AI-based teacher intelligent education literacy training program formulation method and system

Abstract

The present invention provides a method and system for formulating a cultivation plan for teachers' intelligent education literacy based on AI, which relates to the field of artificial intelligence technology, including: obtaining teachers' basic data and intelligent education literacy goals; constructing an intelligent education literacy knowledge map based on text processing of the basic data; evaluating the current level of teachers at various literacy ability nodes based on the intelligent education literacy knowledge map to obtain literacy level evaluation results; performing gap analysis processing based on the literacy level evaluation results and the intelligent education literacy goals to obtain gap analysis results; optimizing processing based on the gap analysis results and the intelligent education literacy knowledge map to obtain a literacy cultivation path; performing resource allocation processing based on the literacy cultivation path to obtain a teacher's intelligent education literacy cultivation plan. The present invention can effectively identify and analyze teachers' ability requirements in different literacy contents by constructing an AI-based teacher intelligent education literacy knowledge map.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method and system for formulating a teacher's intelligent education literacy cultivation program based on AI. Background Art

[0002] With the rapid development of artificial intelligence (AI), the field of education has gradually introduced intelligent means to improve teaching efficiency and personalized learning experience. The application of AI technology in education is mainly concentrated in intelligent teaching systems, personalized learning path recommendations, and educational data analysis. However, in the field of cultivating teachers' intelligent education literacy, the application of AI is still in its early stages. Teachers' intelligent education literacy refers to the ability of teachers to use technical tools, data analysis, and intelligent systems to assist teaching in an intelligent education environment, but most teachers are relatively weak in this area. Existing solutions mostly rely on simple online course recommendation and resource allocation systems. Such systems often recommend learning resources through preset templates or static content, and lack the ability to be personalized and dynamically adjusted. Existing technologies are difficult to accurately plan resource allocation according to the specific literacy level of teachers, and cannot provide real-time feedback, resulting in insufficient optimization of teachers' learning paths, low resource utilization, and difficulty in improving learning effects.

[0003] In response to the above problems, the present invention proposes a method and system for formulating a cultivation plan for teachers' intelligent educational literacy based on AI, aiming to overcome the defects in the prior art and provide a more intelligent, systematic and personalized cultivation path generation and optimization method. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for formulating a teacher's intelligent education literacy training program based on AI to improve the above problems. In order to achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0005] In the first aspect, the present application provides a method for formulating a training program for teachers' intelligent education literacy based on AI, including:

[0006] Obtaining teachers’ basic data and intelligent education literacy goals, including personal information, teaching practice records, and career development goals;

[0007] Performing text processing on the basic data, identifying literacy content and teaching elements related to the teacher's intelligent education literacy, and constructing an intelligent education literacy knowledge graph based on the correlation and influence path between the literacy content and the teaching elements, wherein the nodes in the intelligent education literacy knowledge graph represent various literacy abilities of the teacher, and the edges represent dependencies between abilities;

[0008] According to the intelligent education literacy knowledge graph, the graph neural network is used to infer the knowledge graph, and the current level of the teacher at each literacy ability node is evaluated to obtain a literacy level evaluation result;

[0009] Performing a gap analysis based on the literacy level assessment results and the intelligent education literacy goals, analyzing the gap between the teacher's current literacy and the target literacy and determining the abilities that need to be improved, thereby obtaining a gap analysis result;

[0010] Optimizing the gap analysis results and the intelligent education literacy knowledge graph, using a path optimization algorithm to plan the ability improvement sequence and learning resources to obtain a literacy cultivation path, wherein the literacy cultivation path includes the sequence of each ability improvement and the corresponding learning resource recommendation;

[0011] Resource allocation is performed according to the literacy cultivation path, and the learning plan is planned by utilizing resource priority analysis and time series planning algorithms to obtain a teacher intelligent education literacy cultivation program that includes learning paths, resource allocation, learning progress, and evaluation indicators.

[0012] Secondly, this application also provides an AI-based teacher intelligent education literacy training program formulation system, including:

[0013] An acquisition module is used to acquire the basic data and intelligent education literacy goals of teachers, wherein the basic data includes personal information, teaching practice records and career development goals;

[0014] A processing module, configured to perform text processing according to the basic data, identify literacy content and teaching elements related to the teacher's intelligent education literacy, and construct an intelligent education literacy knowledge graph based on the correlation and influence path between the literacy content and the teaching elements, wherein the nodes in the intelligent education literacy knowledge graph represent various literacy abilities of the teacher, and the edges represent the dependencies between the abilities;

[0015] A reasoning module is used to use a graph neural network to reason about the knowledge graph based on the intelligent education literacy knowledge graph, evaluate the current level of the teacher at each literacy ability node, and obtain a literacy level evaluation result;

[0016] An analysis module is used to perform gap analysis based on the literacy level assessment result and the intelligent education literacy goal, and obtain a gap analysis result by analyzing the gap between the teacher's current literacy and the target literacy and determining the ability that needs to be improved;

[0017] An optimization module, used to perform optimization processing according to the gap analysis results and the intelligent education literacy knowledge map, and use a path optimization algorithm to plan the ability improvement sequence and learning resources to obtain a literacy cultivation path, wherein the literacy cultivation path includes the sequence of each ability improvement and the corresponding learning resource recommendation;

[0018] The planning module is used to perform resource allocation processing according to the literacy cultivation path, and to plan the learning plan by utilizing resource priority analysis and time series planning algorithm to obtain a teacher intelligent education literacy cultivation plan including learning path, resource allocation, learning progress and evaluation indicators.

[0019] The beneficial effects of the present invention are:

[0020] By constructing an AI-based teacher intelligent educational literacy knowledge graph, the present invention can effectively identify and analyze teachers' ability requirements in different literacy contents; by using graph neural networks to perform reasoning and evaluation on the knowledge graph, it can accurately analyze teachers' current literacy levels, and combine the gap analysis results to develop personalized ability improvement paths for each teacher.

[0021] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or be understood by implementing the embodiments of the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0023] Figure 1 A flowchart of a method for formulating a teacher's intelligent education literacy training program based on AI described in an embodiment of the present invention;

[0024] Figure 2 A structural diagram of a system for formulating a training program for teachers' intelligent educational literacy based on AI as described in an embodiment of the present invention.

[0025] Markings in the figure: 1. Acquisition module; 2. Processing module; 21. First processing unit; 22. First calculation unit; 23. First modeling unit; 24. First optimization unit; 3. Reasoning module; 31. Second processing unit; 32. Third processing unit; 33. First classification unit; 34. First integration unit; 4. Analysis module; 41. First matching unit; 42. Fourth processing unit; 43. First judgment unit; 44. Second calculation unit; 5. Optimization module; 51. Fifth processing unit; 52. Second optimization unit; 53. First construction unit; 54. Third optimization unit; 6. Planning module; 61. First analysis unit; 62. Second analysis unit; 63. First planning unit; 64. First design unit. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0027] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings. At the same time, in the description of the present invention, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0028] Example

[0029] This embodiment provides a method for formulating a teacher's intelligent education literacy training program based on AI.

[0030] See also Figure 1 , the figure shows that the method includes steps S100 to S600.

