Intelligent study interaction platform based on AI

Abstract

The invention discloses an intelligent research interaction platform based on AI, and relates to the technical field of research interaction. Comprising a collaborative interaction module, a task analysis module, a learning condition analysis module and an intelligent pushing module. Through keyword extraction and matching degree calculation, the collaborative interaction module realizes rapid and accurate response of student questions, immediately feeds back knowledge base answers when the matching degree reaches the standard, and circulates teachers and supplements a knowledge base when the matching degree is insufficient, so that the question solving efficiency is improved, and knowledge resources are dynamically enriched; the task analysis module performs multi-dimensional disassembly and hierarchical scoring on research and study task results, comprehensively and accurately evaluates the performance of students in the aspects of knowledge mastering, practical operation, innovative thinking and the like, and provides objective evaluation basis for teachers and students; the learning condition analysis module accurately diagnoses weak points and ability dimension performance of students in combination with the questioning frequency and the score trend, and the intelligent pushing module pushes targeted teaching resources based on the learning condition and resource feature matching degree to assist teachers in implementing personalized teaching.

Description

Technical Field

[0001] This invention relates to the field of interactive learning technology, specifically to an AI-based intelligent interactive learning platform. Background Technology

[0002] In the field of study tour education, traditional teaching models have many limitations. On the one hand, when students encounter problems during the study tour, they often rely on teachers' immediate answers. If teachers cannot respond promptly, it will hinder the students' exploration process. Moreover, students' questions are diverse, making it difficult for teachers to quickly and accurately match knowledge resources, resulting in low efficiency in answering questions. On the other hand, the evaluation of students' study tour achievements mostly relies on manual scoring, which is not only time-consuming and laborious but also has a single evaluation dimension, making it difficult to comprehensively and objectively reflect students' performance in multiple dimensions such as knowledge mastery, practical operation, and innovative thinking. At the same time, in terms of learning analysis, traditional methods lack a systematic analysis of the frequency of students' questions and the scoring trends of task dimensions. Teachers find it difficult to accurately grasp students' knowledge weaknesses and ability development trends, making it impossible to achieve targeted teaching.

[0003] In addition, the delivery of teaching resources lacks intelligent means and is mostly selected by teachers based on experience, which makes it difficult to accurately match with students' specific learning situations. This results in low efficiency in the use of teaching resources and fails to effectively assist teachers in carrying out personalized teaching.

[0004] Therefore, there is an urgent need for an AI-based intelligent study tour interaction platform that integrates functions such as collaborative interaction, task analysis, learning analysis, and intelligent push notifications to break through the bottlenecks of traditional study tour education, improve the intelligence and personalization of study tour teaching, and help students learn efficiently and teachers teach precisely. Summary of the Invention

[0005] (a) Technical problems to be solved To address the shortcomings of existing technologies, this invention provides an AI-based intelligent research and learning interaction platform, which solves the problems mentioned in the background section.

[0006] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: an AI-based intelligent research and learning interaction platform, comprising: Collaborative Interaction Module: After a student inputs a question, the platform extracts keywords and retrieves knowledge information from a pre-set knowledge base based on the keywords. It then calculates the matching degree between the question and the knowledge information and compares the matching degree with a corresponding pre-set threshold. Based on the comparison results, the platform provides a preliminary answer to the student or forwards it to the teacher. After the teacher answers, the platform adds knowledge to the knowledge base.

[0007] Task Analysis Module: After students submit their research and study task results, the platform breaks down the results into dimensions. Each dimension has an evaluation indicator. The platform calculates the indicator score, the dimension comprehensive score, and the overall task score using a scoring formula. The task level is then determined based on the overall score. Learning Analysis Module: Based on students' questions, matching degree, comprehensive scores of each dimension, and task level, the module counts the frequency of questions on knowledge points to identify weak knowledge points; it collects comprehensive scores of each dimension at multiple time points, calculates the change in scores between adjacent time points, and analyzes the performance trends of each dimension, including continuous rise, continuous fall, fluctuation, and overall trend. Intelligent push module: Based on the weak knowledge points and performance trends of students' learning situation analysis, it calculates the matching degree between teaching resources and students' learning situation, and pushes teaching resources that meet the matching degree to teachers to assist in targeted teaching.

