A modular content generation method and system for embedded instruction

Abstract

The application discloses a modular content generation method and system for embedded teaching, and relates to the technical field of teaching semantic generation.A modular content generation system for embedded teaching comprises an embedded teaching module division module and an embedded teaching content generation module.The application solves the problems of loose organization and poor reusability of traditional embedded teaching content by semantic analysis and structured processing of original embedded teaching content;by introducing a skill dependency analysis mechanism, an explicit pre-skill set and output skill set are allocated to each initial teaching module, a teaching path network oriented to ability growth is constructed, and the clarity of teaching logic and the controllability of learning path are improved;learning fatigue is evaluated by combining multi-dimensional factors such as skill span, path complexity and cognitive load, and cognitive overload or content redundancy in the learning process of students is effectively avoided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of teaching semantic generation, and particularly relates to a modular content generation method and system for embedded teaching. BACKGROUND

[0002] The existing embedded teaching mostly adopts linear and fixed course structure, which is difficult to dynamically adjust the teaching content according to the individual differences of students, resulting in problems such as mismatched learning rhythm and unbalanced cognitive load. At the same time, the teaching resources are mostly static courseware, experimental instruction books or scattered code instances, which lack structured organization and skill correlation analysis, and cannot support modular reconstruction and personalized generation. In addition, the traditional teaching mode lacks real-time feedback mechanism for the learning state of students, and the teaching content is disconnected from the ability of learners, which affects the teaching effect.

[0003] Therefore, there is an urgent need for a modular embedded teaching method and system which can dynamically optimize the teaching structure according to the cognitive law based on semantic analysis and skill modeling, and generate adaptive learning content combined with student portraits. SUMMARY

[0004] The present application aims to provide a modular content generation method and system for embedded teaching, which can dynamically optimize the teaching structure according to the cognitive law based on semantic analysis and skill modeling, and generate adaptive learning content combined with student portraits.

[0005] A modular content generation method for embedded teaching, comprising the following steps:

[0006] Obtaining original teaching content related to embedded teaching; performing semantic analysis and structured processing on the original teaching content, dividing the original teaching content into a plurality of initial teaching modules; assigning a corresponding pre-skills set and output skills set to each initial teaching module; sorting all initial teaching modules according to the teaching sequence relationship;

[0007] For any initial teaching module, based on the pre-skills set and the output skills set, performing learning fatigue evaluation on the initial teaching module to obtain a module fatigue evaluation result; comparing the module fatigue evaluation result with a preset fatigue threshold; if the module fatigue evaluation result exceeds the preset fatigue threshold, the corresponding initial teaching module is further split to obtain at least two split embedded teaching modules; if the module fatigue evaluation result is much lower than the preset fatigue threshold, a plurality of associated initial teaching modules are selected for combination to obtain a combined embedded teaching module;

[0008] Based on the step of comparing the module fatigue evaluation result with the preset fatigue threshold, all initial teaching modules are iteratively compared and judged, and all split embedded teaching modules and combined embedded teaching modules obtained are renamed to obtain a plurality of embedded teaching modules.

[0009] According to the student learning portrait and the embedded teaching module to be learned, corresponding embedded module learning content is generated; learning is carried out based on the corresponding embedded module learning content, and learning feedback data is collected to update the student learning portrait.

[0010] As a preferred technical solution of the present application, the specific steps of assigning a corresponding pre-skills set and output skills set to each initial teaching module include:

[0011] The first n initial teaching modules and the last n initial teaching modules of the initial teaching module are identified;

[0012] The semantic analysis is performed on the teaching content contained in the initial teaching module to obtain the knowledge set corresponding to the initial teaching module;

[0013] The semantic analysis is performed on the first n initial teaching modules to obtain the output basic ability set, and the semantic analysis is performed on the last n initial teaching modules to obtain the pre-basic ability set; based on the dependency relationship between knowledge points, the knowledge set, the pre-basic ability set and the output basic ability set are comprehensively analyzed to obtain the pre-skills set and the output skills set of the initial teaching module.

[0014] As a preferred technical solution of the present application, the specific steps of learning fatigue evaluation of the initial teaching module based on the pre-skills set and the output skills set include:

[0015] A skill learning span model is constructed based on prior knowledge; a quantitative analysis is performed on the skill difference of the pre-skills set and the output skills set to obtain a skill cognitive transition intensity value; the average cognitive load bearing capacity of the student is analyzed based on big data, and a preset fatigue threshold is determined based on the average cognitive load bearing capacity;

[0016] The number and length of the skill conversion paths contained in the initial teaching module are calculated, and the number and length of the skill conversion paths are quantified as a learning fatigue index;

[0017] The skill cognitive transition intensity value and the learning fatigue index are feature fused to obtain a module fatigue evaluation result.

[0018] As a preferred technical solution of the present application, if the module fatigue evaluation result exceeds the preset fatigue threshold, the specific steps include:

[0019] Based on the skill span distribution between the pre-skills set and the output skills set, a skill transition mutation point in the initial teaching module is identified; based on the skill transition mutation point, a cognitive overload sub-region in the initial teaching module is determined;

[0020] The learning content link graph of the cognitive overload sub-region is generated by using a graph neural network; and clustering analysis is performed based on the learning content link graph to extract a teaching sub-module candidate set with a long operation chain;

[0021] The number of shortest paths, longest paths and cross paths of the learning content link graph is identified, and the teaching sub-module candidate set is subjected to load balancing splitting processing based on the identified number, to obtain at least two split embedded teaching modules;

[0022] The split embedded teaching modules after splitting are re-established for teaching sequence relationship, and the pre-skills set and output skills set corresponding to the split embedded teaching modules are updated.

