TensorFlow and Keras, the model structure comprises a plurality of convolution layers Conv2D and a pooling layer MaxPooling D for feature extraction, a Dense layer is used for classification, an optimizer selects Adam, a loss function adopts categorical _ crossentropy, a complex mode in unstructured data is learned, and a deep learning model is generated;

The composite reasoning application submodule carries out mixed reasoning through a multi-mode reasoning library in Python based on a knowledge logic framework and a deep learning model, comprises verification of symbolic logic reasoning and analysis of a deep learning result, analyzes medical concepts of traditional Chinese medicine and generates a composite reasoning result.

Preferably, the effect prediction module comprises a model training sub-module, an effect prediction sub-module and an evidence base analysis sub-module;

the model training submodule carries out model training by using a random forest algorithm based on a composite reasoning result, random forest parameters are set through a scikit-learn library of Python, the number of trees is set to be 100, the maximum depth is 10, and a model in structural data is learned through training data to generate a prediction model prototype;

The effect prediction submodule is based on a prediction model prototype, a support vector machine algorithm is used, SVM parameters are set through scikit-learn libraries, a kernel function is selected as RBF, regularization parameter C is 1.0, effect prediction is refined, nonlinear relations in a feature space are processed, and a refined effect prediction model is generated;

The evidence base analysis submodule executes cross verification based on the refined effect prediction model, adopts a K-fold cross verification method, sets a K value to be 5, segments a data set by using KFold functions in a model_selection module in Python, trains and verifies the model for multiple times, evaluates the generalization capability of the model on unseen data and generates an effect prediction analysis result.

Preferably, the nonlinear learning path generation module comprises a knowledge graph construction sub-module, a path recommendation algorithm sub-module and a user support sub-module;

The knowledge graph construction submodule adopts a graph database technology based on an effect prediction analysis result, uses Neo4j, and generates a traditional Chinese medicine knowledge graph structure by writing a Cypher query statement, wherein parameters comprise knowledge point identifications and relationship types, and a graph structure of knowledge points and interrelationships thereof in the traditional Chinese medicine field is constructed to represent connection and hierarchy among the knowledge points;

The path recommendation algorithm submodule is used for carrying out study path design by applying a graph algorithm based on a traditional Chinese medicine knowledge graph structure, dijkstra algorithm and Louvain community discovery algorithm are used by using a networkx library of Python, parameter setting comprises starting nodes and target nodes and community modularity optimization, and optimal study path recommendation is generated by referring to user interests and knowledge point importance;

The user support submodule builds an interactive user interface based on optimized learning path recommendation, and creates a dynamic front-end interface by using a reaction frame, so that a user can browse a knowledge graph according to interests, select a recommended learning path, and simultaneously support dynamic preview and selection of the knowledge path to generate a nonlinear learning path.

Preferably, the dynamic diagram network construction module comprises a network construction sub-module, a time sequence data processing sub-module and a trend analysis prediction sub-module;

The network construction submodule is based on a nonlinear learning path, utilizes a dynamic graph construction technology, uses Gephi software to visualize the evolution relation among traditional Chinese medicine knowledge points, sets time window parameters to show the change of the knowledge points along with time, uses a force-oriented algorithm to optimize the network layout, shows the process of knowledge evolution and generates a knowledge evolution dynamic network;

the knowledge evolution tracking sub-module is based on a knowledge evolution dynamic network, performs time sequence data analysis, tracks the change trend of knowledge points by adopting an ARIMA model through a statsmodels library of Python, determines model parameters by the periodic characteristics of a data set, reveals the process of the development and the transition of the knowledge points along with time, and generates a knowledge change trend analysis result;

The trend analysis prediction sub-module predicts future development of knowledge points by applying a machine learning algorithm based on knowledge change trend analysis results, utilizes scikit-learn library configuration of Python through a random forest model, sets the number of random forest parameters including trees as 100, limits the depth to be none, captures complex trends and modes, predicts the development direction of the knowledge points, and generates a dynamic graph network.

Preferably, the time sequence analysis module comprises a time sequence data collection sub-module, a data analysis sub-module and a trend prediction sub-module;

The time sequence data collection submodule executes an automatic data collection flow based on a dynamic graph network, operates through a Python script and pandas library, comprises capturing traditional Chinese medicine knowledge points and time-varying data thereof from a plurality of data sources, performs data cleaning by using pandas, standardizes a date format, and generates a standardized time sequence data set;

The knowledge point change analysis submodule is used for analyzing the change trend of the knowledge points by utilizing an autoregressive moving average model based on a normalized time sequence data set, executing through a statsmodels library in Python, selecting model parameters p and q, revealing the dynamic characteristics of knowledge in the traditional Chinese medicine field by analyzing the change rule of the knowledge points along with time, and generating a knowledge point change trend analysis result;

The trend prediction submodule adopts a random forest algorithm to configure through a scikit-learn library of Python based on knowledge point change trend analysis results, sets the number of the included trees as 100, does not limit the maximum depth of the trees, predicts the future development directions of traditional Chinese medicine knowledge and application thereof according to the past and existing change trends, and generates a time sequence analysis result.

Preferably, the knowledge updating mechanism module comprises a knowledge base updating sub-module, an updating strategy making sub-module and an execution effect monitoring sub-module;

The knowledge base updating sub-module updates the knowledge base by using a content management tool based on a time sequence analysis result, performs version control by Git, records the updated content each time, comprises newly added knowledge points and updating information, and generates a knowledge base updating record;

The update strategy making submodule designs an update strategy based on the update record of the knowledge base, automatically analyzes the update requirement by using pandas libraries in Python, and sets the frequency and the priority of reference update by parameters so as to enable the timely update of the knowledge base to be matched with the requirement of a user and generate an update strategy scheme;

The execution effect monitoring sub-module executes quality control of updated contents based on an update strategy scheme, and performs Web interface test by adopting an automatic test tool Selenium, including selecting proper test cases and determining test frequency, so as to generate an updated traditional Chinese medicine knowledge base.