[0031] Step S100: Obtaining the teacher's basic data and intelligent education literacy goals, where the basic data includes personal information, teaching practice records, and career development goals;

[0032] It is understandable that basic data includes teachers’ personal information (such as education, professional background, years of teaching, etc.), teaching practice records (such as course design, classroom activities, student feedback, etc.) and teachers’ career development goals (such as professional title promotion plans, research directions, etc.); intelligent education literacy goals refer to the ability requirements that teachers need to achieve in an intelligent education environment based on their current teaching ability, technology application level, and future career development plans. Specifically, intelligent education literacy goals include teachers’ ability to effectively use technical tools (such as intelligent teaching platforms, data analysis tools, etc.) in teaching, and to assist teaching decisions through intelligent systems (such as personalized teaching feedback, classroom performance evaluation, etc.). These goals not only cover the technical operation level, but also include how teachers can combine data analysis to improve teaching effectiveness, promote student learning, and apply AI technology for teaching innovation.

[0033] Step S200: Perform text processing based on the basic data, identify literacy content and teaching elements related to the teacher's intelligent education literacy, and construct an intelligent education literacy knowledge graph based on the correlation and influence path between the literacy content and the teaching elements. The nodes in the intelligent education literacy knowledge graph represent the various literacy abilities of the teacher, and the edges represent the dependencies between the abilities.

[0034] It should be noted that this step, through the construction of the knowledge graph, can transform the basic data of teachers into a structured capability landscape, which not only displays the current literacy capabilities of teachers, but also reveals the dependencies and transfer paths between these capabilities. In this way, it is possible to clearly identify which capabilities are core capabilities and which capabilities need to be improved first, which lays a solid foundation for subsequent capability assessment and optimization of improvement paths.

[0035] Step S300: Based on the intelligent education literacy knowledge graph, the graph neural network is used to infer the knowledge graph, and the current level of the teacher at each literacy ability node is evaluated to obtain the literacy level evaluation result;

[0036] It is understandable that by reasoning about the intelligent education literacy knowledge graph through graph neural networks, teachers' performance at each literacy node can be efficiently evaluated. Not only a single ability can be focused on, but also the level of teachers' intelligent education literacy can be comprehensively evaluated from the perspective of the overall graph. This graph-based reasoning process can capture the potential dependencies between multiple abilities, making the evaluation more accurate and comprehensive, and suitable for complex education literacy assessment scenarios. In addition, graph neural networks can reason based on the personalized data of different teachers, generate customized literacy assessment results, and provide accurate input data for subsequent ability improvement path planning.

[0037] Step S400: performing gap analysis based on the literacy level assessment results and the intelligent education literacy goals, analyzing the gap between the teacher's current literacy and the target literacy and determining the abilities that need to be improved, thereby obtaining gap analysis results;

[0038] It is understandable that the role of the gap analysis method is to combine multiple dimensions for trade-offs, not only considering the direct gap in ability improvement, but also ensuring the rationality and scientificity of the improvement path through the analysis of the dependence and importance of ability nodes. This makes the improvement of teachers' literacy more systematic and precise, laying a solid foundation for subsequent path optimization and resource allocation.

[0039] Step S500: Optimize the gap analysis results and the intelligent education literacy knowledge map, and use the path optimization algorithm to plan the ability improvement sequence and learning resources to obtain the literacy cultivation path. The literacy cultivation path includes the sequence of each ability improvement and the corresponding learning resource recommendation;

[0040] It can be understood that this step is based on the literacy nodes that need to be improved determined by the gap analysis, combined with the node dependencies in the intelligent education literacy knowledge graph, to design the optimal ability improvement sequence, and recommend corresponding learning resources for each node to ensure the scientificity and efficiency of the improvement path.

[0041] Step S600: perform resource allocation processing according to the literacy cultivation path, plan the learning plan by utilizing resource priority analysis and time series planning algorithm, and obtain a teacher intelligent education literacy cultivation plan including learning path, resource allocation, learning progress and evaluation indicators.

[0042] It should be noted that this step not only optimizes the allocation of learning resources, but also reasonably arranges the time sequence and progress of learning to ensure that teachers can complete capacity improvement within a reasonable time while achieving efficient use of resources.

[0043] Further, step S200 includes step S210 to step S240.

[0044] Step S210: Based on the basic data, a natural language processing algorithm is used to process the text, the text data is converted into a vector representation through a word embedding model, and a named entity recognition algorithm is used to identify key terms and concepts related to the teacher's intelligent education literacy, so as to obtain a vector representation of the literacy content and teaching elements;

[0045] Specifically, we first process the unstructured text data in the basic data. The text data mainly comes from teachers' teaching records, professional development goals, course descriptions, etc. These data exist in the form of natural language, and need to be converted into vector representations that can be mathematically operated through word embedding models (such as BERT or Word2Vec). The advantage of the word embedding model is that it can generate word vectors with semantic similarity by capturing the semantic information of words in context. This process helps to understand the similarity and correlation between words in subsequent analysis.

[0046] Next, the named entity recognition (NER) algorithm is used to automatically extract key terms and concepts related to teachers' intelligent education literacy from the text data. These terms and concepts may include "technology application", "personalized teaching design", "data analysis", etc. The NER algorithm can automatically identify these entities from complex texts and mark their locations in the text. By using predefined dictionaries and model training, the algorithm can distinguish different types of entities and classify them into areas related to intelligent education literacy.

[0047] After text data is vectorized, it can not only capture the semantic information of each key term and concept, but also provide accurate data input for the subsequent knowledge graph node generation. This vectorization process simplifies the complex structure of the language into a computable numerical form, allowing subsequent reasoning and analysis to operate in mathematical space.

[0048] Step S220: Based on the vectorized representation, semantic relevance modeling is performed using a semantic similarity calculation algorithm, by calculating the similarity between different literacy contents and teaching elements, and selecting correlation combinations based on a preset threshold, to obtain a correlation matrix between literacy contents and teaching elements;

[0049] It should be noted that this step first calculates the similarity between literacy content and teaching elements by cosine similarity. Cosine similarity is a common algorithm for measuring the angular similarity between two vectors, and is suitable for identifying semantically similar content after text vectorization. Its calculation process is based on the angle between the two vectors. The smaller the angle, the higher the similarity between the two contents. Then the calculated similarity results are filtered according to the preset similarity threshold. The role of the preset threshold is to filter out those content combinations with low similarity or irrelevant content, ensuring that the final correlation matrix only retains those literacy content and teaching elements with high semantic relevance. Through this screening process, noise information can be effectively reduced and the accuracy of the data can be improved. The resulting correlation matrix can be regarded as a network structure, and each element in the matrix represents the similarity between two literacy contents or teaching elements. Through this semantic correlation modeling, not only can the direct relationship between the contents be constructed, but also the hidden semantic associations can be excavated.