[0023] As a preferred technical solution of the present application, if the module fatigue evaluation result is far below the preset fatigue threshold, the specific steps include:

[0024] Other initial teaching modules with similar output skill sets of the current initial teaching module and module fatigue evaluation results far below the preset fatigue threshold are extracted as associated initial teaching modules;

[0025] The learning content link graph of the plurality of associated initial teaching modules is identified, and a module merging candidate pair corresponding to the current initial teaching module is constructed;

[0026] A learning coherence scoring model is introduced; the module merging candidate pair is screened by using the learning coherence scoring model, and the module merging candidate pair meeting the preset learning coherence requirement is selected;

[0027] The module merging candidate pair meeting the preset learning coherence requirement is combined to obtain a combined embedded teaching module; the teaching sequence relationship of the combined embedded teaching module is re-established, and the pre-skills set and output skills set corresponding to the combined embedded teaching module are updated.

[0028] As a preferred technical solution of the present application, for the student learning portrait and the embedded teaching module to be learned, the specific steps of generating the corresponding embedded module learning content include:

[0029] The student learning portrait is obtained; the student learning portrait contains the current learning cognitive load, learning style preference and prior knowledge level; the semantic generation granularity is matched based on the prior knowledge level; the semantic generation preference is generated based on the learning style preference; and the semantic generation density is matched based on the current learning cognitive load;

[0030] The corresponding embedded module learning content of the embedded teaching module is generated based on the semantic generation granularity, semantic generation preference and semantic generation density.

[0031] A modular content generation system for embedded teaching, comprising:

[0032] The embedded teaching module division module comprises an initial module generation unit and an embedded teaching module generation unit; the initial module generation unit is used for obtaining original teaching content related to embedded teaching; the original teaching content is subjected to semantic analysis and structured processing, and the original teaching content is divided into a plurality of initial teaching modules; a corresponding prerequisite skill set and an output skill set are assigned to each initial teaching module; all the initial teaching modules are sorted according to a teaching sequence relationship; the embedded teaching module generation unit is used for, for any initial teaching module, performing learning fatigue evaluation on the initial teaching module based on the prerequisite skill set and the output skill set to obtain a module fatigue evaluation result; the module fatigue evaluation result is compared with a preset fatigue threshold to determine whether the module fatigue evaluation result exceeds the preset fatigue threshold; if the module fatigue evaluation result exceeds the preset fatigue threshold, the corresponding initial teaching module is further split to obtain at least two split embedded teaching modules; if the module fatigue evaluation result is much lower than the preset fatigue threshold, a plurality of associated initial teaching modules are selected for combination to obtain a combined embedded teaching module;

[0033] Based on the step of comparing the module fatigue evaluation result with the preset fatigue threshold, all the initial teaching modules are iteratively compared and determined, and all the split embedded teaching modules and the combined embedded teaching modules obtained are renamed to obtain a plurality of embedded teaching modules;

[0034] The embedded teaching content generation module comprises a module content generation unit; the module content generation unit is used for generating corresponding embedded module learning content for a student learning portrait and an embedded teaching module to be learned; learning is performed based on the corresponding embedded module learning content, and learning feedback data is collected for updating the student learning portrait.

[0035] The present application has the following advantages:

[0036] 1. The present application converts unstructured and multi-modal teaching resources into analyzable and operable teaching modules through semantic analysis and structured processing of original embedded teaching content, solves the problem of loose organization and poor reusability of traditional embedded teaching content, and assigns a clear prerequisite skill set and output skill set to each initial teaching module by introducing a skill dependency analysis mechanism, thereby constructing a teaching path network oriented to ability growth and improving the clarity of teaching logic and the controllability of learning path; further, learning fatigue evaluation is performed by taking into account multi-dimensional factors such as skill span, path complexity and cognitive load, which can dynamically determine whether the module is too heavy or too light, support intelligent splitting and reconstruction of the module, and effectively avoid cognitive overload or content redundancy in the student learning process.

[0037] 2、The present application can realize the personalized dynamic generation of the granularity, density and form of embedded teaching content by introducing a student learning portrait including prior knowledge level, cognitive load and learning style preference, so that the learning content is more suitable for individual differences, and the participation and learning efficiency of students are improved; at the same time, the feedback data collected in the learning process will continuously update the learning portrait, realize the two-way adaptation and closed-loop optimization of teaching content and learning state, and promote the formation of a precise teaching mechanism with students as the center, task-driven and skill-oriented. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 A structural schematic diagram of a modular content generation system for embedded teaching used in an embodiment of the present application. DETAILED DESCRIPTION

[0039] In order to enable personnel in the technical field to better understand the technical solutions in the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application.

[0040] Embodiment 1, a modular content generation method for embedded teaching, comprising the following steps:

[0041] Obtain the original teaching content related to embedded teaching; the original teaching content includes but is not limited to text, code, schematic diagram, embedded development board simulation model, development board teaching video, audio and 3D hardware model; the original teaching content of embedded teaching is usually collected, screened and integrated step by step through multiple channels; for example, the systematic embedded course materials can be obtained through the official teaching resource library of a university, which contains detailed teaching outline, course lecture notes, experiment guide book, schematic diagram file and supporting source code; secondly, large technical communities and professional forums such as electronic engineering, embedded development or single-chip application websites can be accessed, in which original teaching content, development board training documents, real project code and circuit diagram shared by engineers or teachers can be found, which are usually published in the form of posts, compressed packages or file attachments, with demonstration videos or development board simulation model files, including Proteus simulation diagram, Multisim engineering file, etc., which can be used for local learning or simulation experiment after downloading; in addition, the original code, circuit diagram and three-dimensional hardware model of the development board can also be obtained through the open source hardware platform.