Compared with the prior art, the invention has the beneficial effects that: the chaos model construction module is introduced, so that the system can accurately analyze the nonlinearity and dynamic change of the traditional Chinese medicine data, deep dig the complex relation between the treatment effect and the physique reaction, and greatly improve the depth and breadth of data analysis. The application of the composite knowledge reasoning module combines the symbolic logic and the neural network technology, effectively improves the capability of the system for processing complex medical concepts and unstructured data, and provides deeper and accurate knowledge analysis for users. The effect prediction module remarkably improves the accuracy of the effect prediction of the treatment method and the medicine formula through a machine learning algorithm, and enhances the scientificity and reliability of the application of the traditional Chinese medicine. The nonlinear learning path generation module provides flexible path selection for learning by utilizing a knowledge graph technology, optimizes the logicality and systemicity of learning content and improves the learning efficiency. The dynamic graph network construction and the time sequence analysis module are combined, so that the evolution trend of the knowledge of the traditional Chinese medicine along with time is effectively captured, the timeliness and the foresight of the content of the science popularization are ensured, and the knowledge is updated more timely and accurately. Through comprehensive application of the technology, the invention not only improves the propagation efficiency and accuracy of the knowledge of the traditional Chinese medicine, but also strengthens the dynamic updating capability of a science popularization system and the scientificity of knowledge content, and provides firm support for the public to understand the traditional Chinese medicine.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed for the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a block diagram of a traditional Chinese medicine knowledge science popularization system utilizing artificial intelligence and big data analysis;

FIG. 2 is a diagram showing a system frame of a traditional Chinese medicine knowledge science popularization system using artificial intelligence and big data analysis;

FIG. 3 is a schematic diagram showing a chaotic model construction module in a traditional Chinese medicine knowledge science popularization system by utilizing artificial intelligence and big data analysis;

FIG. 4 is a schematic diagram showing a knowledge granularity adjustment module in a traditional Chinese medicine knowledge science popularization system by utilizing artificial intelligence and big data analysis;

FIG. 5 is a schematic diagram of a compound knowledge reasoning module in a traditional Chinese medicine knowledge science popularization system utilizing artificial intelligence and big data analysis;

FIG. 6 is a schematic diagram showing an effect prediction module in a traditional Chinese medicine knowledge science popularization system by utilizing artificial intelligence and big data analysis;

FIG. 7 is a schematic diagram of a nonlinear learning path generation module in a traditional Chinese medicine knowledge science popularization system utilizing artificial intelligence and big data analysis;

FIG. 8 is a schematic diagram of a dynamic graph network construction module in a traditional Chinese medicine knowledge science popularization system utilizing artificial intelligence and big data analysis;

FIG. 9 is a schematic diagram showing a timing analysis module in a traditional Chinese medicine knowledge science popularization system using artificial intelligence and big data analysis;

fig. 10 is a schematic diagram of a knowledge updating mechanism module in a traditional Chinese medicine knowledge science popularization system using artificial intelligence and big data analysis.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

Referring to fig. 1, the present invention provides a technical solution: a traditional Chinese medicine knowledge science popularization system utilizing artificial intelligence and big data analysis comprises a chaos model construction module, a knowledge granularity adjustment module, a compound knowledge reasoning module, an effect prediction module, a nonlinear learning path generation module, a dynamic graph network construction module, a time sequence analysis module and a knowledge updating mechanism module;

The chaotic model construction module analyzes the dynamic property of the traditional Chinese medicine data by adopting a chaotic dynamics theory based on the historical case data, extracts key time characteristics by a time sequence analysis method, maps a nonlinear structure of the data by utilizing a phase space reconstruction technology, calculates and confirms the chaotic characteristic of the system by Lyapunov indexes, and generates a chaotic dynamics model;

The knowledge granularity adjustment module is used for collecting user feedback and learning behavior data by adopting a user modeling technology based on a chaotic dynamics model, analyzing learning preference and cognition level of a user by using a content adaptation algorithm, evaluating difficulty of knowledge content by using a cognition load theory, adjusting knowledge granularity to match user requirements, and generating dynamically adjusted knowledge content;

the effect prediction module analyzes the structured data by adopting a random forest algorithm based on the composite reasoning result, processes the nonlinear relation in the multidimensional feature space by using a support vector machine, and generates an effect prediction analysis result by cross-verifying the optimized model parameters;

The nonlinear learning path generation module integrates knowledge points and relations thereof in the traditional Chinese medicine field by adopting a knowledge graph construction technology based on the effect prediction analysis result, designs a learning path by a shortest path and a community mining algorithm, recommends learning content for a user by referring to user interests and knowledge point importance, and generates a nonlinear learning path;

The time sequence analysis module is used for analyzing the traditional Chinese medicine knowledge and the change of the application of the traditional Chinese medicine knowledge along with time by adopting a time sequence analysis method based on a dynamic graph network, predicting the development direction of the knowledge by adopting an autoregressive moving average model and a trend analysis technology, and generating a time sequence analysis result;

The chaotic dynamics model comprises stability parameters, characteristic quantities of chaotic attractors and key dynamic behavior indexes, the dynamically adjusted knowledge content comprises knowledge units with multiple difficulty levels, learning materials adjusted according to cognitive load and knowledge points matched with user preferences, the composite reasoning result comprises a logic expression of a traditional Chinese medicine rule, mode characteristics of deep learning identification, knowledge conclusion obtained through comprehensive reasoning, the effect prediction analysis result comprises treatment effect scores of symptoms, confidence intervals of the prediction model and priority ordering of treatment suggestions, the nonlinear learning path comprises learning target setting, links of recommended learning resources and learning sequences of key knowledge points, the dynamic graph network comprises a dependency graph among the knowledge points, knowledge evolution tracks in time dimension and evolution analysis of key nodes, the time sequence analysis result comprises time sequence prediction of knowledge point change, key time points of trend change and potential and new knowledge mining areas, and the updated traditional Chinese medicine knowledge base comprises newly added treatment method items, revised medicine information and a verified knowledge update list.