[0008] As a further aspect of the present invention, the specific implementation of the collaborative interaction module is as follows: Students enter their questions on the platform's question interface. The platform first extracts keywords from the input questions and sets the extracted student question keywords as K={k1,k2,……,k n}; where n is the number of keywords; Then, based on keyword K, the platform's pre-set knowledge base is searched to obtain knowledge information; the knowledge base stores knowledge information related to study tours, common questions, and answers. The retrieved knowledge information is grouped into a set and labeled as U={i1,i2,……,i m}, where m is the number of knowledge information retrieved; Next, the matching degree between the student's question keywords and the knowledge information in the knowledge base is calculated and denoted as S(i). j (Q) When the matching degree S(i) j When Q) ≥ T, the platform will i j This serves as preliminary feedback to the students, i.e., retrieving the corresponding solution content; Where T is the preset matching threshold; When the matching degree S(i) j If Q) < T, then the question will be forwarded to the corresponding teacher; after receiving the question pushed by the student, the teacher will directly answer the student's question and add it to the knowledge base.

[0009] As a further aspect of the present invention: when calculating the matching degree between student question keywords and knowledge information in the knowledge base, first determine the set of keywords extracted from the student question, and the set of keywords corresponding to the knowledge information. Next, find the common keywords in the two sets, that is, the intersection keywords, and add up the preset weight values ​​of each intersection keyword to obtain the numerator; Then, compare the length of the student's question keyword set with the length of the knowledge information keyword set, and take the maximum of the two as the denominator; Finally, divide the numerator by the denominator and multiply by 100% to get the degree of matching between the knowledge information and the student's question. The formula is: Where: i j It is the j-th piece of information in set U; K(i) j ) is information i j The corresponding set of keywords; w k This refers to the preset weight value corresponding to keyword k; len(K) refers to the length of the set of student question keywords K; len[K(i j [] is information i j Keyword set K(i) j ) length; molecular part Indicates student question keywords and knowledge information i j The sum of the weights of each keyword in the intersection of keywords; the denominator. This is to normalize keyword sets of different lengths, taking the maximum value between the length of the student question keyword set and the length of the knowledge information keyword set; Matching score is used to determine whether the answer can be directly fed back to the student.

[0010] As a further aspect of the present invention, the dimensional decomposition method is as follows: The decomposed set of dimensions is labeled as D={d1,d2,……,d} p}, where r = 1, 2, ..., p, and p is the number of dimensions; For each dimension d r The platform sets a corresponding set of evaluation indicators G. r ={g r1 ,g r2 ,……,g rq}, Where v = 1, 2, ..., q, and q is the number of evaluation indicators in the r-th dimension; Then, using the scoring formula: For each evaluation index g rv Scoring is performed to obtain the evaluation index g. rv Score S rv ; Where: N rv The student's task results in the evaluation indicator g rvThe actual number achieved; N rv,max This is the maximum achievable quantity for this evaluation indicator; multiplying by 100 is to present the score in a percentage format. Next, for each evaluation dimension, first obtain the score of each evaluation indicator under the corresponding dimension, as well as the preset weight value of each evaluation indicator in the corresponding dimension; then multiply the score of each evaluation indicator by its corresponding weight value, and finally sum all the results to obtain the comprehensive score of the corresponding dimension, denoted as S. r ; The overall score S for each dimension r The calculation formula is: ; Where, β rv It is the indicator g rv In dimension d r The corresponding preset weight value, ; After obtaining the overall score for each evaluation dimension, the pre-defined weight value for each dimension in the overall task evaluation is obtained; then, the overall score for each dimension is multiplied by its corresponding weight value, and all products are summed to obtain the overall task score, denoted as S. Z .

[0011] Overall score of the task S Z The calculation formula is: ; Where, α r It is dimension d r The preset weight values ​​in the overall task evaluation, ; The levels are determined based on the overall score of the task: When S Z If the score is ≥90, the task level is excellent. When 80≤S Z If the score is less than 90, the task level is considered good. When 60≤S Z If the score is less than 80, then the task level is considered satisfactory. S Z If the score is less than 60, the task level is considered unqualified.

[0012] As a further aspect of this invention: the dimensions encompass knowledge mastery, standardization of practical operations, and manifestation of innovative thinking.

[0013] As a further aspect of this invention, the method for identifying weak knowledge points is as follows: Count the number of times students input corresponding questions for each knowledge point, and record it as N. k At the same time, the total number of times the corresponding questions for all knowledge points are input is counted and denoted as N1; Divide the number of times a student enters a corresponding question for a specific knowledge point by the total number of times that corresponding question is entered for all knowledge points to obtain the frequency of student questions for each knowledge point, and denote it as F. k ; The formula is: ; Then, the frequency F of students entering corresponding questions on each knowledge point will be calculated. k Compare with a pre-set questioning frequency threshold FY: When F k If the value is greater than FY, the corresponding knowledge point will be marked as a weak point in the student's knowledge mastery; otherwise, it will not be marked.

