Classroom teaching quality evaluation index system and use method thereof

Abstract

The invention discloses a classroom teaching quality evaluation index system and a use method thereof, and the method comprises the steps: building classroom teaching quality evaluation indexes based on subject literatures, and collecting classroom teaching practice data according to the classroom teaching quality evaluation indexes, constructing a classroom teaching quality evaluation index system according to the subject literature and classroom teaching practice data; the method comprises the following steps: dividing a classroom teaching quality evaluation index system into a plurality of dimension indexes according to an evaluation standard, determining the weight of each dimension index and calculating the weight of a classroom teaching evaluation index by using an analytic hierarchy process, carrying out quantitative analysis on the quality of classroom words, and constructing a comprehensive evaluation model based on the index system; and calculating a weighted difference degree by adopting score calculation of a weighted Euclidean distance model, and carrying out quantitative comparison on the multi-dimensional classroom behavior data and an expert reference classroom to obtain a teaching quality evaluation result. And a foundation is laid for intelligent evaluation of classroom teaching. And a technical basis is provided for development of large-scale classroom teaching evaluation and improvement work.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent education, and more particularly relates to a classroom teaching quality evaluation index system and a method for using the same. Background Art

[0002] The related research on teacher classroom teaching evaluation originated early and there are relatively mature theories and frameworks. As early as the 1970s, Sinclair and Coulthard proposed the IRF classroom discourse interaction model through the analysis of classroom language, including initiation, response, and feedback. This model has been widely used by researchers and has gradually formed various structural variants, such as the no-response structure I-R, the multi-response structure I-R1-R2…Rn-F, the multi-feedback structure I-R1-F1…Rn-Fn, and IRE (initiation-response-evaluation), etc. Another widely influential classroom teaching evaluation framework is the Flanders Interaction Analysis System (FIAS) proposed by the American scholar Flanders. Flanders believes that the most important information in the classroom is language, and classifies classroom language into teacher language, student language, and silence, where teacher language includes expressing emotions, encouraging and praising, adopting opinions, asking questions, lecturing, giving instructions, and criticizing and safeguarding rights. Through the review of classroom teaching evaluation frameworks, it is found that most of the existing classroom teaching evaluation frameworks are adaptations of the Flanders framework.

[0003] Problems and disadvantages of the prior art: 1. The existing classroom teaching evaluation systems have various problems, such as a large number of types, inconsistent naming methods and classification criteria for indicators, and difficult-to-interoperate analysis results. The coexistence of different evaluation indicators with different naming methods and classification criteria makes it difficult to effectively compare and communicate evaluation results. This weakens the evaluation system and limits the role of teaching evaluation in promoting the improvement of teaching quality. 2. The existing analysis frameworks pay more attention to the forms of classroom teaching behaviors, such as frequency, duration, sequence, etc., and ignore the connection between teaching behaviors and students' cognitive development and emotional experience, resulting in evaluation results that are difficult to comprehensively reflect the actual effects of classroom teaching. 3. The existing classroom teaching evaluation systems cannot give quantitative evaluation results based on teaching ability. First, the existing classroom teaching evaluation systems usually record classroom teaching behaviors, and the output evaluation reports are mostly trivial behavior frequencies, but the evaluation is not intuitive and does not directly point to the evaluation of teaching ability. Second, the existing classroom teaching evaluation systems judge whether the teaching of the evaluated teacher needs to be optimized by comparing the evaluated teacher with the existing sample means in the database or the pre-set reference values, but neither the sample means nor the set reference values are scientific quantifications of high-quality classroom teaching. The evaluation results are mainly qualitative descriptions, lacking specific quantitative data support and being difficult to compare and promote among different classrooms, schools, and even regions. 4. Most of the existing classroom teaching evaluation systems are "one-size-fits-all", that is, one system can be applied to all disciplines, ignoring the differences in the requirements for teachers' abilities by different subject contents. It cannot reflect the requirements for teachers' abilities by the teaching subject content (such as mathematics), cannot reflect subject differences, and cannot give the requirements for teachers' abilities by the mathematics subject content, making it impossible for teachers of different disciplines to obtain highly targeted teaching improvement suggestions. 5. The existing classroom discourse evaluation process is highly subjective and inefficient. The evaluators rely on personal experience and intuition to make judgments, lacking objective quantitative criteria, resulting in low consistency and reliability of evaluation results. Therefore, there is an urgent need to construct a general classroom teaching quality evaluation index system to achieve the mutual recognition and sharing of evaluation results. Summary of the Invention