[0050] Step S230: Based on the correlation matrix, the dependency path between the literacy content and the teaching elements is modeled by the shortest path algorithm in graph theory, and the influence path diagram between the literacy content and the teaching elements is obtained by analyzing the transitivity and hierarchical relationship between different contents, where the nodes represent each specific literacy content and the edges represent the dependency paths between the literacy contents;

[0051] It can be understood that this step first uses the similarity information in the association matrix as the "edge weight" input in graph theory, which represents the strength of association between literacy content or teaching elements. Each node represents a specific literacy ability or teaching element, such as "data analysis ability" or "personalized teaching design ability". At this time, the best path from one node to another is determined. This path represents the dependency relationship between abilities or elements. For this purpose, the shortest path algorithm, such as the Dijkstra algorithm or the Bellman-Ford algorithm, is used to calculate the optimal path from one node to another. This algorithm can find the path with the least cost in the graph structure (in this scenario, the cost refers to the dependency strength or difficulty between literacy content).

[0052] The shortest path algorithm can determine the transitivity between different literacy contents based on the weight information in the correlation matrix. For example, assuming that there is a dependency relationship between "basic technical operation ability" and "smart teaching platform application", the shortest path algorithm can identify which ability nodes (such as "basic technical operation") teachers need to master first in order to smoothly transition to higher-level abilities (such as "smart teaching platform application"). This analysis not only reveals the direct dependency between literacy abilities, but also constructs the path that teachers need to improve in sequence through hierarchical relationships. In addition, by analyzing the transitivity and hierarchical relationships in the path, a complete impact path diagram can be further constructed. In this diagram, the more basic literacy content is at a lower level, while the high-level or complex abilities are at a higher level. This hierarchical path can clearly show which ability nodes teachers need to improve in sequence in order to ultimately achieve their smart education literacy goals. For example, before teachers master "personalized teaching design ability", they may first need to have "teaching data analysis ability". This type of hierarchical dependency is clarified through the shortest path algorithm and ultimately reflected in the impact path diagram.

[0053] Step S240: According to the impact path diagram, the graph neural network is used to optimize the graph structure, and the nodes are embedded and learned through the graph convolutional network to optimize the node relationship between each literacy content and teaching elements to obtain the intelligent education literacy knowledge graph.

[0054] It should be noted that the embedding learning process of the graph convolutional network is completed through multi-layer convolution operations. Each layer of convolution collects information from neighboring nodes and aggregates it into the current node, and then updates the node representation vector layer by layer. Multi-layer convolution can ensure that each node can not only capture the information of directly adjacent nodes during the embedding learning process, but also integrate the information of multi-hop neighbors to capture a wider range of dependencies. Another key step in the graph neural network is node relationship optimization. The graph neural network continuously adjusts the weights between nodes during the training process to ensure that the dependencies in the impact path diagram can more accurately reflect the actual literacy improvement path.

[0055] Further, step S300 includes step S310 to step S340.

[0056] Step S310: Initialize the graph neural network according to the intelligent education literacy knowledge graph, assign an initial feature vector to the relationship between each literacy content node and its adjacent nodes, and use the initial parameter setting of the graph convolutional network to embed the nodes and their edges to obtain the initial feature matrix of the graph neural network;

[0057] It can be understood that, first, an initial feature vector is assigned to each literacy content node in the knowledge graph. These feature vectors can reflect the basic information of the ability or teaching element represented by each node. The feature vector of the node is generated by the embedded information obtained from the text processing results and vectorized representation. For example, in the previous steps, natural language processing has been used to process and vectorize literacy content such as "data analysis ability" or "smart teaching platform application", so the feature vectors of these nodes can represent their semantic information in multidimensional space. The feature vectors of adjacent nodes are also allocated based on the similarity information in the correlation matrix, so that adjacent nodes share part of the feature information, reflecting their dependency.

[0058] Next, set the initial parameters for the graph convolutional network. The main function of the graph convolutional network is to aggregate the features of each node and its neighboring nodes through convolution operations. To achieve this process, first initialize the parameters for each node and its connected edges. These parameters include the initial weight of the node, the weight of the edge, and the learning rate of the convolution layer. At this stage, random initialization or pre-trained model initialization can be used to assign initial values ​​to these parameters.

[0059] Once the initial feature vectors and weights of nodes and edges are assigned, the initial feature matrix of the graph neural network is calculated using these initial features. The rows of this matrix represent each competency node, and the columns represent the initial features of the node in each dimension. The initial feature matrix is ​​the basis for further learning and updating of the graph convolutional network. In the subsequent graph convolution process, the feature vector of the node will be continuously updated with multiple rounds of convolution and information aggregation of neighboring nodes, gradually integrating information from surrounding nodes, and finally forming a deeper embedded representation of the node.

[0060] Step S320: Perform graph convolution processing according to the initial feature matrix, aggregate and transmit the information of each node and its adjacent nodes through a multi-layer graph convolution network, update the feature vector of each node, and obtain a node feature matrix reflecting the literacy content node and its related nodes;

[0061] It should be noted that, first of all, the working principle of the graph convolutional network is to combine the feature vector of a node with the features of its neighboring nodes through information aggregation and transmission, so that the features of each node not only rely on its own information, but also integrate the information of its neighboring nodes. In this process, the graph convolution operation will gradually transfer the features of the neighboring nodes to the target node layer by layer, and update the feature representation of the target node after weighted summation. Specifically, the convolution operation can be defined as: for each layer, the node The eigenvector of Its neighboring nodes The eigenvector of According to weight Perform weighted summation and update the node feature vector It can be expressed as:

[0062] ;

[0063] in, Representation Node The set of neighbor nodes; It is a nonlinear activation function to ensure the nonlinear characteristics of information transmission. Through this operation, each node will fuse the information of neighboring nodes layer by layer, and gradually capture a wider range of dependencies in the entire graph after multiple layers of convolution; Indicates the current time.

[0064] In specific applications, the features of literacy content nodes (such as "data analysis ability") will be updated through information transmission from adjacent nodes (such as "technical tool operation ability" or "teaching feedback ability"). Multi-layer convolutional networks allow information to be transmitted step by step, ensuring that information from more distant nodes can also be obtained by the target node through multi-hop propagation. The representation of such a literacy node is no longer isolated, but integrates the information of its surrounding dependent nodes. In addition, graph convolutional networks can optimize the way of information transmission through weight adjustment, so that more important neighbor nodes contribute more to the update of the target node, while less important nodes have less impact. This weighted aggregation ensures that after multi-layer convolution, the dependency relationship between nodes can be more reasonably reflected in the node features. As multi-layer convolutions proceed, each layer of the convolutional network will learn more complex and deep relationships between nodes and their adjacent nodes, and finally form an updated node feature matrix. This matrix not only contains the characteristics of each literacy node itself, but also the transmission influence of its adjacent nodes on its characteristics, thereby realizing the capture and representation of global dependencies.

[0065] Step S330: Based on the node feature matrix, the classification layer of the graph neural network is used to classify the literacy nodes. The classification label of each node is obtained by performing a classification operation on the final feature vector of each node and predicting the teacher's ability level at each literacy content node.

[0066] It can be understood that after the graph neural network has been processed by multiple layers of graph convolution, the node feature matrix generated already contains the final feature representation of each node. These feature vectors not only reflect the properties of each literacy node, but also contain the dependency and association features with its adjacent nodes, so they can more comprehensively describe the characteristics of each competency node. Next, the features of these nodes are further processed through the classification layer to predict the teacher's ability level at each literacy content node.