[0042] Perform semantic analysis and structural processing on the original teaching content, and divide the original teaching content into a plurality of initial teaching modules;

[0043] Specific steps: First, different types of content need to be converted into a unified semantic data format. For text content such as teaching notes, experiment guides, source code comments, etc., we can use natural language processing techniques such as word segmentation, syntax analysis, and named entity recognition to extract key terms, concept definitions, and knowledge system structures. For video and audio materials, we need to perform speech recognition and image recognition first, transcribe the speech into text, and identify the operation steps, device labels, or code demonstration screens in the picture, and extract the teaching information in combination with the speech content. Circuit schematics and three-dimensional hardware models are extracted through image recognition or model analysis tools to extract components, connection relationships, and hardware composition logic, and then converted into data that can describe hardware structure and function. After analyzing the parsed semantic content, identify knowledge points and skill points based on knowledge graph construction methods, and label information such as embedded system theory concepts (such as interrupts, timers, serial communication), programming techniques (such as register operations, peripheral device drivers, and bare-metal development), and hardware operations (such as circuit soldering, development board debugging, and simulation platform usage) in the semantic structure. Combine the actual project operation process to further identify practical skills, such as implementing LED flicker control based on a timer, and performing data loopback testing using a serial debugging tool. This process includes both rule-based reasoning methods and deep learning models to identify implicit skill expressions in unstructured content.

[0044] After forming a complete knowledge point and skill point map, the content needs to be preliminarily structured based on teaching logic. This division is based on the hierarchical relationship of knowledge and skills, the order of learning, and the dependency between content to analyze and determine the corresponding unit boundaries. Finally, each initial teaching module should include a main knowledge point, several related skill points, and a complete practical goal, forming a basic unit structure with clear content, clear goals, and easy teaching organization.

[0045] Assign a corresponding set of pre-skills and output skills to each initial teaching module. The pre-skills set refers to the basic skills or prerequisite abilities that learners should have before learning a certain initial teaching module. These skills are the prerequisites for successfully learning, understanding, and completing practical tasks in this unit. Pre-skills may include basic programming syntax, basic circuit knowledge, development environment configuration, and common tool usage methods. Without these skills, learners may not be able to effectively follow the teaching content, resulting in knowledge gaps or practical failures. The output skill set refers to the new skills or improved abilities that learners should master after completing a certain initial teaching module. These skills are the embodiment of the module's teaching goals and typically include understanding of specific knowledge, hands-on operation ability, problem-solving ability, or practical control of certain hardware modules. Output skills can be basic (such as lighting an LED) or complex (such as implementing a master-slave communication mechanism based on a serial protocol), depending on the depth of teaching and the complexity of the task in the unit.

[0046] All initial teaching modules are sorted according to the teaching sequence relationship; in order to sort all initial teaching modules according to the teaching sequence relationship, a complete skill dependency graph needs to be constructed; the graph takes each initial teaching module as a node, and the dependency relationship between the prerequisite skill set and the output skill set of the modules as a directed edge, thereby forming a directed graph structure; by comparing the skill sets of each module, if the prerequisite skill of one module is exactly a subset or completely contained in the output skill of another module, it can be determined that the former should precede the latter in the teaching order; the completed skill dependency graph not only describes the flow direction of the teaching logic, but also reveals the order and skill evolution path between the modules.

[0047] The specific steps for assigning the corresponding prerequisite skill set and output skill set to each initial teaching module include:

[0048] The first n initial teaching modules and the last n initial teaching modules of the initial teaching modules are identified; the specific value of n is set by professional technicians; these modules represent the possible leading teaching content and subsequent connecting content of the module; by constructing a local teaching window in this way, the system can understand the logical connection and dependency path between the module and the surrounding knowledge.

[0049] The semantic analysis of the teaching content contained in the initial teaching module is performed to obtain the knowledge set corresponding to the initial teaching module; natural language processing technology is used to analyze the text content, code comments, experimental step instructions and other information in the module to extract the core knowledge units with semantic meaning and construct the knowledge set of the module; the knowledge set can include theoretical concepts, operation processes, interface mechanisms, hardware characteristics, etc., which are the basic corpus resources for subsequent skill reasoning.

[0050] The semantic analysis of the first n initial teaching modules is performed to obtain the output basic ability set; the semantic analysis of the last n initial teaching modules is performed to obtain the prerequisite basic ability set; based on the dependency relationship between knowledge points, the knowledge set, the prerequisite basic ability set and the output basic ability set are comprehensively analyzed to obtain the prerequisite skill set and the output skill set of the initial teaching module; the semantic analysis of the first n initial teaching modules of the current module is performed one by one to extract the ability unit set that should be possessed after these modules are completed, which is called the output basic ability set; this set represents the prerequisite knowledge and operation experience that the learner of the current module may already have; at the same time, the semantic analysis of the last n modules is performed to obtain the prerequisite ability set that they depend on, i.e. the basic skills required for these modules to successfully learn, which can reflect the skill transfer role of the current module in the future teaching path and provide a basis for the setting of skill output.

[0051] Fusion analysis is performed on the knowledge set of the current initial teaching module, the output basic ability set of the previous n modules, and the pre-requisite basic ability set of the next n modules. During the analysis process, the logical dependency relationship between knowledge points and the skill development path diagram are used to identify the minimum ability units that must be mastered by the current module as the pre-requisite skill set, and to extract the new skills that should be output after learning the current module as the output skill set. This comprehensive analysis ensures that the skill chain forms a continuous and effective transmission mechanism between teaching modules, thereby constructing a skill structure system that conforms to the cognitive law and has teaching logic.

[0052] For any initial teaching module, learning fatigue assessment is performed on the initial teaching module based on the pre-requisite skill set and the output skill set to obtain a module fatigue assessment result. The module fatigue assessment result is compared with a preset fatigue threshold. If the module fatigue assessment result exceeds the preset fatigue threshold, the corresponding initial teaching module is further split to obtain at least two split embedded teaching modules. If the module fatigue assessment result is much lower than the preset fatigue threshold, multiple associated initial teaching modules are selected for combination to obtain a combined embedded teaching module.