[0004] The present invention aims to overcome the defects of the prior art. In view of the above problems, the present invention provides a classroom teaching quality evaluation index system and its usage method. Based on curriculum standards in the evaluation framework, corresponding indicators for discourse subjects, discourse forms, and discourse functions are designed. Through a standardized process and by combining existing evaluation quantitative indicators, a quantitative evaluation model is constructed to convert the quality of classroom discourse into quantifiable data, so as to improve the reliability and validity of the evaluation, achieve the mutual recognition and sharing of evaluation results, reduce subjective bias, improve the efficiency and accuracy of the evaluation, and have the characteristics of objectivity.

[0005] To solve the above technical problems, the present invention provides the following technical solutions: A classroom teaching quality evaluation index system, the dimensions of which include the discourse subject, discourse form, and discourse function. The discourse subject includes teachers and students; the discourse form includes the first-level discourse form indicators and corresponding second-level discourse form indicators; the discourse function includes the first-level discourse function indicators and corresponding second-level discourse function indicators; the first-level discourse form indicators include statements, questions, responses, feedback, and management; the first-level discourse function indicators include knowledge understanding, personal expression, analysis and argumentation, comparison and induction, and transfer and innovation.

[0006] The present invention also provides a method for using classroom teaching quality evaluation indicators, including: constructing classroom teaching quality evaluation indicators based on preset subject literature, then collecting classroom teaching practice data according to the classroom teaching quality evaluation indicators, and constructing a classroom teaching quality evaluation index system based on the subject literature and classroom teaching practice data; Dividing the classroom teaching quality evaluation index system into multiple dimension indicators according to the evaluation criteria, using the analytic hierarchy process to determine the weights of each dimension indicator and calculate the weights of the classroom teaching evaluation indicators, quantitatively analyzing the quality of classroom discourse, and constructing a comprehensive evaluation model based on the index system; Adopting the scoring calculation of the weighted Euclidean distance model, calculating the weighted difference degree, and quantitatively comparing the multi-dimensional classroom behavior data with the expert benchmark classroom to obtain the teaching quality evaluation result.

[0007] Optionally, the preset subject literature is the evaluation indicators in the teaching evaluation model.

[0008] Optionally, the classroom teaching quality evaluation indicators refer to extracting classroom teaching quality evaluation indicators from the text and existing evaluation systems using the measurement research method according to the teaching curriculum standards; demonstrating the classroom teaching quality evaluation indicators by an expert team and conducting reliability and validity tests; then encoding the classroom teaching according to the classroom teaching quality evaluation indicators, constructing a quantitative comprehensive evaluation model, converting the quality of classroom discourse into quantifiable data, and defining the operability of each indicator.

[0009] Optionally, converting the quality of classroom discourse into quantifiable data means that after encoding the classroom discourse according to the classroom teaching quality evaluation indicators, presenting the data in different dimensions according to the frequency of each evaluation indicator and the time sequence between evaluation indicators, and calculating the weighted distance according to the frequency of the classroom teaching quality evaluation indicators to obtain the overall score of the teaching quality of a class.

[0010] Optionally, the use of the analytic hierarchy process to determine the weights of indicators in each dimension and to calculate the weights of classroom teaching evaluation indicators includes: constructing a judgment matrix on the classroom teaching quality evaluation indicator system, and combining goal-oriented analysis and expert team argumentation to conduct expert empowerment and consistency practice tests, and using evaluation software to determine the weights of indicators in each dimension of the analytic hierarchy process based on the expert evaluation results; Using evaluation software to determine the weights of indicators in each dimension of the hierarchical analysis method includes: using the evaluation software to first calculate the weights of the three first-level indicators of discourse subject, discourse form and discourse function, and then calculate the weights of each second-level indicator separately.

