Foreign language teaching resource intelligent planning method and system based on knowledge graph

Abstract

The invention discloses an intelligent foreign language teaching resource planning method and system based on a knowledge graph, and relates to the technical field of foreign language teaching and intelligence. The planning method comprises the following steps: constructing a knowledge system and data acquisition: collecting foreign language teaching knowledge, carrying out preprocessing, constructing a foreign language knowledge graph by means of a graph database, and carrying out data acquisition; the method comprises the following steps: establishing an association relationship between knowledge, acquiring learning feature data of learning basis, style, progress and preference of students through student registration information, learning behavior data acquisition and learning ability test, and acquiring daily curriculum schedule and idle time data of the students in real time; the foreign language knowledge graph is constructed, scattered teaching knowledge is systematically integrated, network association among knowledge is formed, and learning requirements of students can be accurately positioned and intelligent screening and efficient adaptation of teaching resources can be realized in combination with learning feature data, including learning basis, style, progress and preference, of the students.

Description

Technical Field

[0001] The present invention relates to the field of foreign language teaching and intelligent technology, and specifically to a method and system for intelligent planning of foreign language teaching resources based on knowledge graphs. Background Art

[0002] In today's era of deepening globalization, the importance of foreign language education has become increasingly prominent. With the rapid development of information technology, the traditional foreign language teaching model is undergoing profound changes. In order to adapt to this change and improve the efficiency and effectiveness of foreign language teaching, educators and technology developers are constantly exploring the application of new technologies in teaching. Among them, knowledge graphs, as an emerging information organization and management technology, with their powerful knowledge representation and association capabilities, provide new ideas for the intelligent planning and management of foreign language teaching resources. Knowledge graphs can systematically integrate scattered teaching knowledge and form network connections between knowledge, thereby providing students with more personalized and accurate learning resource recommendations.

[0003] In traditional foreign language teaching, the screening and adaptation optimization of teaching resources and the planning of learning paths mainly rely on the teacher's experience and subjective judgment, and there are obvious limitations. In terms of resource screening and adaptation optimization, it is difficult for teachers to fully grasp the learning characteristics, progress and preferences of each student, and it often lacks systematicity and scientificity, resulting in inaccurate resource recommendations and an inability to meet students' personalized needs. The planning of learning paths is relatively fixed, lacking flexibility and real-time performance, and it is difficult to make timely adjustments based on students' learning status and progress.

[0004] Therefore, a method and system for intelligent planning of foreign language teaching resources based on knowledge graph is developed. Summary of the Invention

[0005] The purpose of this invention is to make up for the shortcomings of the existing technology and provide a method and system for intelligent planning of foreign language teaching resources based on knowledge graphs. By constructing a foreign language knowledge graph to integrate teaching knowledge, and collecting students' learning characteristics and time data, the knowledge graph is used to compare and analyze student data to accurately locate learning needs, realize intelligent screening and adaptive optimization of teaching resources, and at the same time, plan personalized learning paths and perform dynamic scheduling to ensure the rationality and efficiency of learning paths.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: a method for intelligent planning of foreign language teaching resources based on knowledge graphs, the specific steps of the planning method are:

[0007] Building a knowledge system and data collection: We collect and preprocess foreign language teaching knowledge, build a foreign language knowledge graph with the help of a graph database, establish relationships between knowledge, and obtain learning characteristic data on students' learning foundation, style, progress, and preferences through student registration information, learning behavior data collection, and learning ability tests. We also collect real-time data on students' daily schedules and free time.

[0008] Demand analysis and status prediction: Compare students' learning characteristics with the foreign language knowledge graph to analyze and determine their current learning needs. Combined with learning characteristic data and time data, conduct multi-dimensional analysis and prediction of students' future learning status to determine whether their learning progress is lagging or ahead of schedule.