[0067] The classification layer usually consists of one or more fully connected layers, and finally the nodes are classified through the softmax activation function. The core operation in the classification process is to map the final feature vector of each node to a discrete classification label space. Assuming that the teacher's literacy level is divided into three categories (such as "high", "medium", and "low"), the softmax function will calculate the probability of each node belonging to each category. The formula for softmax is:

[0068] ;

[0069] in, It is the output value of the node feature vector after passing through the classification layer; Indicates the total number of categories for classification; Represents a given input feature vector ,node Classified as Probability of class; is a natural constant; Represents the input node After passing through the classification layer and The unnormalized score values ​​associated with the class. Softmax normalizes these output values ​​into a probability distribution. By selecting the class with the highest probability, a final classification label is assigned to the node. Each label represents the teacher's current level on that particular literacy competency.

[0070] Furthermore, in actual operation, different classification schemes can be set according to specific needs. It can be a binary classification (such as "capable" or "incapable"), or a multi-level classification (such as "primary", "intermediate", "advanced"), depending on the complexity and refinement of teaching literacy. Through this node classification process, a corresponding ability label can be generated for each literacy content node. This process not only provides the current ability status of each node, but also provides quantitative data support for subsequent ability gap analysis.

[0071] Step S340: perform integration processing according to the classification labels, integrate the evaluation results of each node through the weighted average method, and combine the importance of the node and the influence of its adjacent nodes to obtain the overall evaluation result of the teacher's literacy level.

[0072] It should be noted that the importance of a node refers to the relative weight of each ability in the overall literacy assessment of teachers. In actual application scenarios, different literacy abilities have different effects on teachers' teaching performance, so the weight of some nodes will be higher than that of other nodes. For example, for a teacher engaged in technology-oriented teaching, "data analysis ability" may be the core of his professional performance, so the importance of this node is relatively high and occupies a large proportion in the overall literacy assessment; while for a teacher who focuses on classroom teaching, "classroom management ability" may be more important than "data analysis ability". To reflect this, the weight of each node will be dynamically set according to factors such as the teacher's job responsibilities, teaching style, and course requirements to ensure that more critical abilities are appropriately amplified in the overall assessment. Importance can be determined in a variety of ways, such as dynamic allocation based on job requirements, career development goals, course nature, etc., or it can be configured based on historical performance data, expert ratings, etc. The influence of adjacent nodes means that in the knowledge graph, a certain ability node is not only affected by its own characteristics, but also by other nodes (adjacent nodes) directly associated with it. This is because the various abilities of teachers are not isolated, and there is often a dependency relationship between abilities. For example, a teacher's performance in "intelligent teaching tool operation ability" directly depends on his "basic technical operation ability". If the "basic technical operation ability" is weak, although the "intelligent teaching tool operation ability" node itself may be given a higher weight, the weakness of the adjacent node (i.e., "basic technical operation ability") will have a negative impact on the node and reduce its final evaluation result. This influence reflects the complex hierarchy of educational ability: when basic ability is insufficient, it is difficult for high-order ability to play its due role. On the contrary, when basic ability is solid, the performance of high-order ability is often amplified. This influence of adjacent nodes ensures the rationality of the evaluation model and avoids the evaluation being too one-sided. This makes the evaluation results more holistic and systematic, and can accurately reflect the comprehensive performance of teachers in multidimensional literacy.

[0073] Further, step S400 includes step S410 to step S440.

[0074] Step S410: matching the literacy level assessment result and the intelligent education literacy target, by comparing the teacher's current ability level at each literacy content node with the target level, and using the Euclidean distance method to calculate the similarity between the current literacy level and the target literacy, to obtain a preliminary matching result;

[0075] It is understandable that this step quantifies the difference between the teacher's current ability and the target ability at each literacy content node through the precise geometric measurement method of Euclidean distance. This quantification method is not only simple and effective, but also can process multi-dimensional and multi-node literacy ability data, providing a clear gap assessment standard. Through the preliminary matching results, it is possible to intuitively identify at which nodes the teacher's ability is close to the target and at which nodes further improvement is needed.

[0076] Step S420: Perform gap quantification processing based on the preliminary matching results, calculate the gap value of each literacy content node and perform quantitative analysis to obtain the gap quantification result of each node;

[0077] It should be noted that this step aims to accurately quantify the difference between the teacher's current literacy and the target literacy by further quantifying the gap of the preliminary matching results. The gap value of each literacy content node is obtained by mathematical calculation, and the gap value is normalized by a standardized formula to measure the size of the gap between 0 and 1. The closer the gap value is to 1, the greater the gap between the current ability level of the node and the target, and it needs to be improved first; the closer the gap value is to 0, the closer the ability level of the node is to the target.

[0078] Step S430: Perform capability improvement demand determination processing based on the gap quantification result, determine whether each literacy content node needs to be improved by setting a threshold, and use a dynamic programming algorithm to determine the optimal capability improvement strategy to obtain a capability node set;

[0079] It is understandable that due to the dependencies between each literacy content node, the improvement of certain abilities requires other related basic abilities first, so the simple one-by-one improvement method is not efficient enough. In order to optimize this process, a dynamic programming algorithm is used to determine the optimal ability improvement order and strategy on a global scale. The basic idea of ​​the dynamic programming algorithm is to decompose a large problem into a series of related sub-problems, and gradually solve these sub-problems to finally obtain the global optimal solution. In this scenario, an optimal ability improvement path is designed based on the dependencies and priorities of the literacy content nodes. For example, if the "smart teaching platform application ability" depends on the "basic technical operation ability" and "data analysis ability", the basic ability is improved first, and then the high-level ability is gradually improved. The introduction of dynamic programming ensures that the improvement of each literacy ability node is carried out according to the optimal path, avoiding inefficient learning caused by the failure to consider ability dependencies.

[0080] Through the above judgment and dynamic planning processing, a set of ability nodes that need to be improved is finally generated. This set contains the literacy content nodes that teachers need to focus on improving in the future. This set not only indicates the abilities that need to be improved, but also determines the optimal improvement order to ensure the consistency and effectiveness of the ability improvement process.

[0081] Step S440: Perform priority analysis based on the capability node set, calculate the priority order of each node based on the importance, difficulty and influence of the node on the overall quality, and obtain the final gap analysis result.

[0082] It is understandable that the result of the gap analysis is a ranked list of competency nodes, each with a clear priority order. The gap analysis results show the competencies that teachers need to improve most at the current stage and the best order to improve these competencies.

[0083] Further, step S500 includes step S510 to step S540.

[0084] Step S510: Process the dependency relationship of capability nodes according to the gap analysis results and the intelligent education literacy knowledge graph, analyze the interdependence between the capability nodes in the knowledge graph, and use the topological sorting algorithm to determine the priority of the improvement path to obtain a preliminary capability improvement order;

[0085] It should be noted that topological sorting is a sorting algorithm for directed acyclic graphs (DAGs) that can determine a linear sequence for a set of nodes with clear dependencies, ensuring that before processing a node, all predecessor nodes that depend on the node have been processed. In this scenario, the intelligent education literacy knowledge graph can be regarded as a directed acyclic graph, in which the edges between each node represent the ability dependency. Topological sorting assigns priority to each node through dependencies to ensure that before improving a certain high-level ability node, the related basic ability nodes have been fully improved. This sorting method ensures that teachers can improve their abilities step by step, without ignoring basic abilities, which will lead to low learning efficiency or difficulty in mastering high-level abilities.