[0053] Based on the step of comparing the module fatigue assessment result with the preset fatigue threshold, all initial teaching modules are iteratively compared and judged. The obtained split embedded teaching modules and combined embedded teaching modules are renamed to obtain a plurality of embedded teaching modules.

[0054] The specific steps of performing learning fatigue assessment on the initial teaching module based on the pre-requisite skill set and the output skill set include:

[0055] A skill learning span model is constructed based on prior knowledge. The skill difference between the pre-requisite skill set and the output skill set is quantitatively analyzed to obtain a skill cognitive transition intensity value. The average cognitive load bearing capacity of students is analyzed based on big data, and the preset fatigue threshold is determined based on the average cognitive load bearing capacity.

[0056] The skill learning span model is used to measure the learning span of a teaching module at the knowledge level, i.e., the knowledge depth and skill type complexity spanned from the existing skills of students to the target skills. This model is usually driven by education experts and domain knowledge graphs, and includes multiple dimensions such as skill difficulty level, knowledge dependency level, operation abstraction level, and cross-domain integration level, to characterize the transition cost between different skills. After construction, this model will serve as a basic tool for evaluating the cognitive jump degree of teaching modules.

[0057] Based on the skill difference relationship between the prerequisite skill set and the output skill set of the initial teaching module, the skill learning span model is used to quantitatively analyze the difference part, and the skill cognitive transition intensity value is obtained. The skill cognitive transition intensity value represents the cognitive transformation intensity required to transition from the prerequisite skill to the output skill, that is, how many cognitive structures and operation models the learner needs to add. The larger the value, the greater the skill promotion span in the module, and the learner needs to understand, reconstruct concepts or skill fusion more deeply, so it is easier to cause learning fatigue;

[0058] Based on the big data analysis of the average cognitive load bearing capacity of the student group, the corresponding preset fatigue threshold is determined. The cognitive load bearing capacity refers to the upper limit of new information or skills that students can effectively absorb, process and master in a unit of learning time or learning stage. This capacity is derived by analyzing a large number of student learning behavior data under different skill difficulty and module structure, including completion time, failure rate, answer accuracy, operation times, etc. Through statistical modeling and cluster analysis, it can be deduced that the fatigue limit that students can generally bear in a certain teaching environment, so as to set the preset fatigue threshold for subsequent fatigue degree judgment.

[0059] Calculate the number and length of skill transformation paths contained in the initial teaching module, and quantify the number and length of skill transformation paths as learning fatigue indexes;

[0060] The skill transformation paths contained in the initial teaching module are calculated, specifically including the number and length of the paths. The number of paths refers to the total number of all feasible learning paths from any skill in the prerequisite skill set to the output skill set. The path length is the number of intermediate skill transformation steps that need to be completed on each path. These two indicators reflect the complexity of the skill structure and the depth of the cognitive process. The more paths and the longer the length, the more complex the internal skill transition chain of the module, the longer the operation load and cognitive concentration time. The path number and path length are combined and weighted to obtain the learning fatigue index, which is used to measure the potential learning fatigue caused by long skill chains or dense paths.

[0061] Feature fusion is performed on the skill cognitive transition intensity value and the learning fatigue index to obtain the module fatigue evaluation result;

[0062] The skill cognitive transition intensity value is fused with the learning fatigue index to form a module fatigue evaluation result by comprehensively considering the information of both aspects; the fusion method can adopt weighted average, fuzzy reasoning or machine learning model, and the specific strategy is flexibly adjusted according to different course properties. The module fatigue evaluation result is a total value index, which is used to judge whether the current module is too dense or the target span is too large; if the value exceeds the preset fatigue threshold, it is prompted that the module is overloaded, and needs to be split and optimized; if the result is low, it means that the module skills are too scattered or repeated, and can be considered to be combined to improve the learning efficiency and content coherence. This evaluation mechanism makes the teaching module design more scientific and individualized, and effectively improves the learner experience and the quality of course implementation.

[0063] If the module fatigue evaluation result exceeds the preset fatigue threshold, the specific steps include:

[0064] Based on the skill span distribution between the pre-skills set and the output skill set, the skill transition mutation points in the initial teaching module are identified; and based on the skill transition mutation points, the cognitive overload sub-regions in the initial teaching module are determined.

[0065] The skill span distribution refers to the complexity level change, correlation degree and continuity of skill transformation demand among various skill points in the initial teaching module; when the skill span of a certain position presents mutation in a local range (i.e. the difficulty or dependency relationship is significantly improved), it is a skill transition mutation point. These mutation points often indicate regions with sharp increase in cognitive load and sudden learning difficulty, which need to be paid attention to;

[0066] After identifying the skill transition mutation points, the cognitive overload sub-regions in the initial teaching module need to be further determined based on these mutation points; the cognitive overload sub-region refers to those teaching content fragments containing high-difficulty skill combinations, knowledge point intensive intersection and complex operation logic; in these regions, students need to complete more skill conversion, information processing and abstract reasoning, which is easy to cause learning fatigue; therefore, these sub-regions need to be separated from the whole module as a separate research and splitting object for further optimization.

[0067] A learning content link graph of the cognitive overload sub-region is generated by using a graph neural network; based on the learning content link graph, clustering analysis is performed to extract a teaching sub-module candidate set with a long operation chain;

[0068] For the above cognitive overload sub-area, the internal teaching content is modeled using graph neural network technology to generate a learning content link graph. The link graph takes skill points and knowledge points in the teaching content as nodes, and skill dependency, concept reference and operation sequence as edges to express the logical relationship and learning path between the contents. The graph neural network not only considers the attribute features of the nodes themselves, but also learns the semantic association and structural embedding between the nodes through a multi-layer message passing mechanism to form a higher-dimensional learning path expression, providing basic data for subsequent content restructuring. After obtaining the learning content link graph, clustering analysis is performed on the graph structure to identify subgraph regions that are similar in structure and highly related in content. These regions are used as a candidate set of teaching sub-modules. A candidate sub-module is usually composed of one or more operation chains, each chain representing an operation process that evolves from a basic skill to a specific target skill. By calculating the length and node density of these chains, the complexity of each sub-module can be evaluated to determine whether it contains an excessively long operation path or too many concurrent skill points.