[0009] Resource screening and adaptation optimization: Preliminary screening of foreign language teaching resources in the teaching resource library based on learning needs, matching the teaching resources with learning needs, and secondary adaptation of the screened resources based on the student's daily study time and real-time schedule;

[0010] Path planning and dynamic scheduling: Using the path value assessment formula, we plan personalized learning paths based on students' learning characteristics, needs, schedules, and matching resources. We build and run a three-dimensional matrix model, make dynamic adjustments through resource adjustment decision formulas, and continuously monitor learning progress.

[0011] Resource push and closed-loop optimization: Push teaching resources in the planned learning path to students through the learning platform, collect students' feedback on resource quality, difficulty, and practicality, combine the three-dimensional matrix model with learning effects, re-analyze learning needs, and adjust learning paths and resource recommendations.

[0012] Furthermore, in the demand analysis and status prediction, the student learning feature data is mapped to the semantic space of the foreign language knowledge graph, the knowledge graph node features are extracted, the graph feature matrix is ​​obtained, the student learning feature vectors are compared with the graph feature matrix, the nodes where the student's knowledge mastery is weak are located, and the located weak nodes and their related knowledge nodes are integrated and analyzed in combination with the data of the student's learning style and learning goals to clarify the student's current learning needs in the specific knowledge areas of vocabulary, grammar, sentence patterns, and cultural background, and determine the final learning needs.

[0013] Furthermore, in the demand analysis and state prediction, the situation of learning progress lagging behind or ahead is judged, wherein: the learning feature data and time data are input into the dynamic learning trend model, the correlation between the learning feature data and time data is analyzed, the trend of learning state changing over time is simulated, and the actual learning progress data X of the past i moments is extracted from the learning feature data. t-i , extract the time features T of the past j moments from the time data t-j, the formula of the dynamic learning trend model is: in, is the predicted state at time t, is the predicted learning state at time t+1, p is the number of moments when actual learning progress data is extracted from learning feature data, q is the number of moments when past time features are extracted from time data, i is the historical moment index of learning progress data, j is the historical moment index of time feature data, δ, θ i 、φ j The predicted learning status data output by the model is compared with the standard learning progress data. The progress range of the standard learning theoretical progress data is 80%-120%. If the predicted data is lower than this range, it is judged that the learning progress is lagging behind. If the predicted data is higher than this range, it is judged that the learning progress is ahead.

[0014] Furthermore, in the resource screening and adaptation optimization, the teaching resource library contains a variety of resources such as electronic textbooks, teaching videos, audio materials and exercises. Based on the students' learning needs, the resource matching score formula is used to perform preliminary resource screening and matching, and the matching score between each resource in the teaching resource library and the learning needs in the knowledge dimension is calculated. The resources are preliminarily sorted according to the score, and the resources with a score higher than the threshold T are screened. S As teaching resources, T S It is a dynamic multi-dimensional adaptive resource screening threshold, and records the available learning time of the teaching resources. It judges the course density based on the total course duration and course interval of the student on that day, introduces the time-resource adaptation coefficient, and calculates the adaptation coefficient of the remaining resources and the student's free time periods on that day, so as to perform secondary adaptation.

[0015] Furthermore, in the resource screening and adaptation optimization, a resource matching score formula is used for preliminary screening and matching. The resource matching score formula is: Among them, S rs is the matching score between resource r and student demand s, I rk is the information content of resource r in knowledge dimension k, D sk is the expected degree of student demand in knowledge dimension k, μ k is the knowledge dimension weight, which is determined by the structural importance analysis of the knowledge graph, l is the total number of knowledge dimensions, and k is the index variable of the knowledge dimension.

[0016] Furthermore, in the resource screening and adaptation optimization, a time-resource adaptation coefficient formula is introduced to perform secondary adaptation, and the formula is: Among them, A tr is the adaptation coefficient between time t and resource r, τ t is the available learning time at time t, σ ris the recommended learning time for resource r;

[0017] The secondary adaptation rule is: when A tr >0.7, it is determined that the resource r is highly adapted to time t, and no secondary adaptation is performed. tr <0.7, it is determined that the resource r is poorly adapted to time t, and the resource A in the candidate resource set is selected. tr The value is adjusted twice, and resources are screened and matched again, where 0.7 is the critical value for whether it is adapted or not.