In the chaotic model construction module, firstly, historical case data is collected and used as input, and the data exists in a time sequence form and comprises information such as the state before and after treatment of a patient, the prescription used, the treatment effect and the like. The dynamics of the data is analyzed by adopting a chaos dynamics theory, and key time characteristics such as the average value, variance, maximum value, minimum value and the like of a time sequence are extracted by a time sequence analysis method. Then, the nonlinear structure of the data is mapped by using a phase space reconstruction technique, and the step reconstructs the phase space by selecting the embedding dimension and the delay time, so as to reveal the inherent dynamic behavior of the data. And determining the chaotic characteristic of the system, namely judging the sensitivity of the system state evolving along with time by calculating the Lyapunov index. The finally generated chaotic dynamics model can reveal the inherent dynamic change rule of the traditional Chinese medicine data, and provides a theoretical basis for subsequent knowledge granularity adjustment and composite knowledge reasoning. The realization of the module effectively improves the capability of the system for analyzing the nonlinear characteristics of the traditional Chinese medicine data, and provides a new view for deep mining and application of the traditional Chinese medicine knowledge.

In the knowledge granularity adjustment module, based on the chaotic dynamics model, user feedback and learning behavior data are collected through a user modeling technology, wherein the data comprise learning history, preference, feedback on knowledge content and the like of a user. The data is analyzed through a content adaptation algorithm, the cognitive level and learning preference of the user are considered by the algorithm, the difficulty level of the knowledge content is comprehensively evaluated, and the factors such as the depth and complexity of the knowledge content, the learning efficiency of the user in history and the like are involved. And dynamically adjusting knowledge granularity according to the cognitive load theory to match the learning requirement of the user. The process generates dynamically adjusted knowledge content that is presented in knowledge units of different difficulty levels, ensuring that users can learn efficiently according to their own capabilities. The module optimizes learning efficiency and user experience by finely adjusting the granularity of knowledge transfer, so that the knowledge is more fit with the actual demands of users.

The composite knowledge reasoning module improves knowledge processing and reasoning capacity of the system by integrating symbol logic and deep neural network technology. First, the symbolic logic part processes the determined logic relationship and rule, and defines a well-defined logic structure for the knowledge of the traditional Chinese medicine, including the dependency relationship among knowledge points and causal reasoning rule. The deep neural network section is then responsible for learning fuzzy, unstructured data patterns, such as text descriptions and image information, extracting features and identifying patterns through the multi-layer network structure. The two parts of results are integrated by a composite reasoning mechanism, and the respective advantages are complemented through the cooperative work of the algorithm, so that a composite reasoning result is finally generated. The result not only comprises a logic expression of the Chinese medicine rule, but also comprises pattern features of deep learning identification, thereby providing more comprehensive and accurate knowledge analysis for users. The module obviously improves the capability of the system for processing complex medical concepts, so that knowledge reasoning is more accurate and deep.

The effect prediction module performs data analysis and prediction by adopting a random forest algorithm and a support vector machine technology based on the composite reasoning result. The random forest algorithm analyzes the structured data by constructing a plurality of decision trees, each decision tree is trained on a random subset of the data set, and the final result is determined by voting or average value, so that the generalization capability of the model is effectively improved. The support vector machine is focused on processing the nonlinear relationship in the multidimensional feature space, and data classification is realized by searching the optimal decision boundary. Model parameters are optimized through a cross verification method, and accuracy and reliability of prediction are guaranteed. And finally, the generated effect prediction analysis result comprises the scores of the treatment effects, the confidence intervals of the prediction model and the priority ordering of the treatment suggestions, so that scientific decision support is provided for the treatment of the traditional Chinese medicine. The implementation of the module greatly enhances the capability of the system in the aspect of traditional Chinese medicine effect prediction, and provides accurate medical advice and treatment scheme.

The nonlinear learning path generation module provides flexible learning path design for users through knowledge graph construction technology. Firstly, knowledge points and correlations thereof in the traditional Chinese medicine field are integrated by adopting a graph database technology, a knowledge graph structure is constructed, and logic relations and dependencies among the knowledge points are defined. And then, dynamically designing a learning path by applying a shortest path algorithm and a community mining algorithm, and recommending a proper learning sequence for the user by the algorithm according to the importance and the interrelationship of the knowledge points. The nonlinear learning path generated by the process not only considers the internal relation among knowledge points, but also considers the comprehensiveness and systematicness of knowledge, and provides an efficient learning scheme for users. The implementation of the module optimizes the logic and systematicness of learning content and improves the learning efficiency and the knowledge grasping depth.

The dynamic diagram network construction module adopts a dynamic network model to capture the evolution relationship among the knowledge points of the traditional Chinese medicine. The time sequence analysis method is used for tracking the variation trend of knowledge points, the development of a knowledge network is simulated by using a network evolution algorithm, and the evolution process of knowledge along with time is displayed. The generated dynamic graph network not only depicts the dependency graph among knowledge points, but also shows the evolution analysis of the knowledge evolution track and key nodes in the time dimension. The realization of the module provides dynamic knowledge updating and development trend prediction capability for the system, and ensures timeliness and foresight of popular science content.