[0069] The number of shortest paths, longest paths and cross paths of the learning content link graph is identified, and the candidate set of teaching sub-modules is balanced and split based on the identified number to obtain at least two split embedded teaching modules.

[0070] The number of shortest paths, longest paths and cross paths in the learning content link graph is identified. The shortest path reflects the minimum skill steps required to achieve the basic learning goal in the module, the longest path reveals the deepest skill dependency chain that may exist in the module, and the number of cross paths represents the degree of skill reuse and multi-path coupling. Based on these three indicators, the system performs balanced and split processing on the candidate sub-modules. This process splits high-complexity areas into at least two split embedded teaching modules with smaller cognitive span, shallower path depth and fewer skill intersections, to balance the skill distribution within the modules and reduce the cognitive burden on learners.

[0071] The split embedded teaching modules after splitting are re-established for teaching sequence relationship, and the pre-skills set and output skills set corresponding to the split embedded teaching modules are updated. The at least two split embedded teaching modules after splitting are re-established for teaching sequence. According to their original position sequence, skill dependency direction and knowledge development logic, a teaching path graph is reconstructed to ensure smooth connection between modules and natural skill progression. The pre-skills set and output skills set are re-assigned to each newly generated split embedded teaching module to adapt to individualized learning path planning, forming controllable teaching units with clear structure and appropriate granularity.

[0072] If the module fatigue evaluation result is much lower than the preset fatigue threshold, the specific steps include:

[0073] Far below the preset fatigue threshold is a criterion for determining whether the initial teaching module has too light a learning load or too small a teaching granularity. The "far below" is not an absolute numerical value, but a quantitative deviation interval relative to the threshold. The judgment is usually set in combination with statistical methods and teaching cognitive models. The specific judgment method is as follows:

[0074] The preset fatigue threshold is a critical value determined based on the average cognitive load bearing capacity of the student group, indicating the upper limit of the reasonable cognitive pressure that the students can bear under normal learning conditions to complete a module. Generally, the threshold is an interval, not a single numerical value. Far below is usually quantitatively judged by the difference ratio method. The difference ratio method means that if the learning fatigue index of a module is lower than a certain percentage (such as 30% or 40%) of the lower limit of the fatigue threshold, it is considered to be far below. For example, if the preset fatigue threshold is 80, and the evaluation score of a module is 45, and the gap is more than 24 points, which is 30% of the difference between 80 and 45, then it can be determined that it is far below.

[0075] Extract other initial teaching modules that have similarity with the output skill set of the current initial teaching module and whose module fatigue evaluation results are all far below the preset fatigue threshold, as the associated initial teaching modules;

[0076] Extract other initial teaching modules that have similarity with the output skill set of the current initial teaching module, and these modules also have low fatigue evaluation results, indicating that these modules are scattered in content, have too small granularity, or have too light cognitive burden, which may cause learning rhythm interruption or content redundancy. At this time, the modules with similar skill output directions can be identified by skill set similarity calculation, such as semantic similarity matching based on knowledge graph or vector embedding model, to build the associated initial teaching module set of the current module, and prepare for subsequent content fusion.

[0077] Identify the learning content link graph of multiple associated initial teaching modules, and construct module merging candidate pairs corresponding to the current initial teaching module. After identifying multiple associated initial teaching modules, the learning content link graphs of these modules need to be extracted. The learning content link graph is a directed graph structure with module internal knowledge points, skill points, operation steps, etc. as nodes, and uses the logical or dependency relationship between them to build a structured flow for representing the internal content of the module. By comparing the link graph of the current module with the link graphs of the associated modules, a number of module merging candidate pairs can be combined, each pair representing two teaching modules that may have merging potential in content logic, skill direction, and learning path.

[0078] The learning continuity scoring model is introduced; the learning continuity scoring model is used for screening the module merging candidate pairs, and the module merging candidate pairs meeting the preset learning continuity requirements are selected; the learning continuity scoring model is introduced, which is used for evaluating the consistency between the module merging candidate pairs in terms of teaching rhythm and cognitive process; the learning continuity scoring model is an evaluation tool constructed based on multiple indexes such as structural similarity, skill transfer fluency and teaching target consistency, and its function is to avoid abrupt learning rhythm or logical jump caused by simple content splicing; the learning continuity scoring model considers the concept continuity, operation connection and semantic transition smoothness between modules, and sets a continuity threshold, so that only under the condition of logical continuity and reasonable skill docking, the candidate pair can be further merged.

[0079] The learning continuity scoring model is constructed to measure whether two or more teaching modules have natural connection in terms of knowledge logic, skill transfer and task rhythm, so as to avoid cognitive jump or learning fragmentation after module merging; the construction method of the model is as follows: firstly, based on the structured characteristics of the teaching module, the knowledge point sequence, skill path graph and operation flow graph of each module are vectorized and coded to generate the semantic representation thereof; on this basis, a deep matching network or a double-tower semantic similarity model is used, two modules are taken as an input pair, and the learning continuity score between them is learned through training; the training process is mainly supervised learning, a batch of module combination data labeled by teaching experts are selected as the training set, the positive examples are module combinations considered to have natural teaching connection and reasonable rhythm, and the negative examples are combinations with abrupt logic and disjointed skills; the model gradually learns the effective parameters for measuring teaching continuity by optimizing the gap between the score and the artificial score; the training set mainly includes historical course design schemes, classic module pairs in the teaching case library, module combinations with good and poor performance in actual student learning path data, and the continuity evaluation contrast set constructed by artificial scoring through the expert evaluation system; finally, the scoring model formed can input any two teaching modules to be merged, and output a continuity score for judging whether they are suitable for fusion into a unified embedded teaching module.