[0014] As a further aspect of the present invention, the trend analysis method is as follows: Within a specified period, collect students' comprehensive scores (S) across various dimensions after completing the study tour tasks at multiple consecutive time points. r ; For each dimension d r The difference between the composite score of the corresponding dimension at the next time point and the composite score of the same dimension at the previous time point is the change in score of that dimension between two adjacent time points, and is denoted as ΔS. r (t); The formula is: Calculate the change in score ΔS between two adjacent time points. r (t); Where: t refers to a point in time within the specified period; ΔS r (t) refers to dimension d r The change in score between time point t and t+1; S r (t) refers to dimension d r The overall score at time point t; S r (t+1) refers to dimension d. r The overall score at time point t+1; If dimension d r ΔS r (t) remains positive, meaning that ΔS corresponds to all two adjacent time points within the specified period. r If all (t) are greater than 0, it indicates that the score in this dimension is continuously rising, and the student is showing a trend of improvement in this dimension; If dimension d r ΔS r (t) remains negative, meaning that ΔS corresponds to all two adjacent time points within the specified period. r If all (t) are less than 0, it indicates that the score in this dimension is continuously decreasing, and the student is showing a downward trend in this dimension; If dimension d rΔS r (t) Alternating positive and negative values, that is, ΔS corresponding to all two adjacent time points within a specified period. r In (t), ΔS r (t) < 0 and ΔS r If (t) > 0 is present simultaneously, it indicates that the score of this dimension fluctuates greatly. Therefore, multiple sets of data should be combined to judge the overall trend.

[0015] As a further aspect of the present invention, the overall trend judgment method is as follows: ΔS corresponding to all two adjacent time points within the specified period r Add (t) together and find the sum; When the sum is positive, the overall trend is judged to be a progressive trend; When the sum is negative, the overall trend is judged to be downward.

[0016] As a further aspect of the present invention: the teaching resources are pre-constructed; each pre-constructed teaching resource contains corresponding feature information, including the set of core knowledge points involved in the teaching resource, the difficulty level of the resource, and the applicable learning scenarios of the resource.

[0017] As a further aspect of the present invention, the method for determining the matching degree between teaching resources and students' learning situation is as follows: Determine the characteristic set of teaching resources and the characteristic set of students' learning situation; Find the interrelated features in the two sets, i.e., the intersection features. Add the preset teaching resource feature weight values ​​corresponding to each intersection feature. The sum is the matching degree between the teaching resource and the student's learning situation, which is used to determine whether to push the resource to the teacher. The calculation formula is as follows: Where R is the feature set of teaching resources, and A is the feature set of students' learning situation, which covers weak knowledge points and performance trends in various dimensions; γ e It is the preset weight value corresponding to the teaching resource feature e, which is preset according to the degree to which the resource compensates for the learning situation.

[0018] As a further aspect of the present invention: when the matching degree C(R,A)≥C0, the teaching resource is pushed to the teacher to assist the teacher in conducting targeted teaching; wherein, C0 is a preset teaching resource matching degree threshold.

[0019] (III) Beneficial Effects This invention provides an AI-based intelligent research and learning interaction platform. Compared with existing technologies, it has the following advantages: The collaborative interaction module achieves precise responses to student questions through keyword extraction and matching degree calculation. When the matching degree is sufficient, the module can quickly retrieve the answer from the knowledge base and provide feedback to the student, improving problem-solving efficiency. When the matching degree is insufficient, the module forwards the answer to the teacher and supplements the knowledge base, ensuring that students' questions receive professional answers and continuously enriching the knowledge base resources. This helps to dynamically improve the research and learning knowledge system and builds an efficient and sustainable knowledge acquisition channel for students.

[0020] The task analysis module achieves a scientific evaluation of the results of study tours through multi-dimensional breakdown and a tiered scoring mechanism. From knowledge acquisition and practical operation to innovative thinking, scores are calculated and levels are assigned based on the weights of each indicator and the overall weight. This comprehensively and accurately reflects students' study tour achievements, allowing students to clearly understand their strengths and weaknesses. It also provides teachers with objective task evaluation criteria, helping to improve the targeted nature of teaching.