[0086] The specific algorithm steps of topological sorting are as follows:

[0087] Step S511, performing in-degree allocation processing according to the intelligent education literacy knowledge graph to obtain an initial in-degree result;

[0088] It can be understood that according to the directed edges in the knowledge graph, the number of predecessor nodes (i.e., in-degree) of each literacy content node is determined, and a node with an in-degree of 0 indicates that it is independent of other nodes and serves as a basic capability node.

[0089] Step S512: Perform basic node selection processing according to the initial in-degree result to obtain a basic node set;

[0090] It should be noted that this step uses a queue algorithm to queue nodes with an in-degree of 0 and process these basic nodes in turn. In this way, the basic ability nodes that teachers currently need to improve first are identified, and the first batch of ability improvement targets are obtained. These nodes will become the starting point for path planning.

[0091] Step S513: perform in-degree update and topological sorting processing according to the basic node set and the gap analysis results to obtain a preliminary capability improvement order.

[0092] Specifically, this step adopts a breadth-first search algorithm to process nodes with an in-degree of 0 in turn, reduce the in-degree of adjacent nodes, and determine their sorting order according to the size of the gap.

[0093] Step S520: Optimize the path according to the preliminary capability improvement sequence, analyze the cost and benefit of each capability node, dynamically plan the optimal capability improvement path, and obtain the capability improvement sequence;

[0094] It should be noted that in this step, the improvement cost of each node is first evaluated, including learning time and resource investment and its corresponding benefits, that is, the contribution of improving the node to the overall improvement of teachers' literacy. Through this analysis, low-cost and high-benefit nodes are identified for priority processing. Then, the dynamic programming algorithm is used to gradually solve each improvement decision in the path on the basis of phased optimization to ensure that the final generated path is the global optimal rather than the local optimal. The advantage of dynamic programming is that it can be calculated recursively to ensure that the most effective improvement path is selected within a given time or resource budget. In the end, an optimal ability improvement order arranged by priority is obtained to ensure that teachers can achieve maximum literacy improvement with the most reasonable resource investment and the shortest time.

[0095] Step S530: constructing a scenario model according to the ability improvement sequence, simulating the ability improvement process of teachers in an actual teaching environment, constructing at least two scenario models using a scenario simulation algorithm, and judging the rationality of the path to obtain path correction suggestions;

[0096] Specifically, firstly, according to the order of ability improvement, the process of teacher's ability improvement in actual teaching is constructed. The process of teachers gradually improving their intelligent education literacy in different teaching situations is simulated by scenario simulation algorithm, and at least two different scenario modes are constructed. Each scenario mode will combine different teaching variables, such as course type, student feedback, frequency of use of teaching technology, etc., to form the trajectory of teacher's ability improvement in the teaching process. Then, by comparing these different scenario modes, the rationality of the ability improvement path in actual operation is analyzed. Furthermore, the scenario simulation not only takes into account the gradual improvement of teachers' abilities, but also introduces limiting factors in the actual teaching environment such as teaching load, resource availability, and learning time allocation to ensure that the path is not only optimal in theory, but also can be efficiently implemented in reality. Finally, according to the results of scenario simulation, path correction suggestions are put forward. If the improvement path under a certain scenario mode shows better rationality and effect, the initial ability improvement order will be adjusted accordingly. Through this scenario mode construction and optimization, it is ensured that the final path is not only reasonable in theory, but also can be effectively implemented in the actual teaching environment.

[0097] Step S540: Optimize the path according to the path correction suggestion, and obtain the literacy development path by comprehensively considering the priority order of each capability node, the feedback information in the scenario mode, and the reasonable arrangement of learning time.

[0098] It is understandable that through the comprehensive treatment of priorities, situational feedback and time arrangements, it is ensured that the capacity improvement path can maximize the teacher's capacity improvement effect within the most reasonable time, while reducing resource waste and time pressure.

[0099] Further, step S600 includes step S610 to step S640.

[0100] Step S610: Perform resource demand analysis according to the literacy cultivation path, analyze the resource type and quantity required for each capability improvement node, and use the content-based collaborative filtering algorithm to match the specific learning resources required for each node in combination with the teacher's existing resource library to obtain the resource matching result for each node;

[0101] It should be noted that this step first analyzes the type and quantity of resources required for each capability improvement node. Resource types include online courses, textbooks, case studies, experimental activities, etc., while the number of resources depends on the complexity of the node and the length of learning. Through this analysis, the specific resource requirements of each capability node in the learning process are determined. Then, combined with the teacher's existing resource library, a content-based collaborative filtering algorithm is used for resource matching. The collaborative filtering algorithm matches resources in the resource library that are similar to the requirements of the target capability node by analyzing the content characteristics of the resources. For example, if a node requires the improvement of "data analysis ability", the system will recommend courses or textbooks related to data analysis based on the theme, difficulty and learning objectives of the resources. The advantage of the collaborative filtering algorithm is that it can accurately find learning resources that meet the needs of the node based on the similarity of the content, thereby improving the accuracy of resource matching. Finally, the resource matching result is obtained to ensure that the learning resource requirements of each node can be met and support the process of improving the teacher's ability.

[0102] Step S620: Perform resource priority analysis and processing according to the resource matching result, comprehensively evaluate the urgency and resource scarcity of each capability improvement node, and use the hierarchical analysis method to prioritize the resources to obtain the resource priority order corresponding to each capability improvement node;

[0103] It should be noted that the urgency is determined by the gap between the teacher's current literacy and the target literacy. The larger the gap and the more urgent the need for improvement, the higher the urgency of the ability node. For example, if the gap of a node (such as "intelligent teaching platform use ability") is large and has a significant impact on the overall teaching performance, then the node is considered to be a priority for improvement. This evaluation process is determined by analyzing the gap analysis results generated earlier and the actual needs of each ability node. Resource scarcity refers to the availability and difficulty of using certain resources. If the learning resources required by a node are less in the resource library, or it takes a long time to obtain or complete, the resource is considered a scarce resource, which in turn affects the priority of resource allocation. For nodes with scarce resources, their resource priorities are specially handled according to the situation of the resource library to ensure that scarce resources can be reasonably scheduled and allocated. Finally, the analytic hierarchy process (AHP) is used to prioritize the resource allocation of each node based on urgency and resource scarcity. The analytic hierarchy process is a multi-criteria decision-making method that can decompose complex decision problems into multiple levels for weight allocation and evaluation. In this step, urgency and resource scarcity are used as the two main criteria for decision-making, each node is scored, and a priority is assigned to each capability node by calculating the weighted score. The analytic hierarchy process can effectively solve the multi-factor trade-off problem faced in resource allocation and ensure that key nodes can obtain resources first.