[0080] The module merging candidate pairs meeting the preset learning continuity requirements are combined to obtain a combined embedded teaching module; the teaching sequence relationship of the combined embedded teaching module is re-established, and the pre-skills set and output skills set corresponding to the combined embedded teaching module are updated; the preset learning continuity requirements are set by professional technicians artificially;

[0081] The learning continuity scoring model is used to screen all candidate pairs, remove module combinations with substandard continuity scores, and retain merging candidate pairs that meet the preset learning continuity requirements. This process ensures that the new modules generated by subsequent merging do not affect learning effectiveness due to mixed content, skill conflicts, or cognitive rhythm disorders, and also ensures the natural connection of the teaching path and the consistency of the content. Module merging operations are performed based on the screened module merging candidate pairs to obtain combined embedded teaching modules. Each combined embedded teaching module should have stronger content integrity, task continuity, and skill effectiveness in terms of teaching function, which improves teaching efficiency and reduces fragmentation risk. After module merging, the teaching sequence relationship needs to be reconstructed to ensure that the combined embedded teaching module is in the appropriate order position in the entire teaching path, and its prerequisite skill set should cover the prerequisite skill set of all original sub-modules, and its output skill set should cover the skills output of all merged modules, thereby ensuring that the combined embedded teaching module has a reasonable input and output interface in terms of skill structure, supporting the smooth connection of subsequent modules and progressive learning.

[0082] All split embedded teaching modules and combined embedded teaching modules obtained are renamed. The meaning of this step is that within the system, the split embedded teaching modules obtained by splitting the initial teaching modules, or the combined embedded teaching modules obtained by recombining the initial teaching modules, and the initial teaching modules that have not been split or combined are integrated and unified as embedded teaching modules for subsequent teaching content generation operations.

[0083] Specifically, the teaching sequence relationship corresponding to the split embedded teaching module, the teaching sequence relationship corresponding to the combined embedded teaching module, and the teaching sequence relationship corresponding to the initial teaching module are taken as the teaching sequence relationship of the corresponding embedded teaching module, and are reordered.

[0084] For the student learning profile and the embedded teaching module to be learned, corresponding embedded module learning content is generated; learning is performed based on the corresponding embedded module learning content and learning feedback data is collected for updating the student learning profile;

[0085] The specific steps for generating corresponding embedded module learning content for the student learning profile and the embedded teaching module to be learned include:

[0086] Obtain the student learning profile; the student learning profile includes the current learning cognitive load, learning style preference, and prior knowledge level; match the semantic generation granularity based on the prior knowledge level; generate semantic generation preference based on the learning style preference; and match the semantic generation density based on the current learning cognitive load;

[0087] A student learning profile is a dynamic feature model constructed based on a student individual, mainly including three core dimensions of current learning cognitive load, learning style preference and prior knowledge level. Among them, the learning cognitive load reflects the psychological carrying capacity of the student in absorbing knowledge, mastering skills and operating execution in a certain learning stage; the learning style preference embodies the preference of the student in information receiving, processing and expression mode, such as preferring to show in the form of diagram, explain in words or interact in practice; the prior knowledge level describes the knowledge reserve and skill mastery depth of the student before entering a certain teaching module; the learning profile is usually obtained by relying on the behavior data, answer records, task completion, learning feedback and stage evaluation results in the learning process.

[0088] Based on the semantic generation granularity, semantic generation preference and semantic generation density, the corresponding embedded module learning content of the embedded teaching module is generated;

[0089] After mastering the prior knowledge level of the student, the semantic generation granularity needs to be matched based on the level. The semantic generation granularity refers to the refinement degree or information decomposition depth when generating learning content, which adjusts the complexity and information density of the teaching materials. If the prior knowledge of the student is weak, a fine-grained strategy is needed to decompose a complex knowledge point into multiple basic concepts and add a large number of explanatory explanations. If the student has a high knowledge base, a coarse-grained expression method can be used to simplify the description, focus on key knowledge and practical operation goals, and improve learning efficiency and challenge, so as to achieve personalized matching.

[0090] According to the learning style preference of the student, the semantic generation preference is determined. The semantic generation preference refers to the generation tendency in the form of content presentation, which ensures that the generated learning content can conform to the way most easily accepted and processed by the student. For example, students who prefer visual learning can generate flowcharts, structural diagrams, circuit diagrams and other graphical content; students who prefer language learning generate detailed text explanations, concept definitions and process explanations; and students who prefer hands-on operation can highlight operation steps, real cases and interactive simulation experiments. This preference matching can significantly improve the learning motivation and acceptance of the student.

[0091] The semantic generation density also needs to be matched according to the current learning cognitive load level. The semantic generation density refers to the amount of information or the level of complexity in unit content, which determines the compactness of the learning content in logical progression. If the student is currently in a high cognitive load state, the semantic density should be appropriately reduced by inserting blank transition points, supplementing examples, simplifying language and other ways to reduce cognitive pressure. Conversely, in the state of sufficient cognitive resources, the density can be increased to integrate knowledge points, increase extension content or increase problem difficulty to maintain challenge and efficiency.

[0092] Integrating the three personalized adjustment parameters of semantic generation granularity, semantic generation preference and semantic generation density, the generation engine is called to dynamically generate content for the knowledge points, skill points and operation processes in the embedded teaching module, and output the learning content of the embedded module that adapts to the student; these contents will be organized in a form suitable for the student's learning style, with reasonable complexity and structure level, to ensure that both the teaching objectives and the student's current cognitive state and growth path are met, thereby effectively supporting the implementation of personalized embedded learning.