The time sequence analysis module is based on a dynamic graph network and carries out deep analysis on the traditional Chinese medicine knowledge and the time change of application thereof through an autoregressive moving average model and a trend analysis technology. Time sequence data related to traditional Chinese medicine is collected and processed, the rule of knowledge point change is analyzed through a model, and the development direction of knowledge is predicted. The generated time sequence analysis result comprises time sequence prediction of knowledge point change and key time points of trend change, and scientific basis is provided for knowledge updating and content optimization. The application of the module enhances the capability of the system in knowledge dynamic monitoring and prediction, and provides powerful support for the propagation and update of the knowledge of the traditional Chinese medicine.

The knowledge updating mechanism module realizes continuous updating and maintenance of the knowledge base through a content management tool and a version control technology. Based on the time sequence analysis result, new knowledge points and update information are introduced periodically, and the accuracy and reliability of the update content are ensured through a quality control flow. Not only an updated traditional Chinese medicine knowledge base is generated, but also timeliness and scientificity of knowledge content are guaranteed. The implementation of the module provides a solid foundation for continuous propagation and science popularization of the knowledge of the traditional Chinese medicine, and ensures high quality and high credibility of the knowledge content.

Referring to fig. 2 and 3, the chaotic model construction module includes a chaotic theory analysis sub-module, a dynamics model construction sub-module, and an effect change rule analysis sub-module;

The chaos theory analysis submodule carries out preliminary analysis of chaos theory based on historical case data, calculates standard deviation and average value of data by using numpy and scipy libraries in Python, estimates initial chaos indexes, draws a time sequence chart by using matplotlib, monitors data volatility, determines whether to carry out chaos analysis or not through the volatility, and generates a chaos analysis result;

The dynamics model construction submodule is used for carrying out phase space reconstruction by using numpy libraries in Python based on chaos analysis results and adopting a dynamics modeling method, setting embedding dimension as 3 and delay time as 2, constructing a phase space image of dynamic behaviors, analyzing the dynamic behaviors and stability of data, and generating a dynamics model;

The effect change rule analysis submodule is based on a dynamics model, adopts time sequence analysis, uses pandas and statsmodels libraries of Python to execute an autoregressive moving average model, automatically determines p values and q values through AIC criteria in parameter selection, analyzes dynamic relations between traditional Chinese medicine treatment effects and physique reaction, and generates a chaotic dynamics model.

In the chaos theory analysis submodule, the chaos characteristics of the traditional Chinese medicine data are identified through preliminary analysis of the historical case data. The process involves using the numpy and scipy libraries of Python to calculate statistical properties of the data, such as standard deviation and average, as a basis for determining data chaos. Then, a matplotlib library is used for drawing a time sequence diagram, the step aims at intuitively showing the change condition of data along with time, and whether chaos analysis is needed or not is preliminarily judged by observing the fluctuation of the data. If the data shows high irregularity and complexity, namely high fluctuation, the system has chaos characteristics, and the necessity of chaos analysis is further determined. The chaos analysis result generated by the submodule provides scientific basis for deep research of traditional Chinese medicine data, reveals nonlinear dynamic rules existing behind the data, and lays a foundation for subsequent chaos dynamics model construction.

The dynamics model construction submodule adopts a dynamics modeling method to deeply mine the internal dynamic behavior of the traditional Chinese medicine data based on the chaos analysis result of the previous step. In this process, phase space reconstruction is performed by using the numpy library of Python, which is accomplished by selecting a suitable embedding dimension and delay time, setting the embedding dimension to 3 and the delay time to 2, and such setting is intended to construct a phase space image capable of reflecting the dynamic behavior of the system. By analyzing the image, the dynamic characteristics of the data, such as periodicity, stability, or chaos, can be more deeply understood. The generated dynamic model not only can reveal the dynamic rule of the traditional Chinese medicine data, but also provides possibility for predicting and controlling the system behavior, and enhances the understanding of the dynamic relationship between the traditional Chinese medicine treatment effect and the physique reaction.

The effect change rule analysis submodule further researches the dynamic relationship between the treatment effect of the traditional Chinese medicine and the physique reaction by using a dynamic model and adopting a time sequence analysis method. In an implementation, the time series data is processed using pandas libraries of Python, and statsmodels libraries perform an autoregressive moving average model, where the process of automatically selecting model parameters p and q is based on the red-pool information criterion (AIC) to determine the best model. The analysis method can deeply explore the law of the change of the therapeutic effect of the traditional Chinese medicine along with time and the dynamic difference of different constitutions on the therapeutic response. Through the complex data analysis process, the generated chaotic dynamics model not only enhances the understanding of deep dynamic characteristics of the traditional Chinese medicine data, but also provides scientific basis for formulating more personalized and accurate treatment schemes, and remarkably improves the capability of the traditional Chinese medicine knowledge science popularization system in data analysis and application.

Referring to fig. 2 and 4, the knowledge granularity adjustment module includes a user model construction sub-module, a knowledge base structural design sub-module, and a dynamic granularity adjustment machine sub-module;

The user model building submodule is used for carrying out user feedback and learning behavior data collected based on a chaotic dynamics model, adopting a decision tree algorithm, executing through a scikit-learn library of Python, setting parameters of a decision tree, including maximum depth of 5 and minimum division sample number of 10, classifying user data, analyzing learning preference and cognition level of a user, and generating a user characteristic model;

The knowledge base structural design submodule applies a K-mean value clustering algorithm based on a user characteristic model, and operates by utilizing scikit-learn base of Python again, wherein the core parameters of the clustering algorithm are set to be the optimal clustering number selected automatically, and the knowledge base is reorganized based on the user characteristic to generate a knowledge base reorganization scheme;

In the user model construction sub-module, user feedback and learning behavior data collected based on a chaotic dynamics model are analyzed, and the data exist in a structured format, wherein the structured format comprises learning progress, preference feedback, interaction logs and the like of a user. The data was classified and analyzed using a decision tree algorithm and the process was performed using the scikit-learn library of Python. Specifically, by setting the maximum depth of the decision tree to 5 and the minimum number of split samples to 10, the algorithm can effectively distinguish between learning preferences and cognitive levels of different users when processing user data. In the operation process, the algorithm generates a series of decision rules, the rules predict the classification of the user based on the characteristics of the user data, a finally generated user characteristic model is stored in a file form, the model details the behavior characteristics and the learning requirements of the user group, and an accurate basis is provided for subsequent knowledge content adjustment.