[0021] The learning analysis module and the intelligent recommendation module work together to achieve accurate diagnosis of learning progress and personalized delivery of teaching resources. Learning analysis identifies weak knowledge points based on question frequency and analyzes ability metrics based on score trends, providing a basis for teaching intervention. The intelligent recommendation module calculates the matching degree between learning progress and resource characteristics to deliver targeted teaching resources to teachers, helping them implement personalized teaching, effectively improving teaching efficiency and student learning outcomes, and promoting the precision and personalization of study-based learning. Attached Figure Description

[0022] Figure 1 This is a system block diagram of an AI-based intelligent research and learning interaction platform according to the present invention. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see Figure 1 As shown, the embodiments of the present invention provide the following technical solutions: As an embodiment of the present invention: This invention is an AI-based intelligent research and learning interaction platform, comprising: The collaborative interaction module allows students to input questions on the platform's question interface. The platform first extracts keywords from the input questions and sets the extracted student question keywords as K={k1,k2,……,k n}; Where n is the number of keywords; Then, based on keyword K, the platform's pre-set knowledge base is searched to obtain knowledge information; The knowledge base stores knowledge points, frequently asked questions, and answers related to study tours. The retrieved knowledge information is grouped into a set and labeled as U={i1,i2,……,i m}, where m is the number of knowledge information retrieved; Next, through: Calculate the matching degree S(i) between student question keywords and knowledge information in the knowledge base. j (Q) in: i j It is the j-th piece of information in set U; K(i) j ) is information i j The corresponding set of keywords; w k It is a preset weight value corresponding to keyword k; len(K) refers to the length of the set K of student question keywords; len[K(i j [] is information i j Keyword set K(i) j ) length; Molecular part Indicates student question keywords and knowledge information i j The sum of the weights of each keyword in the intersection of keywords; denominator This is to normalize keyword sets of different lengths, taking the maximum value between the length of the student question keyword set and the length of the knowledge information keyword set; Multiplying by 100% is to present the match rate as a percentage. In this embodiment, the weight values ​​are preset according to the importance of the keywords in the research and learning knowledge system; When the matching degree S(i) j When Q) ≥ T, the platform will i j This serves as preliminary feedback to the students, i.e., retrieving the corresponding solution content; In this embodiment, T is a preset matching threshold, which takes the value of 80%. When the matching degree S(i) j If Q) < T, then the question will be forwarded to the corresponding teacher; after receiving the question pushed by the student, the teacher will directly answer the student's question and add it to the knowledge base.

[0025] Example 1 establishes an efficient question-and-answer bridge between students and the platform by constructing a collaborative interaction module. On one hand, the platform first extracts keywords from student questions and retrieves them from a pre-defined knowledge base. Then, using a scientific matching formula, it filters out knowledge information with a matching degree of ≥80% as preliminary answers. This allows students to quickly obtain accurate answers related to their study tours, reducing waiting time and improving the immediacy and convenience of self-directed learning. On the other hand, when the matching degree is less than 80%, the question is forwarded to the teacher, who then answers and adds the information to the knowledge base. This ensures that student questions are professionally resolved, continuously enriches the knowledge base, and dynamically optimizes the platform's functions. It provides more comprehensive knowledge support for subsequent student questions, while also reducing the burden on teachers to repeatedly answer common questions and improving teaching response efficiency.

[0026] As a second embodiment of the present invention: In its specific implementation, compared to Embodiment 1, the technical solution of this embodiment differs only in that it further includes: a task analysis module, used to break down the study tour task results by dimension after students submit them. The decomposed set of dimensions is labeled as D={d1,d2,……,d} p}, where r = 1, 2, ..., p, and p is the number of dimensions; In this embodiment, the dimensions include knowledge mastery, standardization of practical operation, and manifestation of innovative thinking; For each dimension d r The platform sets a corresponding set of evaluation indicators G. r ={g r1 ,g r2 ,……,g rq}, Where v = 1, 2, ..., q, and q is the number of evaluation indicators in the r-th dimension; Then, using the scoring formula: For each evaluation index g rv Scoring is performed to obtain the evaluation index g. rv Score S rv ; in: N rv The student's task results in the evaluation indicator g rv The actual number achieved; N rv,max This is the maximum achievable quantity for this evaluation indicator; Multiplying by 100 is to present the score as a percentage. Then, through: Calculate the overall score S for each dimension. r ; Where, β rv It is the indicator g rv In dimension d r The corresponding preset weight value; in this embodiment, it is preset according to the research and study objectives; Passed again: Calculate the overall score S of the task. Z ; Where, α r It is dimension d r Preset weight values ​​in the overall task evaluation; The levels are determined based on the overall score of the task: When S Z If the score is ≥90, the task level is excellent. When 80≤S Z If the score is less than 90, the task level is considered good. When 60≤S Z If the score is less than 80, then the task level is considered satisfactory. S Z If the score is less than 60, the task level is considered unqualified.