[0104] Step S630: Perform time series planning according to the resource priority order, use the time series planning algorithm to combine the teacher's available learning time and learning resource allocation requirements, formulate the learning order and learning time of each ability improvement node, and obtain the learning progress schedule;

[0105] Specifically, first determine the learning order of each capacity improvement node according to the generated resource priority order. Based on this order, plan the allocation of learning time. Since the learning resource requirements and urgency of different nodes are different, it is necessary to reasonably arrange the learning time of each node in combination with the weight of the teacher's learning tasks. Next, use the teacher's available learning time as a constraint to perform time series planning. Teachers' learning time is often limited, especially when teaching tasks are busy, reasonable time allocation becomes crucial. By analyzing the teacher's daily schedule, determine which time periods can be used for learning, and allocate reasonable learning time according to the amount of learning tasks and resource requirements of different nodes.

[0106] The time series planning algorithm can dynamically adjust the order and time allocation of learning tasks according to the continuity and distribution of time. Specifically, the resource priority order is used as the initial input, and combined with the teacher's learning time limit, the most reasonable learning schedule is optimized. Through this planning, not only is the learning order of each node guaranteed to meet the priority, but also the time allocation is ensured to adapt to the actual teaching environment and learning load. Finally, a learning schedule is generated, that is, the specific learning time and sequence planning of each capacity improvement node. This schedule not only takes into account the urgency and resource requirements of the node, but also fully combines the teacher's learning time to ensure that the capacity improvement process can proceed in an orderly manner.

[0107] Step S640: Design and process evaluation indicators according to the learning progress schedule, use regression analysis algorithm to combine the teacher's learning goals and expected results, design phased evaluation and final evaluation, and obtain a learning effect evaluation system for each ability improvement node.

[0108] It should be noted that this step first analyzes the learning schedule and designs corresponding evaluation indicators based on the learning tasks and goals of each capability node to ensure that the progress and performance of teachers in the learning process can be measured. Then, a regression analysis algorithm is used to predict which learning behaviors will lead to better learning results by modeling the relationship between historical data and learning goals, and then determine the time points and evaluation criteria for staged evaluations. Regression analysis also helps identify key evaluation factors, such as the correlation between learning time investment and task completion, to ensure the accuracy of the evaluation design. Finally, a comprehensive learning effect evaluation system is generated for each capability improvement node, including staged evaluations to adjust learning strategies in real time, and final evaluations to verify the achievement of overall capability improvement. This evaluation system can accurately reflect the learning effects of teachers and provide a scientific basis for capability improvement.

[0109] Example

[0110] like Figure 2 As shown, this embodiment provides a system for formulating a training program for teachers' intelligent education literacy based on AI, and the system includes:

[0111] Acquisition module 1 is used to obtain teachers’ basic data and intelligent education literacy goals. The basic data includes personal information, teaching practice records, and career development goals.

[0112] Processing module 2 is used to perform text processing based on basic data, identify literacy content and teaching elements related to teachers' intelligent education literacy, and construct an intelligent education literacy knowledge graph based on the correlation and influence path between literacy content and teaching elements. The nodes in the intelligent education literacy knowledge graph represent the various literacy abilities of teachers, and the edges represent the dependencies between abilities.

[0113] Reasoning module 3 is used to use graph neural network to reason about the knowledge graph based on the intelligent education literacy knowledge graph, evaluate the current level of teachers at each literacy ability node, and obtain the literacy level evaluation result;

[0114] Analysis module 4 is used to perform gap analysis based on the literacy level assessment results and the intelligent education literacy goals, and obtain the gap analysis results by analyzing the gap between the teacher's current literacy and the target literacy and determining the abilities that need to be improved;

[0115] Optimization module 5 is used to optimize the gap analysis results and the intelligent education literacy knowledge map, and use the path optimization algorithm to plan the ability improvement sequence and learning resources to obtain the literacy cultivation path. The literacy cultivation path includes the sequence of each ability improvement and the corresponding learning resource recommendation;

[0116] Planning module 6 is used to allocate resources according to the literacy cultivation path. By using resource priority analysis and time series planning algorithm to plan the learning plan, a teacher intelligent education literacy cultivation plan including learning path, resource allocation, learning progress and evaluation indicators is obtained.

[0117] In a specific embodiment disclosed in the present application, the processing module 2 includes:

[0118] The first processing unit 21 is used to perform text processing using a natural language processing algorithm based on the basic data, convert the text data into a vector representation through a word embedding model, and use a named entity recognition algorithm to identify key terms and concepts related to the teacher's intelligent education literacy, and obtain a vector representation of literacy content and teaching elements;

[0119] The first calculation unit 22 is used to perform semantic relevance modeling based on the vectorized representation using a semantic similarity calculation algorithm, calculate the similarity between different literacy contents and teaching elements, and screen out correlation combinations based on a preset threshold to obtain a correlation matrix between literacy contents and teaching elements;

[0120] The first modeling unit 23 is used to model the dependency path between the literacy content and the teaching elements according to the correlation matrix by using the shortest path algorithm in graph theory, and obtain the influence path diagram between the literacy content and the teaching elements by analyzing the transitivity and hierarchical relationship between different contents, wherein the nodes represent each specific literacy content and the edges represent the dependency paths between the literacy contents;

[0121] The first optimization unit 24 is used to optimize the graph structure using a graph neural network according to the impact path diagram, embed the nodes through a graph convolutional network, optimize the node relationship between each literacy content and teaching elements, and obtain an intelligent education literacy knowledge graph.

[0122] In a specific implementation disclosed in the present application, the reasoning module 3 includes:

[0123] The second processing unit 31 is used to perform initialization processing of the graph neural network according to the intelligent education literacy knowledge graph, by allocating an initial feature vector to the relationship between each literacy content node and its adjacent nodes, and using the initial parameter setting of the graph convolutional network to embed the nodes and their edges to obtain an initial feature matrix of the graph neural network;

[0124] The third processing unit 32 is used to perform graph convolution processing according to the initial feature matrix, aggregate and transmit the information of each node and its adjacent nodes through a multi-layer graph convolution network, update the feature vector of each node, and obtain a node feature matrix reflecting the literacy content node and its related nodes;

[0125] The first classification unit 33 is used to classify the literacy nodes using the classification layer of the graph neural network according to the node feature matrix, and obtain the classification label of each node by performing classification operations on the final feature vector of each node and predicting the ability level of the teacher on each literacy content node;

[0126] The first integration unit 34 is used to perform integration processing according to the classification labels, integrate the evaluation results of each node through the weighted average method, and combine the importance of the node and the influence of its adjacent nodes to obtain the overall evaluation result of the teacher's literacy level.

[0127] In a specific embodiment disclosed in the present application, the analysis module 4 includes:

[0128] The first matching unit 41 is used to perform matching processing according to the literacy level assessment result and the intelligent education literacy target, by comparing the teacher's current ability level at each literacy content node with the target level, and using the Euclidean distance method to calculate the similarity between the current literacy level and the target literacy, to obtain a preliminary matching result;

[0129] The fourth processing unit 42 is used to perform gap quantification processing according to the preliminary matching result, by calculating the gap value of each literacy content node and performing quantitative analysis to obtain the gap quantification result of each node;

[0130] The first determination unit 43 is used to determine the need for capability improvement according to the gap quantification result, determine whether each literacy content node needs to be improved by setting a threshold, and use a dynamic programming algorithm to determine the optimal capability improvement strategy to obtain a capability node set;

[0131] The second calculation unit 44 is used to perform priority analysis based on the capability node set, calculate the priority order of each node based on the importance, difficulty and influence of the node on the overall quality, and obtain the final gap analysis result.