[0093] After the student starts learning the corresponding embedded module learning content, the learning process is monitored and recorded in real time. This process includes the start time, completion time, interruption times, video viewing time, interactive operation frequency, code writing and debugging behavior, simulation experiment running results and other data collection in multiple dimensions, aiming to fully capture the student's actual learning behavior trajectory in the module; secondly, the student is guided to complete the key test questions, experimental operation questions and reflective questions embedded in the module, and the student's knowledge mastery and skill application ability are judged according to the student's answer accuracy, error type, submission delay and other conditions, and the system also extracts features from the error frequency and debugging time during the learning process to reflect the cognitive load state; then, the system inputs the above feedback data into the student portrait update model, dynamically adjusts and updates the student's current cognitive level, skill mastery ability, behavior habit and emotional tendency in combination with the previously existing learning portrait, thereby forming a more accurate learning portrait. The updated portrait will serve as an important basis for the next round of learning content generation and teaching strategy recommendation, ensuring that the teaching process continues to adapt to the student's development state and realizes the closed-loop optimization of personalized embedded teaching.

[0094] Taking embedded teaching as an example, in the intelligent electronic course of a certain university, a basic peripheral control teaching module is built around the STM32 development board, which mainly includes GPIO configuration, LED lighting experiment, key input reading and simple interrupt control; for a student, the student's learning portrait shows that the current cognitive load is at a medium to high level, the prior knowledge mainly masters the basic C language syntax and digital electronics principles, but has not systematically touched embedded development; the student's learning style preference is a combination of text and graphics, and is biased towards practical operation driven learning;

[0095] According to the portrait, first of all, based on the student's prior knowledge level, the semantic generation granularity of the teaching module is set to medium-fine granularity, i.e. the GPIO control process is step-by-step, such as explaining the clock enable, port configuration, pin direction setting and other sub-steps; the semantic generation preference adopts the flowchart priority strategy, with the development board pin diagram, circuit connection diagram and register configuration flowchart; the semantic generation density is controlled at a medium-low level, with appropriate insertion of annotation and common error prompt between key steps to avoid cognitive overload.

[0096] The finally generated personalized embedded learning content includes: a video explanation starting from the real wiring diagram of the development board, gradually guiding the student to configure the project in the Keil environment, complete the GPIO initialization setting in CubeMX, and manually supplement the LED control statement after exporting the code; The meaning of each register bit is shown in the form of a chart in the accompanying text material, and each stage of visual debugging screenshot is attached; After the learning content is put into use in actual teaching, the accuracy rate of code writing, the experiment completion time of the student, and the learning behavior trajectory data generated by the system are collected to feedback for optimizing the generation strategy of subsequent interrupt and timer modules, and finally a personalized teaching path from fixed modules to dynamic content generation is realized, which reflects the feasibility and adaptability of the method in actual embedded teaching.

[0097] Embodiment 2, a modular content generation system for embedded teaching, see Figure 1 as shown, comprising:

[0098] The embedded teaching module division module includes an initial module generation unit and an embedded teaching module generation unit; the initial module generation unit is used to obtain the original teaching content related to embedded teaching; the original teaching content is subjected to semantic analysis and structured processing, and the original teaching content is divided into a plurality of initial teaching modules; each initial teaching module is assigned a corresponding pre-skills set and output skills set; all initial teaching modules are sorted according to the teaching sequence relationship; the embedded teaching module generation unit is used to evaluate the learning fatigue of the initial teaching module based on the pre-skills set and the output skills set for any initial teaching module, and obtain the module fatigue evaluation result; compare the module fatigue evaluation result with the preset fatigue threshold value: if the module fatigue evaluation result exceeds the preset fatigue threshold value, the corresponding initial teaching module is further split to obtain at least two split embedded teaching modules; if the module fatigue evaluation result is much lower than the preset fatigue threshold value, a plurality of associated initial teaching modules are selected for combination to obtain a combined embedded teaching module;

[0099] Based on the step of comparing the module fatigue evaluation result with the preset fatigue threshold value, all initial teaching modules are compared and judged, and the obtained all split embedded teaching modules and combined embedded teaching modules are renamed to obtain a plurality of embedded teaching modules;

[0100] The embedded teaching content generation module includes a module content generation unit; the module content generation unit is used to generate corresponding embedded module learning content for the student learning portrait and the embedded teaching module to be learned; learning based on the corresponding embedded module learning content and collecting learning feedback data for updating the student learning portrait.

[0101] It is to be understood that all of the above modifications and alterations can be made to the above-described arrangements and that all such modifications and alterations are intended to be included within the scope of the present application. Those skilled in the art will readily appreciate that other modifications and alterations can be made to the present application without departing from the scope of the application.