[0018] Furthermore, in the path planning and dynamic scheduling, the specific steps of planning the personalized learning path are:

[0019] The relevant data on students' learning characteristics, needs, time arrangements and matching resources are used as input parameters and substituted into the path value evaluation formula to calculate the value scores of different potential learning paths. Based on the scores, the path with the highest value is selected as the personalized learning path;

[0020] The specific steps of dynamic scheduling are:

[0021] Based on the personalized learning path, data of time dimension, progress dimension and resource dimension are extracted to construct a three-dimensional matrix model. The actual learning progress data is then compared with the progress data predicted in the model. When deviations occur, the resource adjustment decision formula is used to calculate the decision value of the resource adjustment plan based on the degree of deviation, remaining time and resource attribute parameters. The resources in the learning path are dynamically adjusted based on the decision value. When the progress lags behind, according to the knowledge dependency relationship in the knowledge graph, step-by-step consolidation resources from basic to advanced are inserted in the students' subsequent free time. When the progress is ahead of schedule, advanced tasks are pushed based on the students' current mastery and the extended knowledge nodes in the knowledge graph. After each adjustment, the three-dimensional matrix model is updated at the same time to continuously monitor the learning progress.

[0022] Furthermore, in the path planning and dynamic scheduling, a personalized learning path is planned through a path value evaluation formula, wherein the path value evaluation formula is: V p =α·S match +β·T fit +γ·R qual , where V p is the value evaluation score of path p, α, β, γ are weight coefficients, S match is the knowledge demand matching degree, which reflects the fit between the path and students’ learning characteristics and needs. fit is the time fit, which measures the degree of matching between the path and the student's schedule, R qual is the resource quality coefficient, which evaluates the comprehensive quality of resources in the path.

[0023] Furthermore, in the path planning and dynamic scheduling, the resource adjustment decision formula is: Among them, D adj is the resource adjustment decision variable, 1 indicates that adjustment is needed, 0 indicates that no adjustment is needed, ΔX is the deviation vector between the actual learning progress and the predicted progress, ∈1 and ∈2 are the progress deviation thresholds, which are dynamically set according to the learning stage and knowledge difficulty.

[0024] On the other hand, a foreign language teaching resource intelligent planning system based on knowledge graph includes:

[0025] Knowledge graph construction module: collects foreign language teaching knowledge, performs preprocessing, and constructs a foreign language knowledge graph;

[0026] Data acquisition module: This module obtains students' learning characteristics and collects data on their daily schedules and free time in real time. The module has an interface with the school's academic affairs system and has the function of collecting data from the learning platform.

[0027] Learning needs and status analysis module: This module combines knowledge graphs and learning characteristics to analyze students' learning needs. Furthermore, it combines learning progress data and time data to conduct multi-dimensional analysis and prediction of students' learning status.

[0028] Resource screening and matching module: Screen and match teaching resources in the teaching resource library according to learning needs, and optimize the matching based on students' schedules;

[0029] Learning path planning and dynamic scheduling module: Uses path value assessment formulas to plan personalized learning paths, builds and runs a three-dimensional matrix model, and dynamically allocates and adjusts resources;

[0030] Resource push and feedback module: push teaching resources to students, collect students' learning feedback information, and optimize and adjust learning paths and resource recommendations.

[0031] Compared with existing technologies, this literature scene experiential teaching method and system has the following beneficial effects:

[0032] 1. The present invention systematically integrates scattered teaching knowledge by constructing a foreign language knowledge graph, forming a network connection between knowledge, and combining student learning feature data, including learning foundation, style, progress and preferences, to accurately locate students' learning needs, realize intelligent screening and efficient adaptation of teaching resources, and ensure that the recommended teaching resources not only highly match students' learning needs in the knowledge dimension, but also fully consider students' time arrangements, realize personalized and precise learning resource recommendations, and greatly improve the utilization efficiency of learning resources and learning effects.