[0132] In a specific implementation disclosed in the present application, the optimization module 5 includes:

[0133] The fifth processing unit 51 is used to process the dependency relationship of the capability nodes according to the gap analysis results and the intelligent education literacy knowledge graph, analyze the interdependence between the capability nodes in the knowledge graph, and use the topological sorting algorithm to determine the priority of the improvement path to obtain a preliminary capability improvement order;

[0134] The second optimization unit 52 is used to perform path optimization processing according to the preliminary capability improvement sequence, dynamically plan the optimal capability improvement path by analyzing the improvement cost and benefit of each capability node, and obtain the capability improvement sequence;

[0135] The first construction unit 53 is used to construct a scenario model according to the ability improvement sequence, construct at least two scenario models using a scenario simulation algorithm by simulating the ability improvement process of teachers in an actual teaching environment, and judge the rationality of the path to obtain a path correction suggestion;

[0136] The third optimization unit 54 is used to perform path optimization processing according to the path correction suggestion, and obtain the literacy cultivation path by comprehensively considering the priority order of each capability node, the feedback information in the scenario mode and the reasonable arrangement of learning time.

[0137] In a specific implementation disclosed in the present application, the planning module 6 includes:

[0138] The first analysis unit 61 is used to perform resource demand analysis according to the literacy cultivation path, analyze the resource type and quantity required for each capability improvement node, and match the specific learning resources required for each node using a content-based collaborative filtering algorithm in combination with the teacher's existing resource library to obtain a resource matching result for each node;

[0139] The second analysis unit 62 is used to perform resource priority analysis according to the resource matching result, and to prioritize resources by using the hierarchical analysis method by comprehensively evaluating the urgency and resource scarcity of each capability improvement node, so as to obtain the resource priority order corresponding to each capability improvement node;

[0140] The first planning unit 63 is used to perform time series planning processing according to the resource priority order, and use the time series planning algorithm to combine the teacher's available learning time and learning resource allocation requirements to formulate the learning order and learning time of each ability improvement node to obtain a learning progress schedule;

[0141] The first design unit 64 is used to design and process evaluation indicators according to the learning progress schedule, use regression analysis algorithms to combine teachers' learning goals and expected results, design phased evaluations and final evaluations, and obtain a learning effect evaluation system for each ability improvement node.

[0142] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for formulating a teacher's intelligent education literacy training program based on AI, characterized in that: include: Obtaining teachers’ basic data and intelligent education literacy goals, including personal information, teaching practice records, and career development goals; Performing text processing on the basic data, identifying literacy content and teaching elements related to the teacher's intelligent education literacy, and constructing an intelligent education literacy knowledge graph based on the correlation and influence path between the literacy content and the teaching elements, wherein the nodes in the intelligent education literacy knowledge graph represent various literacy abilities of the teacher, and the edges represent dependencies between abilities; According to the intelligent education literacy knowledge graph, the graph neural network is used to infer the knowledge graph, and the current level of the teacher at each literacy ability node is evaluated to obtain a literacy level evaluation result; Performing a gap analysis based on the literacy level assessment results and the intelligent education literacy goals, analyzing the gap between the teacher's current literacy and the target literacy and determining the abilities that need to be improved, thereby obtaining a gap analysis result; Optimizing the gap analysis results and the intelligent education literacy knowledge graph, using a path optimization algorithm to plan the ability improvement sequence and learning resources to obtain a literacy cultivation path, wherein the literacy cultivation path includes the sequence of each ability improvement and the corresponding learning resource recommendation; Perform resource allocation processing according to the literacy cultivation path, plan the learning plan by using resource priority analysis and time series planning algorithm, and obtain a teacher intelligent education literacy cultivation plan including learning path, resource allocation, learning progress and evaluation indicators; Among them, text processing is performed according to the basic data, and the literacy content and teaching elements related to the teacher's intelligent education literacy are identified, and based on the correlation and influence path between the literacy content and the teaching elements, an intelligent education literacy knowledge map is constructed, including: Based on the basic data, a natural language processing algorithm is used to process text, the text data is converted into a vector representation through a word embedding model, and a named entity recognition algorithm is used to identify key terms and concepts related to teachers' intelligent education literacy, so as to obtain a vector representation of literacy content and teaching elements; According to the vectorized representation, semantic relevance modeling is performed using a semantic similarity calculation algorithm, by calculating the similarity between different literacy contents and teaching elements, and selecting correlation combinations based on a preset threshold, to obtain a correlation matrix between literacy contents and teaching elements; According to the correlation matrix, the dependency path between the literacy content and the teaching elements is modeled by the shortest path algorithm in graph theory, and the influence path diagram between the literacy content and the teaching elements is obtained by analyzing the transitivity and hierarchical relationship between different contents, in which the nodes represent each specific literacy content and the edges represent the dependency paths between the literacy contents; According to the impact path diagram, the graph structure is optimized using a graph neural network, and the nodes are embedded and learned through a graph convolutional network to optimize the node relationship between each literacy content and teaching elements, thereby obtaining an intelligent education literacy knowledge graph; Among them, according to the intelligent education literacy knowledge graph, the graph neural network is used to reason the knowledge graph, and the current level of the teacher at each literacy ability node is evaluated to obtain the literacy level evaluation result, including: Initializing the graph neural network according to the intelligent education literacy knowledge graph, allocating initial feature vectors to the relationship between each literacy content node and its adjacent nodes, and using the initial parameter setting of the graph convolutional network to embed the nodes and their edges, thereby obtaining an initial feature matrix of the graph neural network; Performing graph convolution processing according to the initial feature matrix, aggregating and transmitting information of each node and its adjacent nodes through a multi-layer graph convolution network, updating the feature vector of each node, and obtaining a node feature matrix reflecting the literacy content node and its related nodes; According to the node feature matrix, the classification layer of the graph neural network is used to classify the literacy nodes, and the classification label of each node is obtained by performing classification operations on the final feature vector of each node and predicting the ability level of the teacher on each literacy content node; The classification labels are integrated, and the evaluation results of each node are integrated by weighted average method. The importance of the node and the influence of its adjacent nodes are combined to obtain the overall evaluation result of the teacher's literacy level.

2. The method for formulating a teacher's intelligent education literacy training program based on AI according to claim 1 is characterized in that: A gap analysis is performed based on the literacy level assessment results and the intelligent education literacy goals. By analyzing the gap between the teacher's current literacy and the target literacy and determining the abilities that need to be improved, a gap analysis result is obtained, including: Matching is performed according to the literacy level assessment result and the intelligent education literacy goal, by comparing the teacher's current ability level at each literacy content node with the target level, and using the Euclidean distance method to calculate the similarity between the current literacy level and the target literacy, to obtain a preliminary matching result; Performing gap quantification processing according to the preliminary matching results, by calculating the gap value of each literacy content node and performing quantitative analysis, obtaining the gap quantification result of each node; According to the quantified results of the gap, the ability improvement needs are determined and processed, and a threshold is set to determine whether each literacy content node needs to be improved, and a dynamic programming algorithm is used to determine the optimal ability improvement strategy to obtain a set of ability nodes; Priority analysis is performed based on the set of capability nodes, and the priority order of each node is calculated based on the importance, difficulty and influence of the node on the overall quality, to obtain the final gap analysis result.