[0027] The newly added task analysis module in Example 2 provides a standardized and refined solution for evaluating the outcomes of study tours. This module first breaks down task outcomes into key dimensions such as knowledge mastery, practical operation standardization, and innovative thinking. Then, it sets specific evaluation indicators for each dimension and calculates the indicator score, dimension composite score, and overall task score using quantitative scoring formulas. Based on the overall score, it categorizes students into four levels: excellent, good, satisfactory, and unsatisfactory. This not only frees students' study tour outcome evaluation from the limitations of subjectivity and ambiguity, making the evaluation results more objective and fair, and helping students clearly understand their performance in each dimension, but also provides teachers with precise data references on student task completion. This allows teachers to specifically understand students' strengths and weaknesses in knowledge mastery, practical operation, and innovative abilities, providing clear direction for subsequent teaching guidance and promoting the improvement of the quality of study tour teaching.

[0028] As an embodiment of the present invention: In its specific implementation, compared to Embodiment 1 and Embodiment 2, the technical solution of this embodiment combines the solutions of Embodiment 1 and Embodiment 2. The only difference between this embodiment and Embodiment 1 and Embodiment 2 is that this embodiment also includes: a learning analysis module, which analyzes students' weak knowledge points based on their student questions, matching degree, comprehensive scores of various dimensions, and task level. The analysis method for weak knowledge points is as follows: Count the number of times students input corresponding questions for each knowledge point, and record it as N. kAt the same time, the total number of times the corresponding questions for all knowledge points are input is counted and denoted as N1; pass: Calculate the frequency F of students asking corresponding questions on each knowledge point. k ; In this embodiment, the frequency F of questions asked by students on various knowledge points is used to determine the question frequency. k This makes it easier to statistically analyze the distribution of students' weak knowledge points. For example, if the frequency of questions about a particular knowledge point is higher than the frequency of questions about other knowledge points, then that knowledge point is a weak knowledge point for the student. Then, the frequency F of students entering corresponding questions on each knowledge point will be calculated. k Compare with a pre-set questioning frequency threshold FY: When F k If the value is greater than FY, then the corresponding knowledge point will be marked as the student's weak knowledge point; otherwise, it will not be marked. The learning analysis module also determines the students' performance trends in each dimension based on their comprehensive scores in each dimension after completing the study tour tasks at multiple consecutive time points. The trend is determined as follows: Within a specified period, collect students' comprehensive scores (S) across various dimensions after completing the study tour tasks at multiple consecutive time points. r ; For each dimension d r ,pass: Calculate the change in score ΔS between two adjacent time points. r (t); Where: t refers to a point in time within the specified period; ΔS r (t) refers to dimension d r The change in score between time point t and t+1; S r (t) refers to dimension d r The overall score at time point t; S r (t+1) refers to dimension d. r The overall score at time point t+1; In this embodiment, by adjusting ΔS r The sign and magnitude of t can be used to quantitatively assess a student's progress or shortcomings in various dimensions: If dimension d r ΔS r (t) remains positive, meaning that ΔS corresponds to all two adjacent time points within the specified period. r If all (t) are greater than 0, it indicates that the score in this dimension is continuously rising, and the student is showing a trend of improvement in this dimension; For example: Suppose the score changes for dimension d1 corresponding to knowledge mastery are as follows: The fact that the change was positive and stable at two consecutive time points indicates that students have made continuous progress in the corresponding dimension of knowledge mastery. If dimension d r ΔS r (t) remains negative, meaning that ΔS corresponds to all two adjacent time points within the specified period. r If all (t) are less than 0, it indicates that the score in this dimension is continuously decreasing, and the student is showing a downward trend in this dimension; For example: Suppose that innovative thinking is reflected in the following changes in the score of the corresponding dimension d2: The fact that the change was negative and stable at two consecutive time points indicates that students' performance in the corresponding dimension of innovative thinking has declined and there are deficiencies. If dimension d r ΔS r (t) Alternating positive and negative values, that is, ΔS corresponding to all two adjacent time points within a specified period. r In (t), ΔS r (t) < 0 and ΔS r If (t) > 0 exists simultaneously, it indicates that the score of this dimension fluctuates greatly. Then, multiple sets of data should be combined to judge the overall trend. The overall trend is judged as follows: ΔS corresponding to all two adjacent time points within the specified period r Add (t) together and find the sum; When the sum is positive, the overall trend is judged to be a progressive trend; When the sum is negative, the overall trend is judged to be downward. For example: Suppose the score changes for dimension d3 corresponding to the standardization of practical operations are as follows: The total change was +7, showing an overall upward trend, indicating that although students fluctuated in this dimension, they made overall progress.