[0033] 2. The present invention realizes personalized planning and dynamic adjustment of learning paths through the constructed three-dimensional matrix model and path value evaluation and resource adjustment decision-making technology. It can adjust the resource allocation in the learning path in real time according to the student's learning status and progress, ensuring the rationality and efficiency of the learning path. The dynamic planning of the learning path not only improves learning efficiency, but also increases learning flexibility, thereby achieving better learning results.

[0034] Other advantages, objects and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art based on an examination of the following or may be learned from the practice of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort.

[0036] Figure 1 This is a flowchart of a method for intelligent planning of foreign language teaching resources based on knowledge graph;

[0037] Figure 2 This is a framework diagram of an intelligent planning system for foreign language teaching resources based on knowledge graph. DETAILED DESCRIPTION

[0038] In order to further illustrate the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0039] Example 1:

[0040] Constructing a knowledge system and collecting data: The knowledge graph construction module collects the core knowledge of university public English courses, such as basic vocabulary, grammatical tenses, listening, speaking, reading and writing skills, and cultural background. After classification and annotation, it is imported into the graph database to construct a knowledge graph containing the "vocabulary-grammar-sentence pattern-chapter" relationship. The data acquisition module synchronizes students' English class scores through the school's academic affairs system interface, and collects students' daily learning behavior data through the learning platform, such as the duration of listening practice and the frequency of essay submission. It also captures students' class schedules in real time. For example, there are professional classes in the morning from Monday to Friday, public English classes from 3 to 5 in the afternoon, and free time from 7 to 9 in the evening, which are integrated to form learning characteristics and time data.

[0041] Demand Analysis and Status Prediction: Map student learning characteristics, such as listening scores below the class average and insufficient mastery of writing templates, to the knowledge graph to identify weak points. Combined with student learning goals, such as passing the final exam, the needs are clearly defined as "intensive listening training for short conversations" and "practice applying argumentative essay templates." The learning demand and status analysis module simulates learning trends through a dynamic learning trend model based on historical learning data and time data, including the duration of listening practice over the past two weeks, changes in accuracy, and daily idle time. The formula is: in, is the predicted state at time t, is the predicted learning state at time t+1, p is the number of moments when actual learning progress data is extracted from learning feature data, q is the number of moments when past time features are extracted from time data, i is the historical moment index of learning progress data, j is the historical moment index of time feature data, δ, θ i 、 As the weight parameter, the predicted learning status data output by the model is compared with the standard learning progress data. In the standard progress, 80% of the grammar unit learning should be completed by the middle of the semester. If the predicted completion rate is only 50%, it is determined to be lagging behind. Figure 1 shown.

[0042] Resource screening and adaptation optimization: Based on the requirements of "Listening Comprehension" and "Argumentative Essay Template", the matching degree between resources and requirements is calculated using the resource matching score formula. The formula is: Among them, S rs is the matching score between resource r and student demand s, I rk is the information content of resource r in knowledge dimension k, D sk is the expected degree of student demand in knowledge dimension k, μ k is the knowledge dimension weight, determined by the structural importance analysis of the knowledge graph, l is the total number of knowledge dimensions, and k is the index variable of the knowledge dimension. Then the resource screening and matching module retrieves the corresponding resources from the library, including the BBC Six Minutes English Listening Audio and the Argumentative Essay Template Collection, sorts them by knowledge matching degree, and selects those with scores higher than the threshold T. S Resources such as intensive listening audio containing high-frequency test vocabulary and model essays covering test questions.

[0043] Time adaptation: Based on the student's free time between 7:00 and 9:00 p.m. (available time is 2 hours), we calculate the recommended duration of resources (e.g., 80 minutes for intensive audio learning and 100 minutes for model essay learning). We then use the time-resource adaptation coefficient to determine the degree of adaptation. The formula is: Among them, A tr is the adaptation coefficient between time t and resource r, τ t is the available learning time at time t, σr is the recommended learning time for resource r, if A tr <0.7, resource recommendation duration σ r = 3 hours, calculate A tr =0.6, then replace it with shorter resources, such as splitting intensive audio learning and model article learning into multiple sections.