In the knowledge base structural design submodule, a K-means clustering algorithm is adopted to reorganize the knowledge base based on the user characteristic model. The process is also implemented by using scikit-learn library of Python, and the core parameters of the clustering algorithm, such as the optimal clustering number, are automatically selected according to the user characteristics. This means that the algorithm groups knowledge content into a plurality of knowledge units according to different characteristics of the user population, such as learning progress and preference. The reorganization of the knowledge units aims at better matching the learning requirement of the user, the generated knowledge base reorganization scheme is stored in a file form, and the scheme plans the new structure and organization mode of the knowledge content in detail, so that the structure of the knowledge base is more in line with the learning mode and the cognitive structure of the user, and the learning efficiency and the accessibility of the knowledge are improved.

And in the dynamic granularity adjusting machine submodule, the difficulty and the format of the knowledge content are dynamically adjusted by adopting a cognitive load evaluation technology. Parameters such as task type weights and user learning history efficiency are adjusted through the complexity of the Python calculation task and the matching degree of the user learning ability so as to ensure that the difficulty degree of the knowledge content can be adapted to the cognition level of the user. In this process, the system generates dynamically adjusted knowledge content that is optimized according to the learning ability and cognitive load of the user. The generated dynamically adjusted knowledge content is not only static documents or data, but also dynamic learning resources adjusted according to the specific needs of each user, and the resources are presented in a form most suitable for the current cognitive state of the user, so that the pertinence and effect of the user learning are effectively improved, each user can learn at a proper difficulty level, and the learning effect is maximized.

Referring to fig. 2 and 5, the composite knowledge reasoning module includes a symbol logic construction sub-module, a neural network construction sub-module, and a composite reasoning application sub-module;

The symbol logic building submodule adopts a symbol logic method based on dynamically adjusted knowledge content to define the logic structure of the traditional Chinese medicine knowledge through Prolog, and the symbol logic building submodule comprises the steps of defining the dependency rules and logic relations among knowledge points and generating a knowledge logic framework;

The neural network construction submodule constructs a neural network based on a knowledge logic framework by using deep learning frameworks TensorFlow and Keras, the model structure comprises a plurality of convolution layers Conv2D and pooling layers MaxPooling D for feature extraction, a Dense layer is used for classification, an optimizer selects Adam, a loss function adopts categorical _ crossentropy, and a complex mode in unstructured data is learned to generate a deep learning model;

The composite reasoning application submodule carries out mixed reasoning through a multi-mode reasoning base in Python based on a knowledge logic framework and a deep learning model, and the mixed reasoning comprises verification of symbolic logic reasoning and analysis of a deep learning result, analyzes medical concepts of traditional Chinese medicine and generates a composite reasoning result.

In the symbol logic construction sub-module, a logic structure of the Chinese medicine knowledge is defined by adopting a symbol logic method. In the process, prolog programming language is used for specific implementation, and a detailed knowledge logic framework is formed by defining dependency rules and logic relations among knowledge points. The rules and relationships are encoded as facts and rules in Prolog, providing a solid logical basis for subsequent deep learning models. Such a logical framework not only clarifies the inherent links between knowledge points, but also makes the knowledge system more systematic and organized, facilitating machine understanding and reasoning. The generated knowledge logic framework file provides necessary input for the deep learning part of the compound knowledge reasoning module, and ensures the logic and accuracy of the reasoning process.

In the neural network construction sub-module, a neural network model is constructed using TensorFlow and Keras deep learning frameworks based on the previously defined knowledge logic framework. The model includes multiple convolution layers (Conv 2D) and pooling layers (MaxPooling D) for feature extraction, and a full connection layer (Dense) for classification. By setting the optimizer to Adam, the loss function is chosen to categorical _ crossentropy, and the setting of this parameter is based on consideration of the model training effect and convergence speed. In the training process, a fit method is used for training the model, and evaluate method is applied for verifying the effect of the model. The model can learn complex modes in unstructured data, such as Chinese medicine knowledge in texts and images, and the generated deep learning model file provides strong mode recognition capability for compound reasoning, so that the system can process and infer more complex and fuzzy medical concepts.

In the composite reasoning application submodule, a mixed reasoning process is executed through a multi-mode reasoning base in Python based on a knowledge logic framework and a deep learning model which are constructed previously. In the process, the symbol logic reasoning and the deep learning result are comprehensively considered, and not only are the definite rules and structured knowledge processed by utilizing the logic reasoning, but also unstructured complex data are understood and inferred through a deep learning model. The mixed reasoning mechanism fully plays respective advantages of symbol logic and deep learning, and improves comprehensiveness and depth of reasoning. The generated composite reasoning result not only comprises the logic expression and the classification result of the medical concept, but also analyzes the deep knowledge and rules of the traditional Chinese medicine in detail, and provides more accurate and deep knowledge solutions for users. The reasoning result file provides firm support for knowledge propagation and application, and the intelligent level and user experience of the system are remarkably improved.

Referring to fig. 2 and 6, the effect prediction module includes a model training sub-module, an effect prediction sub-module, and an evidence base analysis sub-module;

The model training sub-module performs model training by using a random forest algorithm based on a composite reasoning result, random forest parameters are set through a scikit-learn library of Python, the number of trees is set to be 100, the maximum depth is 10, and a model in structural data is learned through training data to generate a prediction model prototype;

The evidence base analysis submodule executes cross verification based on the refinement effect prediction model, adopts a K-fold cross verification method, sets a K value to be 5, segments a data set by using KFold functions in a model_selection module in Python, trains and verifies the model for multiple times, evaluates the generalization capability of the model on unseen data and generates an effect prediction analysis result.