[0029] Example 3 integrates the solutions from Examples 1 and 2 and adds a learning analysis module, enabling in-depth insight and dynamic tracking of students' learning progress. In analyzing weak knowledge points, by statistically analyzing the number of times students ask questions about each knowledge point, calculating the frequency of questions, and comparing this with thresholds, the system can accurately pinpoint students' weak knowledge areas, making teaching support more targeted. In judging performance trends, by collecting dimensional scores from multiple consecutive time points, calculating the changes in scores between adjacent time points, and combining the positive or negative sign, magnitude, and sum of these changes, the system quantitatively analyzes students' progress, decline, or fluctuation trends in each dimension. This allows students to understand their learning dynamics in real time and adjust their learning strategies accordingly, while also enabling teachers to comprehensively grasp students' learning development trajectories and intervene in advance to address declining learning, providing a scientific basis for the development of personalized teaching plans and further improving the accuracy and effectiveness of learning-based teaching.

[0030] As an embodiment of the present invention: In specific implementation, compared with Embodiment 1, Embodiment 2 and Embodiment 3, the only difference between this embodiment and Embodiment 1, Embodiment 2 and Embodiment 3 is that this embodiment also includes: an intelligent push module, which pushes targeted teaching resources to teachers based on students' learning situation corresponding to weak knowledge points and performance trends in various dimensions; The teaching resources are pre-built. In this embodiment, the teaching resources are built by experts and teachers in the field of study tours before the platform is put into use. They are systematically organized and created by combining study tour objectives, knowledge system and students' cognitive patterns. Each pre-built teaching resource contains corresponding feature information, including but not limited to the set of core knowledge points involved in the teaching resource, the difficulty level of the resource, and the applicable learning scenarios of the resource; The matching degree between teaching resources and students' learning situation is calculated as follows: Where R is the feature set of teaching resources, and A is the feature set of students' learning situation, which covers weak knowledge points and performance trends in various dimensions; γ e It is the preset weight value corresponding to the teaching resource feature e, which is preset according to the degree to which the resource compensates for the learning situation; When the matching degree C(R,A)≥C0, the teaching resource is pushed to the teacher to assist the teacher in carrying out targeted teaching; Wherein, C0 is a preset teaching resource matching threshold, and in this embodiment, the value of C0 is 75%.

[0031] Example 4 adds an intelligent push module to the previous three examples, constructing a closed loop of "learning analysis - resource matching - precise push". This module, based on students' weak knowledge points and performance trends in various dimensions, uses a weighted matching formula to filter out teaching resources with a matching degree ≥75% and push them to teachers. Furthermore, these teaching resources are constructed by experts in the research and study field and a team of teachers, combining research and study objectives, knowledge systems, and students' cognitive patterns, ensuring the professionalism and applicability of the resources. This function not only solves the problem of "difficulty in finding resources" and "unsuitable resources" for teachers in targeted teaching, allowing them to quickly obtain teaching resources that match students' learning situations, saving resource selection time and improving teaching preparation efficiency; it also helps teachers conduct more targeted teaching activities, effectively addressing students' weak knowledge points and improving their performance shortcomings in various dimensions, promoting the transformation of research and study teaching from "broad-based" to "precision-based," and further improving the quality of research and study teaching and student learning outcomes.

[0032] As a fifth embodiment of the present invention: In specific implementation, compared with Embodiment 1, Embodiment 2, Embodiment 3 and Embodiment 4, the technical solution of this embodiment is to combine the solutions of Embodiment 1, Embodiment 2, Embodiment 3 and Embodiment 4.