[0044] Path planning and dynamic scheduling: The learning path planning and dynamic scheduling module takes the relevant data of students' learning characteristics, needs, time arrangements and matching resources as input parameters and substitutes them into the path value evaluation formula: V p =α·S match +β·T fit +γ·R qual , where V p is the value evaluation score of path p, α, β, γ are weight coefficients, S match is the knowledge demand matching degree, which reflects the fit between the path and students’ learning characteristics and needs. fit is the time fit, which measures the degree of matching between the path and the student's schedule, R qual The resource quality coefficient evaluates the overall quality of resources in the path, thereby generating a personalized path: at 7 pm that night, a 40-minute audio lesson will be pushed (matching listening needs), and at 8 pm, a 50-minute argumentative essay sample will be pushed (matching writing needs);

[0045] If students report that the intensive listening audio is too fast, the learning path planning and dynamic scheduling module triggers adjustments through the resource adjustment decision formula. The resource adjustment decision formula is: Among them, D adj is the resource adjustment decision variable, 1 indicates that adjustment is needed, 0 indicates that no adjustment is needed, ΔX is the deviation vector between the actual learning progress and the predicted progress, ∈1 and ∈2 are the progress deviation thresholds, which are dynamically set according to the learning stage and knowledge difficulty. The "high-frequency vocabulary shorthand" resource is inserted during the idle time of the next day, and the three-dimensional matrix model is updated at the same time to continuously monitor the subsequent progress.

[0046] Resource push and closed-loop optimization: Push resources along the path through the learning platform, automatically send intensive listening audio links and practice questions at 7 o'clock, push model documents and structural analysis diagrams at 8 o'clock, collect student feedback (such as "the model template is stiff"), combine with the three-dimensional matrix model analysis, reposition the demand as "logical connection training for argumentative essays", adjust the path to push "logical word collocation practice" resources, and update the related resources of the "Writing-Logical Connection" node in the knowledge graph.

[0047] To sum up, in the university public English course, the foreign language knowledge graph is constructed and relevant data is collected through the collaborative work of the knowledge graph construction module and the data acquisition module. Then, the learning demand and status analysis module uses the dynamic learning trend model to analyze the demand and predict the progress. Then, the resource screening and matching module uses the resource matching score formula and the time-resource adaptation coefficient formula to process resources. Then, the path value evaluation formula and resource adjustment decision formula of the learning path planning and dynamic scheduling module are used to generate and adjust the path. Finally, the resource push and feedback module completes the push and optimization, realizing the intelligent planning of university public English course resources and improving learning efficiency.

[0048] Example 2:

[0049] Constructing a knowledge system and collecting data: The knowledge graph construction module integrates children's English enlightenment knowledge, including the pronunciation of 26 letters, basic words, and simple sentence patterns, to construct a visual graph. At the same time, the data acquisition module obtains the age (such as 6 years old) and English foundation before school entry (none) through student registration information, collects classroom interaction data (such as the accuracy of letter spelling and the enthusiasm for oral reading) through the learning platform, and collects class schedules in real time (English classes are from 9 to 11 am on weekends and free time after school from 4 to 5 pm on weekdays) to form learning characteristics (concrete learning preferences and short-term attention concentration) and time data.

[0050] Demand Analysis and Status Prediction: Map student characteristics (high error rate in pronouncing the letter A and strong interest in animal words) to the graph, identify weak points ("pronunciation of the letter A" and "recognition of animal words"), and combine this with the learning goal (mastering 20 basic animal words) to identify the needs as "strengthening pronunciation of the letter A" and "memorizing animal words through fun." The learning demand and status analysis module uses last week's learning data (5 words memorized daily, 60% accuracy) and time data (1 hour of free time on weekday afternoons) to implement a dynamic learning trend model: Simulate the progress and compare it with the standard (14 words should be mastered in two weeks). If it is predicted that only 8 words will be mastered, it is judged that the progress is lagging behind.