In the model training sub-module, the composite reasoning result is analyzed and learned through a random forest algorithm to construct a model prototype capable of predicting the treatment effect of the traditional Chinese medicine. The process is embodied using the scikit-learn library of Python, where the key parameters of the random forest algorithm include setting the number of trees to 100 and the maximum depth to 10. The setting of the parameters is based on the generalization ability of the optimization model and the consideration of avoiding overfitting. By inputting structured data, including case characteristics, treatment methods, treatment results, etc., the algorithm learns patterns in the data and builds a predictive model prototype. The prototype model can predict the treatment effect according to given traditional Chinese medicine treatment data, and the generated model file provides a basis for subsequent effect prediction, so that the effect prediction process is more accurate and efficient.

In the effect prediction sub-module, based on the prediction model prototype, a Support Vector Machine (SVM) algorithm is further used for refining effect prediction. Also implemented by the scikit-learn library, the core parameters of the SVM algorithm include selecting a Radial Basis Function (RBF) as the kernel function, and the regularization parameter C is set to 1.0. The parameters are chosen to address non-linear relationships in feature space and optimize the predictive performance of the model. By further training and adjusting the model prototype, a refined effect prediction model is generated, which is focused on improving the accuracy and reliability of the prediction. The generated refinement effect prediction model not only improves the accuracy of the prediction result, but also enables the model to better understand and process the complexity of the treatment effect of the traditional Chinese medicine, and creates conditions for providing scientific and accurate treatment suggestions.

In the evidence base analysis submodule, the refinement effect prediction model is evaluated and verified by executing cross verification, so that the generalization capability and accuracy of the model are ensured. A K-fold cross-validation method is adopted, wherein the K value is set to be 5, and a KFold function in a model_selection module in Python is used for dividing a data set, so that multiple training and validation processes are carried out. The method can ensure the performance of the model on unseen data, and evaluate the stability and reliability of the model through training and testing on different data subsets. The effect prediction analysis result generated by the process provides a solid evidence base for the prediction of the treatment effect of the traditional Chinese medicine, ensures that the medical advice provided for the user is not only based on advanced analysis of deep learning and compound reasoning, but also goes through a strict verification process, and enhances the credibility and application value of the prediction result.

Referring to fig. 2 and 7, the nonlinear learning path generation module includes a knowledge graph construction sub-module, a path recommendation algorithm sub-module, and a user support sub-module;

The path recommendation algorithm submodule is used for designing a learning path by applying a graph algorithm based on a traditional Chinese medicine knowledge graph structure, and generating optimization learning path recommendation by using Dijkstra algorithm and Louvain community discovery algorithm through a networkx library of Python, wherein parameter setting comprises starting nodes and target nodes and modularity optimization of communities, and referring to user interests and knowledge point importance;

the user support sub-module builds an interactive user interface based on optimized learning path recommendation, and creates a dynamic front-end interface by using a reaction frame, so that a user can browse a knowledge graph according to interests, select a recommended learning path, and simultaneously support dynamic preview and selection of the knowledge path to generate a nonlinear learning path.

In the knowledge graph construction submodule, graph database technology Neo4j is used for combining effect prediction analysis results to construct graph structures of knowledge points and correlations of the knowledge points in the traditional Chinese medicine field. The specific operation comprises the step of writing a Cypher query statement to define knowledge point identification and relationship types, wherein the knowledge points and the relationships are organized according to multidimensional information such as therapeutic effects of traditional Chinese medicines, pharmaceutical compositions, disease types and the like. By the method, the knowledge graph reflecting the internal logic and hierarchical structure of the traditional Chinese medicine knowledge system is constructed, the graph is stored in the form of nodes and edges in the graph database, the generated traditional Chinese medicine knowledge graph file is convenient for subsequent inquiry, analysis and study path recommendation, the knowledge retrieval efficiency and accuracy are improved, and an intuitive and systematic study resource is provided for users.

In the path recommendation algorithm submodule, a personalized learning path is designed by applying a graph algorithm based on the constructed traditional Chinese medicine knowledge graph. And (3) path planning and community optimization are carried out through a networkx library of Python and combining a Dijkstra algorithm and a Louvain community discovery algorithm. The Dijkstra algorithm is used for calculating the shortest paths among knowledge points, so that the high efficiency of the learning paths is ensured; and the Louvain community discovery algorithm is used for grouping according to the similarity and the closeness of the knowledge points and optimizing the consistency and the systematicness of the learning path. The parameters set in the process, such as the starting node and the target node, are based on the interests and knowledge demands of the user, and the generated optimized learning path recommendation provides a learning route which accords with personal interests and covers important knowledge points for the user, so that the pertinence and the effectiveness of learning are enhanced.

In the user support sub-module, a dynamic front-end user interface is built using the reaction framework based on the optimized learned path recommendations. The interface allows the user to browse the knowledge graph, select a recommended learning path, and support dynamic previewing and selection of knowledge paths according to personal interests and learning needs. Through the interactive design, the user can intuitively see the relationship and path recommendation among knowledge points, so that the learning process is more personalized and flexible. In addition, the realization of the user interface promotes the deep interaction between the user and the knowledge content, improves the initiative and participation of learning, and the generated nonlinear learning path is not only convenient for the user to learn according to own rhythm and preference, but also enables the knowledge to be mastered more comprehensively and systematically.