[0011] Optionally, the method of using the analytic hierarchy process to determine the weight of each dimension indicator and calculate the weight of the classroom teaching evaluation indicator includes: constructing a hierarchical structure model; Designing the questionnaire and collecting quantifiable data include: based on the characteristics of mathematics classroom teaching, evaluating the relative importance of each dimension and sub-dimension, and determining the scoring criteria of experts for each indicator; designing the questionnaire based on the pairwise comparison of the nine scales of the hierarchical analysis method, and constructing the judgment matrix of the preset dimensions based on the collected questionnaire comparison results; The consistency check and indicator weight calculation in the judgment matrix. The weight calculation of each dimension indicator includes: inputting the group decision of each expert in the judgment matrix of each dimension into the expert matrix in the group decision panel of the evaluation software. The system automatically generates a judgment matrix for each expert weight result based on the algorithm, and generates the weight result of each dimension indicator through the consistency check. The weight result of each dimension indicator is generated, which includes: calculating the maximum eigenvalue of the judgment matrix and its corresponding eigenvector, normalizing the eigenvector, and calculating the dimensional weight to obtain the weight vector; The screening and correction of judgment matrix data include: according to the order of judgment matrix, query the average random consistency index to calculate the consistency ratio; when the consistency ratio is less than the set value, the consistency of the judgment matrix is ​​acceptable, otherwise the expert weight data is screened and corrected: abnormal scale items are automatically fed back and re-evaluation is required; expert weight data that seriously deviates from the group consistency threshold is eliminated; In the evaluation software, the geometric mean calculation results of the expert weights that have passed the consistency test are aggregated to generate global weights; the weights of each dimension and sub-dimension of the evaluation index are calculated and embedded in the comprehensive evaluation model.

[0012] Optionally, constructing a hierarchical model includes: based on the constructed classroom teaching quality evaluation index system, using the hierarchical analysis method to construct a three-level hierarchical model in the evaluation environment of the evaluation software, decomposing the decision-making problem in the teaching evaluation problem into a target layer, primary indicators and secondary indicators, and presetting the logical relationship between the evaluation indicators layer by layer.

[0013] Optionally, the scoring calculation using the weighted Euclidean distance model is adopted to calculate the weighted difference degree, and the multi-dimensional classroom behavior data is quantitatively compared with the expert benchmark classroom, and the teaching quality evaluation results include: Teaching data collection and processing: Through the multi-modal data collection device arranged in the environment, high-definition teaching audio and video are generated, and after the video recording and voice are collected, they are connected to the video library to realize the all-round multi-modal data collection of teacher teaching; Machine training is carried out on the teaching evaluation model, and according to the trained teaching evaluation model, the audio and video of the all-round multi-modal data of teacher teaching uploaded are automatically transcribed, analyzed and encoded to generate the index frequency vector of the classroom to be evaluated; Dynamically match the reference set according to the index frequency statistical results of the classroom to be evaluated; Construct a benchmark database: Collect the audio and video data of multiple expert demonstration classrooms, extract teaching behavior indicators and generate standardized frequency vectors, and store them in the benchmark database; Generate a scoring standard reference vector: Obtain the standard reference vector for scoring of the same type of classroom from the collected expert vectors; Dynamically match the reference set: According to the subject type, teaching stage and student age level of the classroom to be evaluated, screen the set of expert classrooms of the same type from the benchmark database; Normalized scoring mapping: Based on historical data and empirical values, obtain the maximum possible distance, output the final score, and generate an evaluation analysis report.

[0014] Optionally, before the normalized scoring mapping, the weighted difference degree is calculated, and the calculation of the weighted difference degree includes: calculating the difference degree between the classroom to be evaluated and each expert benchmark classroom by using the weighted Euclidean distance model.

[0015] Compared with the prior art, the present invention has the following beneficial effects: The mathematical classroom discourse evaluation method of the present invention, based on the index system, combines the Analytic Hierarchy Process (AHP), and through constructing a judgment matrix and consistency test, can obtain the weights of each dimension and index according to the judgment of experts, making the evaluation results more objective and scientific. Through the classroom teaching quality evaluation index method, the quality of classroom discourse can be quantitatively analyzed, and a standardized and quantitative tool in the mathematical classroom evaluation system can be constructed. Description of the drawings

[0016] Figure 1 It is the hierarchical diagram of the classroom teaching quality evaluation index system of the present invention.

[0017] Figure 2 It is the framework diagram of the usage method of the classroom teaching quality evaluation index of the present invention.

[0018] Figure 3 It is a schematic diagram for constructing an index system of the usage method of the classroom teaching quality evaluation index of the present invention.

[0019] Figure 4 It is a schematic diagram for determining the index weights of the usage method of the classroom teaching quality evaluation index of the present invention.