3. The method for formulating a teacher's intelligent education literacy training program based on AI according to claim 1 is characterized in that: According to the gap analysis results and the intelligent education literacy knowledge map, optimization processing is performed, and the ability improvement sequence and learning resources are planned using a path optimization algorithm to obtain a literacy cultivation path, including: Processing the dependency relationship of capability nodes according to the gap analysis results and the intelligent education literacy knowledge graph, analyzing the interdependence between the capability nodes in the knowledge graph, and using a topological sorting algorithm to determine the priority of the improvement path, thereby obtaining a preliminary capability improvement order; Perform path optimization processing according to the preliminary capability improvement sequence, dynamically plan the optimal capability improvement path by analyzing the improvement cost and benefit of each capability node, and obtain the capability improvement sequence; According to the ability improvement sequence, scenario model construction is performed, by simulating the ability improvement process of teachers in an actual teaching environment, at least two scenario models are constructed using a scenario simulation algorithm, and the rationality of the path is judged to obtain path correction suggestions; The path is optimized according to the path correction suggestion, and the literacy development path is obtained by comprehensively considering the priority order of each ability node, the feedback information in the scenario mode and the reasonable arrangement of learning time.

4. A system for formulating a teacher's intelligent education literacy training program based on AI, characterized in that: include: An acquisition module is used to acquire the basic data and intelligent education literacy goals of teachers, wherein the basic data includes personal information, teaching practice records and career development goals; A processing module, configured to perform text processing according to the basic data, identify literacy content and teaching elements related to the teacher's intelligent education literacy, and construct an intelligent education literacy knowledge graph based on the correlation and influence path between the literacy content and the teaching elements, wherein the nodes in the intelligent education literacy knowledge graph represent various literacy abilities of the teacher, and the edges represent the dependencies between the abilities; A reasoning module is used to use a graph neural network to reason about the knowledge graph based on the intelligent education literacy knowledge graph, evaluate the current level of the teacher at each literacy ability node, and obtain a literacy level evaluation result; An analysis module is used to perform gap analysis based on the literacy level assessment result and the intelligent education literacy goal, and obtain a gap analysis result by analyzing the gap between the teacher's current literacy and the target literacy and determining the ability that needs to be improved; An optimization module, used to perform optimization processing according to the gap analysis results and the intelligent education literacy knowledge map, and use a path optimization algorithm to plan the ability improvement sequence and learning resources to obtain a literacy cultivation path, wherein the literacy cultivation path includes the sequence of each ability improvement and the corresponding learning resource recommendation; A planning module is used to perform resource allocation processing according to the literacy cultivation path, plan the learning plan by using resource priority analysis and time series planning algorithm, and obtain a teacher intelligent education literacy cultivation plan including learning path, resource allocation, learning progress and evaluation indicators; Wherein, the processing module includes: A first processing unit is used to perform text processing using a natural language processing algorithm based on the basic data, convert the text data into a vector representation through a word embedding model, and use a named entity recognition algorithm to identify key terms and concepts related to the teacher's intelligent education literacy, so as to obtain a vector representation of literacy content and teaching elements; A first calculation unit is used to perform semantic relevance modeling based on the vectorized representation using a semantic similarity calculation algorithm, calculate the similarity between different literacy contents and teaching elements, and screen out correlation combinations based on a preset threshold to obtain a relevance matrix between literacy contents and teaching elements; A first modeling unit is used to model the dependency path between the literacy content and the teaching elements according to the correlation matrix by using the shortest path algorithm in graph theory, and obtain an influence path diagram between the literacy content and the teaching elements by analyzing the transitivity and hierarchical relationship between different contents, wherein the nodes represent each specific literacy content and the edges represent the dependency paths between the literacy contents; A first optimization unit is used to optimize the graph structure by using a graph neural network according to the impact path graph, embed the nodes through a graph convolutional network, optimize the node relationship between each literacy content and teaching elements, and obtain an intelligent education literacy knowledge graph; Wherein, the reasoning module includes: The second processing unit is used to perform initialization processing of the graph neural network according to the intelligent education literacy knowledge graph, by allocating an initial feature vector to the relationship between each literacy content node and its adjacent nodes, and using the initial parameter setting of the graph convolutional network to embed the nodes and their edges to obtain an initial feature matrix of the graph neural network; The third processing unit is used to perform graph convolution processing according to the initial feature matrix, aggregate and transmit information of each node and its adjacent nodes through a multi-layer graph convolution network, update the feature vector of each node, and obtain a node feature matrix reflecting the literacy content node and its related nodes; The first classification unit is used to classify the literacy nodes using the classification layer of the graph neural network according to the node feature matrix, and obtain the classification label of each node by performing classification operations on the final feature vector of each node and predicting the ability level of the teacher on each literacy content node; The first integration unit is used to perform integration processing according to the classification labels, integrate the evaluation results of each node through the weighted average method, and combine the importance of the node and the influence of its adjacent nodes to obtain the overall evaluation result of the teacher's literacy level.

5. The AI-based teacher intelligent education literacy cultivation program formulation system according to claim 4 is characterized in that: The analysis module comprises: A first matching unit is used to perform matching processing according to the literacy level assessment result and the intelligent education literacy target, by comparing the teacher's current ability level at each literacy content node with the target level, and using the Euclidean distance method to calculate the similarity between the current literacy level and the target literacy, so as to obtain a preliminary matching result; A fourth processing unit is used to perform gap quantification processing according to the preliminary matching result, by calculating the gap value of each literacy content node and performing quantitative analysis to obtain a gap quantification result of each node; A first determination unit is used to perform capability improvement demand determination processing according to the gap quantification result, determine whether each literacy content node needs to be improved by setting a threshold, and use a dynamic programming algorithm to determine the optimal capability improvement strategy to obtain a capability node set; The second calculation unit is used to perform priority analysis based on the capability node set, calculate the priority order of each node based on the importance, difficulty and influence of the node on the overall quality, and obtain the final gap analysis result.

6. The AI-based teacher intelligent education literacy cultivation program formulation system according to claim 4 is characterized in that: The optimization module includes: A fifth processing unit is used to process the dependency relationship of capability nodes according to the gap analysis result and the intelligent education literacy knowledge graph, analyze the interdependence between the capability nodes in the knowledge graph, and use a topological sorting algorithm to determine the priority of the improvement path to obtain a preliminary capability improvement order; A second optimization unit is used to perform path optimization processing according to the preliminary capability improvement sequence, dynamically plan the optimal capability improvement path by analyzing the improvement cost and benefit of each capability node, and obtain the capability improvement sequence; The first construction unit is used to perform scenario model construction processing according to the ability improvement sequence, construct at least two scenario models using a scenario simulation algorithm by simulating the ability improvement process of teachers in an actual teaching environment, and judge the rationality of the path to obtain path correction suggestions; The third optimization unit is used to perform path optimization processing according to the path correction suggestion, and obtain the literacy cultivation path by comprehensively considering the priority order of each ability node, the feedback information in the scenario mode and the reasonable arrangement of learning time.

Images