Referring to fig. 2 and 8, the dynamic diagram network construction module includes a network construction sub-module, a time sequence data processing sub-module, and a trend analysis prediction sub-module;

The network construction sub-module is based on a nonlinear learning path, utilizes a dynamic graph construction technology, uses Gephi software to visualize the evolution relation among traditional Chinese medicine knowledge points, sets time window parameters to show the change of the knowledge points along with time, uses a force-oriented algorithm to optimize the network layout, shows the knowledge evolution process and generates a knowledge evolution dynamic network;

the knowledge evolution tracking sub-module is used for executing time sequence data analysis based on a knowledge evolution dynamic network, tracking the change trend of a knowledge point by adopting an ARIMA model through a statsmodels library of Python, determining model parameters by the periodic characteristics of a dataset, revealing the process of the development and the transition of the knowledge point along with time, and generating a knowledge change trend analysis result;

In the network construction submodule, by utilizing a dynamic diagram construction technology and combining Gephi software, the evolution relation between the traditional Chinese medicine knowledge points is visualized. The data format is nodes and edges in a graph database, wherein the nodes represent knowledge points of traditional Chinese medicine, and the edges represent logical relations or interactions between the knowledge points. By setting the time window parameters, the change and evolution of knowledge points along with time can be shown, and the network layout is optimized by using a force-oriented algorithm, so that the evolution process of knowledge is visual and easy to understand. The operation can generate a knowledge evolution dynamic network, the network not only reveals the internal structure and dynamic change of the traditional Chinese medicine knowledge system, but also provides a basis for recommendation of learning paths and deep analysis of knowledge points, and the visualization and interactivity of the knowledge system are enhanced.

In the knowledge evolution tracking sub-module, time sequence data analysis is executed to track the change trend of the knowledge points based on the knowledge evolution dynamic network. The ARIMA model is adopted and implemented through statsmodels libraries of Python, and the time sequence data format comprises the creation time, the update time, the correlation change and the like of knowledge points. The selection of model parameters, such as autoregressive terms, differential times, and moving average terms, is determined based on the periodic characteristics and volatility of the data set. By the method, the development and transformation process of knowledge points along with time can be revealed, and knowledge change trend analysis results are generated. The result provides scientific basis for updating the knowledge base and optimizing the content, and helps to understand and predict the development direction of the traditional Chinese medicine knowledge system.

And in the trend analysis and prediction sub-module, based on the knowledge change trend analysis result, a machine learning algorithm is applied to predict future development of the knowledge points. Configuration is performed by using a random forest model with a scikit-learn library of Python, wherein the parameter settings include the number of trees set to 100 and the depth limit to none, to accommodate the complexity and uncertainty of knowledge evolution. The method can capture the complex trend and mode of knowledge point development, predict the future development direction of knowledge points, and the generated dynamic graph network not only reflects the current state of a knowledge system, but also provides the prediction of future changes. The prediction result has important significance for guiding continuous updating of the knowledge base, optimizing the learning path and making the education strategy, and improves the adaptability and the foresight of the system to the knowledge development trend.

Referring to fig. 2 and 9, the timing analysis module includes a timing data collection sub-module, a data analysis sub-module, and a trend prediction sub-module;

The time sequence data collection submodule executes an automatic data collection flow based on a dynamic graph network, and operates through a Python script and pandas library, and comprises the steps of capturing knowledge points of traditional Chinese medicines and time-varying data thereof from a plurality of data sources, performing data cleaning by using pandas, standardizing a date format and generating a standardized time sequence data set;

the knowledge point change analysis submodule is used for analyzing the change trend of the knowledge points by utilizing an autoregressive moving average model based on a normalized time sequence data set, executing the analysis by utilizing a statsmodels library in Python, selecting model parameters p and q, and revealing the dynamic characteristics of knowledge in the traditional Chinese medicine field by analyzing the change rule of the knowledge points along with time to generate a knowledge point change trend analysis result;

In the time sequence data collection sub-module, the time sequence data in the traditional Chinese medicine field is collected by executing an automatic data collection flow. The process utilizes Python script in combination with pandas library to automatically capture data relating to knowledge points of traditional Chinese medicine and its changes over time from multiple data sources. The data format is CSV or JSON, and key information such as identification, description, change date and the like of the knowledge points is contained. Data cleansing, such as removal of missing values, unifying date formats, is performed using pandas to generate a normalized time series dataset. The data set provides a basis for subsequent trend analysis and prediction, and ensures the accuracy and effectiveness of analysis.

In the knowledge point change analysis submodule, based on a normalized time sequence data set, an autoregressive moving average (ARIMA) model is adopted to conduct deep analysis on the change trend of the knowledge point. The process is performed by means of a statsmodels library in Python, the selection of model parameters p (autoregressive terms) and q (moving average terms) being determined based on a preliminary analysis of the time series data, aimed at optimally capturing trends and seasonal patterns in the data. By analyzing the change rule of knowledge points along with time through the method, the dynamic characteristics of knowledge in the field of traditional Chinese medicine are revealed, and the generated analysis result of the change trend of the knowledge points is stored in a file form, so that insight is provided for updating and developing the knowledge of the traditional Chinese medicine.

And in the trend prediction sub-module, based on the knowledge point change trend analysis result, predicting the future development direction by using a random forest algorithm. A random forest model is configured by the scikit-learn library of Python, where the number of trees is set to 100, without limiting the maximum depth of the trees, to adequately capture complex data relationships and patterns. Such a configuration facilitates model learning of past and existing trends, thereby making predictions of future development of traditional Chinese medicine knowledge and its applications. The generated time sequence analysis result is stored in a file form, so that scientific prediction of future changes in the field of traditional Chinese medicine is provided, important references are provided for planning and strategy formulation of knowledge content, and understanding and grasping of the development trend of traditional Chinese medicine are enhanced.

Referring to fig. 2 and 10, the knowledge updating mechanism module includes a knowledge base updating sub-module, an updating policy making sub-module, and an execution effect monitoring sub-module;

Based on the time sequence analysis result, the knowledge base updating sub-module uses a content management tool to update the knowledge base, carries out version control through Git, records the content updated each time, comprises newly added knowledge points and updating information, and generates a knowledge base updating record;

The execution effect monitoring sub-module executes quality control of the updated content based on the update policy scheme, and performs Web interface testing by adopting an automatic testing tool Selenium, including selecting appropriate test cases and determining test frequency, to generate an updated traditional Chinese medicine knowledge base.