[0051] Resource screening and adaptation optimization: Based on the demand, the resource matching scoring formula is used: Calculate the resource matching degree, retrieve "Letter A Pronunciation Animation" and "Animal Word Puzzle Game" from the resource library, sort them by matching degree, and filter out the ones with scores higher than the threshold T. S Resources: animations with pronunciation and lip-sync demonstrations, puzzle games with word spelling prompts, and combined with the free time between 4 and 5 pm on weekdays (1 hour), using the time-resource adaptation coefficient formula: Calculate the recommended duration of resources (20 minutes per episode for animation and 20 minutes per round for puzzle games), ensuring that a single learning session does not exceed 30 minutes (suitable for children's attention span). If A tr <0.7, such as resource recommendation duration σ r = 40 minutes, calculate A tr =0.4, then replace it with a shorter resource, such as a 10-minute animation + a 20-minute game.

[0052] Path planning and dynamic scheduling: Learn the path planning and dynamic scheduling module through the path value evaluation formula: V p =α·S match +β·T fit +γ·R qual , generate learning paths: push the letter A pronunciation animation (10 minutes) at 4 pm on Monday, push the animal word puzzle game (20 minutes) at 4:15 pm;

[0053] If students frequently make mistakes in the puzzle game, the learning path planning and dynamic scheduling module uses the resource adjustment decision formula: Dynamically adjust the resources in the learning path, and insert the "Letter A Tracing Practice" resource the next day based on the knowledge graph dependency. At the same time, shorten the game time, increase pronunciation repetition training, and update the three-dimensional matrix model, such as Figure 2 shown.

[0054] Resource push and closed-loop optimization: The learning platform pushes animation videos (with pronunciation and follow-up functions) and game links (spelling must be completed to unlock new levels) according to the path, and collects parental feedback ("children have low interest in animation"). Combined with three-dimensional matrix model analysis, the demand is repositioned as "a more interactive learning method", and the path is adjusted to push "Letter A Puppet Show" (real-person demonstration of pronunciation) and "Animal Word Real-life Cards", and the resource association of the "Letter Learning-Interactive Form" node in the map is updated.

[0055] To sum up, in the children's English enlightenment scenario, the knowledge graph construction module and the data acquisition module first construct the foreign language knowledge graph, and collect students' basic, behavioral and time data. The learning needs and status analysis module locates the weak knowledge nodes and uses the dynamic learning trend model to judge the progress. The resource screening and matching module uses the resource matching score formula and the time-resource adaptation coefficient formula to screen and adapt resources. The learning path planning and dynamic scheduling module uses the path value evaluation formula to generate the path, and uses the resource adjustment decision formula to deal with progress deviations. Finally, the resource push and feedback module completes the push and optimization, realizing intelligent resource planning that conforms to the characteristics of children's learning.

[0056] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as above in terms of a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art can, without departing from the scope of the technical solution of the present invention, make some changes or modifications to equivalent embodiments using the technical contents disclosed above. However, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are within the scope of the technical solution of the present invention.

Claims

1. A method for intelligent planning of foreign language teaching resources based on knowledge graph, characterized in that: The specific steps of this planning method are: Building a knowledge system and data collection: We collect and preprocess foreign language teaching knowledge, build a foreign language knowledge graph with the help of a graph database, establish relationships between knowledge, and obtain learning characteristic data on students' learning foundation, style, progress, and preferences through student registration information, learning behavior data collection, and learning ability tests. We also collect real-time data on students' daily schedules and free time. Demand analysis and status prediction: Compare students' learning characteristics with the foreign language knowledge graph to analyze and determine their current learning needs. Combined with learning characteristic data and time data, conduct multi-dimensional analysis and prediction of students' future learning status to determine whether their learning progress is lagging or ahead of schedule. Resource screening and adaptation optimization: Preliminary screening of foreign language teaching resources in the teaching resource library based on learning needs, matching the teaching resources with learning needs, and secondary adaptation of the screened resources based on the student's daily study time and real-time schedule; Path planning and dynamic scheduling: Using the path value assessment formula, we plan personalized learning paths based on students' learning characteristics, needs, schedules, and matching resources. We build and run a three-dimensional matrix model, make dynamic adjustments through resource adjustment decision formulas, and continuously monitor learning progress. Resource push and closed-loop optimization: Push teaching resources in the planned learning path to students through the learning platform, collect students' feedback on resource quality, difficulty, and practicality, combine the three-dimensional matrix model with learning effects, re-analyze learning needs, and adjust learning paths and resource recommendations.