[0033] Example 5 integrates all the solutions from Examples 1 to 4, constructing a fully functional and highly efficient intelligent learning interaction system. This system encompasses four core functions: real-time question-and-answer interaction, standardized assessment of task analysis, in-depth insights into learning progress, and precise resource support through intelligent push notifications. The modules are interconnected and share data: the collaborative interaction module provides student question data for learning progress analysis, the task analysis module provides task score data, the results of the learning progress analysis module provide matching criteria for the intelligent push notification module, and the resources from the intelligent push notification module in turn support collaborative interaction and task-based teaching. This comprehensive functional integration not only provides students with end-to-end learning support from question answering and outcome assessment to learning progress feedback, helping them learn efficiently and make continuous progress; it also provides teachers with full-scenario teaching assistance from teaching response and learning progress monitoring to resource acquisition, significantly improving teaching efficiency and accuracy; and it also makes the platform a comprehensive learning service tool integrating "learning-assessment-analysis-teaching," greatly enhancing the platform's practicality and value, and providing strong support for the digital and intelligent development of learning education.

[0034] It should be stated that all user data collected in this application was collected with the user's consent and authorization, and the use of user data is legal and compliant, and the use and processing of user data comply with the relevant laws, regulations and standards of the relevant regions.

[0035] Furthermore, any content not described in detail in this specification is existing technology known to those skilled in the art.

[0036] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0037] The above formulas are all dimensionless calculations. Dimensionless calculation involves introducing a reference benchmark, such as the maximum, minimum, standard deviation, or theoretical extreme value of a physical quantity, to transform the original physical quantity into a dimensionless relative value. This value is usually mapped to a specific interval, such as [0,1] or [-1,1], which eliminates the influence of units while preserving the relative size relationship of the physical quantities. The formula is derived from software simulation based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0038] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

[0039] The various embodiments of this application have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. An AI-based intelligent research and learning interaction platform, characterized in that, include: Collaborative Interaction Module: After a student inputs a question, the platform extracts keywords and retrieves knowledge information from a pre-set knowledge base based on the keywords. Then, it calculates the matching degree between the question and the knowledge information and compares the matching degree with the corresponding pre-set threshold. Based on the comparison results, the platform provides a preliminary answer to the student or forwards it to the teacher. After the teacher answers, the platform adds knowledge to the knowledge base. Task Analysis Module: After students submit their research and study task results, the platform breaks down the results into dimensions. Each dimension has an evaluation indicator. The platform calculates the indicator score, the dimension comprehensive score, and the overall task score using a scoring formula. The task level is then determined based on the overall score. Learning Analysis Module: Based on students' questions, matching degree, comprehensive scores of each dimension, and task level, the module counts the frequency of questions on knowledge points to identify weak knowledge points; it collects comprehensive scores of each dimension at multiple time points, calculates the change in scores between adjacent time points, and analyzes the performance trends of each dimension, including continuous rise, continuous fall, fluctuation, and overall trend. Intelligent push module: Based on the weak knowledge points and performance trends of students' learning situation analysis, it calculates the matching degree between teaching resources and students' learning situation, and pushes teaching resources that meet the matching degree to teachers to assist in targeted teaching.

2. The AI-based intelligent research and learning interaction platform according to claim 1, characterized in that: The specific methods for the collaborative interaction module are as follows: Students enter their questions on the platform's question interface. The platform first extracts keywords from the input questions and sets the extracted student question keywords as K={k1,k2,……,k n }; where n is the number of keywords; Then, based on keyword K, the platform's pre-set knowledge base is searched to obtain knowledge information; the knowledge base stores knowledge information related to study tours, common questions, and answers. The retrieved knowledge information is grouped into a set and labeled as U={i1,i2,……,i m }, where m is the number of knowledge information retrieved; Next, the matching degree between the student's question keywords and the knowledge information in the knowledge base is calculated and denoted as S(i). j (Q) When the matching degree S(i) j When Q) ≥ T, the platform will i j This serves as preliminary feedback to the student; where T is the preset matching threshold. When the matching degree S(i) j If Q) < T, then the question will be forwarded to the corresponding teacher. After receiving the question pushed by the student, the teacher will directly answer the student's question and add it to the knowledge base.

3. The AI-based intelligent research and learning interaction platform according to claim 2, characterized in that: When calculating the matching degree between student question keywords and knowledge information in the knowledge base, first determine the set of keywords extracted from the student question and the set of keywords corresponding to the knowledge information. Next, find the common keywords in the two sets, that is, the intersection keywords, and add up the preset weight values ​​of each intersection keyword to obtain the numerator; Then, compare the length of the student's question keyword set with the length of the knowledge information keyword set, and take the maximum of the two as the denominator; Finally, divide the numerator by the denominator and multiply by 100% to get the matching degree between the knowledge information and the student's question.