In the knowledge base update sub-module, updates of the traditional Chinese medicine knowledge base are managed and recorded by an automated script and version control system Git. Specifically, the sub-module uses the results of the timing analysis as an update trigger point, integrates the Git through a content management tool (such as the CMS system), and automatically records each submitted change, including newly added knowledge points and updated information. The record is stored in a version control system, and a new version number is generated for each update, so that the change history of the knowledge base is ensured to be accurately recorded and tracked. The generated knowledge base update record not only provides a detailed change log, but also supports backtracking and version comparison, thereby enhancing the maintenance efficiency and accuracy of the knowledge base.

In the update policy making submodule, the pandas library of Python is used to automatically analyze the update requirements by analyzing the knowledge base update records. The sub-module automatically analyzes and determines the update strategy based on the content and frequency of the update records, as well as the feedback and needs of the user. Policies include determining priority of knowledge points, frequency of updates, and associated allocation of responsibilities. By the method, the updating strategy scheme can ensure that the updating of the knowledge base meets the requirements of users, and resources can be effectively utilized, so that the knowledge base is kept in the latest state, and meanwhile, the waste of the resources is avoided.

In the execution effect monitoring sub-module, the updated knowledge base is quality controlled by an automated test tool Selenium. The submodule selects proper test cases according to the updated strategy scheme, determines the test frequency and automatically executes the test of the Web interface. In this way, it is ensured that the knowledge base updated each time can reach the expected quality standard, and potential problems can be found and corrected in time. The generated updated traditional Chinese medicine knowledge base is rich in content, high in accuracy, friendly in interface and good in user experience, and the practicability and the user satisfaction of the system are further improved.

In the description of the present specification, the descriptions of the terms "one embodiment," "example," "specific example," and the like, mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

The preferred embodiments of the invention disclosed above are intended only to assist in the explanation of the invention. The preferred embodiments are not exhaustive or to limit the invention to the precise form disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best understand and utilize the invention. The invention is limited only by the claims and the full scope and equivalents thereof.

Claims

1. A traditional Chinese medicine knowledge science popularization system utilizing artificial intelligence and big data analysis is characterized by comprising a chaotic model construction module, a knowledge granularity adjustment module, a compound knowledge reasoning module, an effect prediction module, a nonlinear learning path generation module, a dynamic graph network construction module, a time sequence analysis module and a knowledge updating mechanism module;

2. The system for scientific knowledge of traditional Chinese medicine using artificial intelligence and big data analysis according to claim 1, wherein: the chaotic dynamics model comprises stability parameters, characteristic quantities of chaotic attractors and key dynamic behavior indexes, the dynamically adjusted knowledge content comprises knowledge units with multiple difficulty levels, learning materials adjusted according to cognitive load and knowledge points matched with user preferences, the composite reasoning result comprises a logic expression of a traditional Chinese medicine rule, mode characteristics of deep learning identification and knowledge conclusion obtained through comprehensive reasoning, the effect prediction analysis result comprises treatment effect scoring of symptoms, confidence intervals of a prediction model and priority ordering of treatment suggestions, the nonlinear learning path comprises learning target setting, links of recommended learning resources and learning sequences of the key knowledge points, the dynamic graph network comprises a dependency relation graph among the knowledge points, knowledge evolution tracks on time dimension and evolution analysis of the key nodes, the time sequence analysis result comprises time sequence prediction of the change of the knowledge points, key time points of the change of the trend and potential knowledge and new knowledge mining areas, and the updated traditional Chinese medicine knowledge base comprises newly added treatment method items, revised medicine information and verified updated lists.

3. The system for scientific knowledge of traditional Chinese medicine using artificial intelligence and big data analysis according to claim 1, wherein: the chaotic model construction module comprises a chaotic theory analysis sub-module, a dynamics model construction sub-module and an effect change rule analysis sub-module;

4. The system for scientific knowledge of traditional Chinese medicine using artificial intelligence and big data analysis according to claim 1, wherein: the knowledge granularity adjustment module comprises a user model construction sub-module, a knowledge base structural design sub-module and a dynamic granularity adjustment machine sub-module;

5. The system for scientific knowledge of traditional Chinese medicine using artificial intelligence and big data analysis according to claim 1, wherein: the composite knowledge reasoning module comprises a symbol logic construction sub-module, a neural network construction sub-module and a composite reasoning application sub-module;

6. The system for scientific knowledge of traditional Chinese medicine using artificial intelligence and big data analysis according to claim 1, wherein: the effect prediction module comprises a model training sub-module, an effect prediction sub-module and an evidence base analysis sub-module;

7. The system for scientific knowledge of traditional Chinese medicine using artificial intelligence and big data analysis according to claim 1, wherein: the nonlinear learning path generation module comprises a knowledge graph construction sub-module, a path recommendation algorithm sub-module and a user support sub-module;

8. The system for scientific knowledge of traditional Chinese medicine using artificial intelligence and big data analysis according to claim 1, wherein: the dynamic diagram network construction module comprises a network construction sub-module, a time sequence data processing sub-module and a trend analysis prediction sub-module;

9. The system for scientific knowledge of traditional Chinese medicine using artificial intelligence and big data analysis according to claim 1, wherein: the time sequence analysis module comprises a time sequence data collection sub-module, a data analysis sub-module and a trend prediction sub-module;

10. The system for scientific knowledge of traditional Chinese medicine using artificial intelligence and big data analysis according to claim 1, wherein: the knowledge updating mechanism module comprises a knowledge base updating sub-module, an updating strategy making sub-module and an execution effect monitoring sub-module;