2. The method for intelligent planning of foreign language teaching resources based on knowledge graph according to claim 1, characterized in that: In the demand analysis and status prediction, the student learning feature data is mapped to the semantic space of the foreign language knowledge graph, the knowledge graph node features are extracted, the graph feature matrix is ​​obtained, the student learning feature vector is compared with the graph feature matrix, the nodes where the student's knowledge mastery is weak are located, and the located weak nodes and their related knowledge nodes are integrated and analyzed in combination with the data of the student's learning style and learning goals to clarify the student's current learning needs in the specific knowledge areas of vocabulary, grammar, sentence patterns, and cultural background, and determine the final learning needs.

3. The method for intelligent planning of foreign language teaching resources based on knowledge graph according to claim 1, characterized in that: In the demand analysis and state prediction, the learning progress lags behind or advances. Here, the learning feature data and time data are input into the dynamic learning trend model, the correlation between the learning feature data and time data is analyzed, the trend of learning state changing over time is simulated, and the actual learning progress data X at the past i moments is extracted from the learning feature data. t-i , extract the time features T of the past j moments from the time data t-j , the formula of the dynamic learning trend model is: in, is the predicted state at time t, is the predicted learning state at time t+1, p is the number of moments when actual learning progress data is extracted from learning feature data, q is the number of moments when past time features are extracted from time data, i is the historical moment index of learning progress data, j is the historical moment index of time feature data, δ, θ i 、φ j The predicted learning status data output by the model is compared with the standard learning progress data. The progress range of the standard learning theoretical progress data is 80%-120%. If the predicted data is lower than this range, it is judged that the learning progress is lagging behind. If the predicted data is higher than this range, it is judged that the learning progress is ahead.

4. The method for intelligent planning of foreign language teaching resources based on knowledge graph according to claim 1, characterized in that: In the resource screening and adaptation optimization, the teaching resource library contains a variety of resources such as electronic textbooks, teaching videos, audio materials and exercises. Based on the students' learning needs, the resource matching score formula is used to perform preliminary resource screening and matching, and the matching score between each resource in the teaching resource library and the learning needs in the knowledge dimension is calculated. The resources are preliminarily sorted according to the score, and the resources with a score higher than the threshold T are screened. S As teaching resources, T S It is a dynamic multi-dimensional adaptive resource screening threshold, and records the available learning time of the teaching resources. It judges the course density based on the total course duration and course interval of the student on that day, introduces the time-resource adaptation coefficient, and calculates the adaptation coefficient of the remaining resources and the student's free time periods on that day, so as to perform secondary adaptation.

5. The method for intelligent planning of foreign language teaching resources based on knowledge graph according to claim 4 is characterized in that: In the resource screening and adaptation optimization, a resource matching score formula is used for preliminary screening and matching. The resource matching score formula is: Among them, S rs is the matching score between resource r and student demand s, I rk is the information content of resource r in knowledge dimension k, D sk is the expected degree of student demand in knowledge dimension k, μ k is the knowledge dimension weight, which is determined by the structural importance analysis of the knowledge graph, l is the total number of knowledge dimensions, and k is the index variable of the knowledge dimension.