4. The AI-based intelligent research and learning interaction platform according to claim 1, characterized in that: The dimensional decomposition method is as follows: The decomposed set of dimensions is labeled as D={d1,d2,……,d} p }, where r = 1, 2, ..., p, and p is the number of dimensions; For each dimension d r The platform sets a corresponding set of evaluation indicators G. r ={g r1 ,g r2 ,……,g rq }, Where v = 1, 2, ..., q, and q is the number of evaluation indicators in the r-th dimension; Then, for each evaluation indicator of the study tour task results, first count the number of times the students actually achieved the corresponding evaluation indicator, and the maximum number of times the corresponding evaluation indicator can be achieved. Then divide the number of times the students actually achieved the indicator by the maximum number of times the indicator can be achieved, and then multiply by 100 to obtain the score value of the corresponding evaluation indicator. Next, for each evaluation dimension, first obtain the score of each evaluation indicator under the corresponding dimension, as well as the preset weight value of each evaluation indicator in the corresponding dimension; then multiply the score of each evaluation indicator by its corresponding weight value, and finally sum all the results to obtain the comprehensive score of the corresponding dimension, denoted as S. r ; After obtaining the overall score for each evaluation dimension, the pre-defined weight value for each dimension in the overall task evaluation is obtained; then, the overall score for each dimension is multiplied by its corresponding weight value, and all products are summed to obtain the overall task score, denoted as S. Z .

5. The AI-based intelligent research and learning interaction platform according to claim 1, characterized in that: The levels are determined based on the overall score of the task: When S Z If the score is ≥90, the task level is excellent. When 80≤S Z If the score is less than 90, the task level is considered good. When 60≤S Z If the score is less than 80, then the task level is considered satisfactory. S Z If the score is less than 60, the task level is considered unqualified.

6. The AI-based intelligent research and learning interaction platform according to claim 1, characterized in that: The methods for identifying weak knowledge points are as follows: Count the number of times students entered corresponding questions for each knowledge point, and also count the total number of times they entered corresponding questions for all knowledge points; Divide the number of times a student enters a corresponding question for a specific knowledge point by the total number of times that corresponding question is entered for all knowledge points to obtain the frequency of student questions for each knowledge point, and denote it as F. k ; Then, the frequency F of students entering corresponding questions on each knowledge point will be calculated. k Compare with a pre-set questioning frequency threshold FY: When F k If the value is greater than FY, the corresponding knowledge point will be marked as a weak point in the student's knowledge mastery; otherwise, it will not be marked.

7. The AI-based intelligent research and learning interaction platform according to claim 1, characterized in that: The performance trend analysis method is as follows: Within a specified period, collect students' comprehensive scores (S) across various dimensions after completing the study tour tasks at multiple consecutive time points. r ; For each dimension d r The difference between the composite score of the corresponding dimension at the next time point and the composite score of the same dimension at the previous time point is the change in score of that dimension between two adjacent time points, and is denoted as ΔS. r (t); If dimension d r ΔS r If (t) remains positive, then the student shows a trend of improvement in this dimension; If dimension d r ΔS r If (t) remains negative, then the student's performance in this dimension shows a downward trend. If dimension d r ΔS r (t) If positive and negative values ​​alternate, then the overall trend can be determined.

8. The AI-based intelligent research and learning interaction platform according to claim 7, characterized in that: The overall trend is judged as follows: ΔS corresponding to all two adjacent time points within the specified period r Add (t) together and find the sum; When the sum is positive, the overall trend is judged to be a progressive trend; When the sum is negative, the overall trend is judged to be downward.

9. The AI-based intelligent research and learning interaction platform according to claim 1, characterized in that: The method for determining the matching degree between teaching resources and students' learning situation is as follows: Determine the characteristic set of teaching resources and the characteristic set of students' learning situation; Find the interrelated features in the two sets, i.e., the intersection features. Add the preset teaching resource feature weight values ​​corresponding to each intersection feature, and the sum is the matching degree between teaching resources and students' learning situation.

10. The AI-based intelligent research and learning interaction platform according to claim 9, characterized in that: When the matching degree C(R,A)≥C0, the teaching resource is pushed to the teacher; where C0 is the preset teaching resource matching degree threshold.

Images