Claims

1. A method for modular content generation for embedded instruction, the method comprising: The method comprises the following steps: Obtaining original teaching content related to embedded teaching; performing semantic analysis and structural processing on the original teaching content, dividing the original teaching content into a plurality of initial teaching modules; assigning a corresponding pre-skills set and output skills set to each initial teaching module; and sorting all initial teaching modules according to a teaching sequence relationship; For any initial teaching module, performing learning fatigue evaluation on the initial teaching module based on the pre-skills set and the output skills set to obtain a module fatigue evaluation result; Comparing the module fatigue evaluation result with a preset fatigue threshold value to determine whether the module fatigue evaluation result exceeds the preset fatigue threshold value; if the module fatigue evaluation result exceeds the preset fatigue threshold value, the corresponding initial teaching module is further split to obtain at least two split embedded teaching modules; If the module fatigue evaluation result is much lower than the preset fatigue threshold value, a plurality of associated initial teaching modules are combined to obtain a combined embedded teaching module; Based on the step of comparing the module fatigue evaluation result with the preset fatigue threshold value, all initial teaching modules are iteratively compared and determined, and all split embedded teaching modules and combined embedded teaching modules obtained are renamed to obtain a plurality of embedded teaching modules; Generating corresponding embedded module learning content for a student learning profile and embedded teaching modules to be learned; Learning based on the corresponding embedded module learning content and collecting learning feedback data for updating the student learning profile; The specific steps of assigning a corresponding pre-skills set and output skills set to each initial teaching module include: Identifying the first n initial teaching modules and the last n initial teaching modules of the initial teaching module; Performing semantic analysis on the teaching content contained in the initial teaching module to obtain a knowledge set corresponding to the initial teaching module; Performing semantic analysis on the first n initial teaching modules to obtain an output basic ability set, and performing semantic analysis on the last n initial teaching modules to obtain a pre-basic ability set; based on the dependency relationship between knowledge points, the knowledge set, the pre-basic ability set and the output basic ability set are comprehensively analyzed to obtain the pre-skills set and the output skills set of the initial teaching module.

2. The method for modular content generation for embedded instruction of claim 1, wherein, The specific steps of performing learning fatigue evaluation on the initial teaching module based on the pre-skills set and the output skills set include: Constructing a skill learning span model based on prior knowledge; quantitatively analyzing the skill difference between the pre-skills set and the output skills set to obtain a skill cognitive transition intensity value; based on the average cognitive load bearing capacity of the students, determining a preset fatigue threshold value based on the average cognitive load bearing capacity; Calculating the number of skill conversion paths and path lengths contained in the initial teaching module, and quantifying the number of skill conversion paths and path lengths into a learning fatigue index; Feature fusion of the skill cognitive transition intensity value and the learning fatigue index to obtain a module fatigue evaluation result.

3. The method of claim 2, wherein, If the module fatigue evaluation result exceeds the preset fatigue threshold value, the specific steps include: Based on the skill span distribution between the pre-skills set and the output skills set, identifying the skill transition mutation point in the initial teaching module; based on the skill transition mutation point, determining a cognitive overload sub-region in the initial teaching module; The learning content link graph of the cognitive overload sub-region is generated by using a graph neural network; clustering analysis is performed based on the learning content link graph, and a teaching sub-module candidate set with a long operation chain is extracted; The number of shortest paths, longest paths and cross paths of the learning content link graph of the learning content link graph is identified, and the teaching sub-module candidate set is balanced and split based on the identified number to obtain at least two split embedded teaching modules; The split embedded teaching modules after splitting are re-established to establish a teaching sequence relationship, and the pre-skills set and the output skills set corresponding to the split embedded teaching modules are updated.

4. The method of claim 3, wherein, If the module fatigue evaluation result is much lower than the preset fatigue threshold, the specific steps include: Extracting other initial teaching modules that have similarity with the output skills set of the current initial teaching module and whose module fatigue evaluation results are all much lower than the preset fatigue threshold, as associated initial teaching modules; Identifying the learning content link graph of the plurality of associated initial teaching modules, and constructing a module merging candidate pair corresponding to the current initial teaching module; Introducing a learning continuity scoring model; using the learning continuity scoring model to screen the module merging candidate pairs, and selecting the module merging candidate pairs that meet the preset learning continuity requirements; Combining the module merging candidate pairs that meet the preset learning continuity requirements to obtain a combined embedded teaching module; re-establishing the teaching sequence relationship of the combined embedded teaching module, and updating the pre-skills set and the output skills set corresponding to the combined embedded teaching module.

5. The method of claim 4, wherein, For the student learning portrait and the embedded teaching module to be learned, the specific steps of generating the corresponding embedded module learning content include: Obtaining a student learning portrait; the student learning portrait includes current learning cognitive load, learning style preference and prior knowledge level; matching semantic generation granularity based on the prior knowledge level; generating semantic generation preference based on the learning style preference; matching semantic generation density based on the current learning cognitive load; Generating the corresponding embedded module learning content of the embedded teaching module based on the semantic generation granularity, the semantic generation preference and the semantic generation density.

6. A modular content generation system for embedded instruction, characterized by, The system applies the modular content generation method for embedded teaching of any one of claims 1-5, including: The embedded teaching module division module comprises an initial module generation unit and an embedded teaching module generation unit; the initial module generation unit is configured to obtain original teaching content related to embedded teaching; the original teaching content is subjected to semantic analysis and structural processing, and the original teaching content is divided into a plurality of initial teaching modules; each initial teaching module is assigned a corresponding prerequisite skill set and an output skill set; all initial teaching modules are sorted according to a teaching sequence relationship; the embedded teaching module generation unit is configured to, for any initial teaching module, perform learning fatigue assessment on the initial teaching module based on the prerequisite skill set and the output skill set to obtain a module fatigue assessment result; the module fatigue assessment result is compared with a preset fatigue threshold; if the module fatigue assessment result exceeds the preset fatigue threshold, the corresponding initial teaching module is further split to obtain at least two split embedded teaching modules; if the module fatigue assessment result is much lower than the preset fatigue threshold, a plurality of associated initial teaching modules are combined to obtain a combined embedded teaching module; Based on the step of comparing the module fatigue assessment result with the preset fatigue threshold, all initial teaching modules are iteratively compared and judged, and all split embedded teaching modules and combined embedded teaching modules obtained are renamed to obtain a plurality of embedded teaching modules; The embedded teaching content generation module comprises a module content generation unit; the module content generation unit is configured to generate corresponding embedded module learning content for a student learning profile and embedded teaching modules to be learned; learning is performed based on the corresponding embedded module learning content, and learning feedback data is collected for updating the student learning profile.

Images