6. The method for intelligent planning of foreign language teaching resources based on knowledge graph according to claim 4, characterized in that: In the resource screening and adaptation optimization, the time-resource adaptation coefficient formula is introduced for secondary adaptation, and the formula is: Among them, A tr is the adaptation coefficient between time t and resource r, τ t is the available learning time at time t, σ r is the recommended learning time for resource r; The secondary adaptation rule is: when A tr >0.7, it is determined that the resource r is highly adapted to time t, and no secondary adaptation is performed. tr <0.7, it is determined that the resource r is poorly adapted to time t, and the resource A in the candidate resource set is selected. tr The value is adjusted twice, and resources are screened and matched again, where 0.7 is the critical value for whether it is adapted or not.

7. The method for intelligent planning of foreign language teaching resources based on knowledge graph according to claim 1, characterized in that: In the path planning and dynamic scheduling, the specific steps for planning a personalized learning path are as follows: The relevant data on students' learning characteristics, needs, time arrangements and matching resources are used as input parameters and substituted into the path value evaluation formula to calculate the value scores of different potential learning paths. Based on the scores, the path with the highest value is selected as the personalized learning path; The specific steps of dynamic scheduling are: Based on the personalized learning path, data of time dimension, progress dimension and resource dimension are extracted to construct a three-dimensional matrix model. The actual learning progress data is then compared with the progress data predicted in the model. When deviations occur, the resource adjustment decision formula is used to calculate the decision value of the resource adjustment plan based on the degree of deviation, remaining time and resource attribute parameters. The resources in the learning path are dynamically adjusted based on the decision value. When the progress lags behind, according to the knowledge dependency relationship in the knowledge graph, step-by-step consolidation resources from basic to advanced are inserted in the students' subsequent free time. When the progress is ahead of schedule, advanced tasks are pushed based on the students' current mastery and the extended knowledge nodes in the knowledge graph. After each adjustment, the three-dimensional matrix model is updated at the same time to continuously monitor the learning progress.

8. The method for intelligent planning of foreign language teaching resources based on knowledge graph according to claim 7, characterized in that: In the path planning and dynamic scheduling, the personalized learning path is planned through the path value evaluation formula, where the path value evaluation formula is: V p =α·S match +β·T fit +γ·R qual , where V p is the value evaluation score of path p, α, β, γ are weight coefficients, S match is the knowledge demand matching degree, which reflects the fit between the path and students’ learning characteristics and needs. fit is the time fit, which measures the degree of matching between the path and the student's schedule, R qual is the resource quality coefficient, which evaluates the comprehensive quality of resources in the path.

9. The method for intelligent planning of foreign language teaching resources based on knowledge graph according to claim 7, characterized in that: In the path planning and dynamic scheduling, the resource adjustment decision formula is: Among them, D adj is the resource adjustment decision variable, 1 indicates that adjustment is needed, 0 indicates that no adjustment is needed, ΔX is the deviation vector between the actual learning progress and the predicted progress, ∈1 and ∈2 are the progress deviation thresholds, which are dynamically set according to the learning stage and knowledge difficulty.

10. A foreign language teaching resource intelligent planning system based on knowledge graph, the system being applicable to the foreign language teaching resource intelligent planning method based on knowledge graph according to any one of claims 1 to 9, characterized in that: The system includes: Knowledge graph construction module: collects foreign language teaching knowledge, performs preprocessing, and constructs a foreign language knowledge graph; Data acquisition module: This module obtains students' learning characteristics and collects data on their daily schedules and free time in real time. The module has an interface with the school's academic affairs system and has the function of collecting data from the learning platform. Learning needs and status analysis module: This module combines knowledge graphs and learning characteristics to analyze students' learning needs. Furthermore, it combines learning progress data and time data to conduct multi-dimensional analysis and prediction of students' learning status. Resource screening and matching module: Screen and match teaching resources in the teaching resource library according to learning needs, and optimize the matching based on students' schedules; Learning path planning and dynamic scheduling module: Uses path value assessment formulas to plan personalized learning paths, builds and runs a three-dimensional matrix model, and dynamically allocates and adjusts resources; Resource push and feedback module: push teaching resources to students, collect students' learning feedback information, and optimize and adjust learning paths and resource recommendations.

Images