[0020] Figure 5 It is a schematic diagram for quantifying the teaching quality of the usage method of the classroom teaching quality evaluation index of the present invention.

[0021] Figure 6 It is a flowchart of the usage method of the classroom teaching quality evaluation index of the present invention. Detailed implementation manners

[0022] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0023] As Figure 3 shown, the present invention constructs a mathematics classroom quality evaluation index system with discourse subject, discourse form and discourse function. It can be used as an important basis for mathematics classroom quality evaluation and improvement, is the foundation for implementing large-scale classroom teaching monitoring and evaluation, and provides data support for improving the quality of education and teaching. Three principles of universality, scientificity and operability are adhered to in the construction process. First, taking the mathematics curriculum standard as the core basis, adopting a measurement research method, extracting general and core indicators from complex texts and existing evaluation systems to ensure that the evaluation system has good universality. Second, this evaluation system has been demonstrated by an expert team and has passed reliability and validity tests through a large number of practices to ensure its scientificity. Finally, more than 100 mathematics classes have been coded based on this evaluation system, and the operational definitions of each indicator have been continuously improved in actual operation to ensure that the coding categories are clear, specific, the wording is clear, and the semantics are unambiguous.

[0024] Example 1 As Figure 2 、 Figure 6 shown, a usage method of a classroom teaching quality evaluation index includes: S10. Based on preset subject literature, construct a classroom teaching quality evaluation index, then collect classroom teaching practice data according to the classroom teaching quality evaluation index, and construct a classroom teaching quality evaluation index system based on the subject literature and classroom teaching practice data; S20. Divide the classroom teaching quality evaluation index system into multiple index dimensions according to the evaluation criteria, use the analytic hierarchy process to determine the weights of each dimension index and calculate the weights of the classroom teaching evaluation index, quantitatively analyze the quality of classroom discourse, and construct a comprehensive evaluation model based on the index system; S30. Calculate the score using the weighted Euclidean distance model, calculate the weighted difference degree, quantitatively compare the multi-dimensional classroom behavior data with the expert benchmark classroom, and obtain the teaching quality evaluation result.

[0025] Optionally, the preset subject literature is an evaluation index in the teaching evaluation model.

[0026] Optionally, the classroom teaching quality evaluation index refers to extracting the classroom teaching quality evaluation index from the text and the existing evaluation system by using the measurement research method according to the teaching curriculum standard; demonstrating the classroom teaching quality evaluation index by an expert team and conducting reliability and validity tests; then encoding the classroom teaching according to the classroom teaching quality evaluation index to construct a quantitative comprehensive evaluation model, converting the quality of classroom discourse into quantifiable data, and continuously improving the definition of the operability of each index in actual operation. The subject literature refers to the literature related to classroom video analysis and classroom teaching interaction evaluation. The text refers to the subject literature.

[0027] Optionally, the conversion of the quality of classroom discourse into quantifiable data means that after encoding the classroom discourse according to the classroom teaching quality evaluation index, presenting the data in a dimension-by-dimension manner based on the frequency of each evaluation index and the time sequence between the evaluation indexes, calculating the weighted distance according to the frequency of the classroom teaching quality evaluation index, and obtaining the overall score of the teaching quality of a class. Among them, the quantifiable data is the number of times, time, and order of each evaluation index appearing in the classroom. By comparing with the corresponding data of the expert classroom (specifically, see the part of weighted distance calculation), the teaching quality of this class is scored.

[0028] Optionally, as Figure 1 shown, a classroom teaching quality evaluation index system, the dimensions of the classroom teaching quality evaluation index system include the discourse subject, discourse form, and discourse function. The discourse subject includes teachers and students; the discourse form includes the first-level discourse form index and the corresponding second-level discourse form index; the discourse function includes the first-level discourse function index and the corresponding second-level discourse function index; the first-level discourse form index includes statement, question, response, feedback, and management; the first-level discourse function index includes knowledge understanding, personal expression, analysis and demonstration, comparison and induction, and transfer and innovation.

[0029] Optionally, using the analytic hierarchy process to determine the weights of each dimension index and calculate the weights of the classroom teaching evaluation index includes: constructing a judgment matrix on the classroom teaching quality evaluation index system, and conducting expert empowerment and consistency practice tests in combination with goal-oriented analysis and expert team demonstration. According to the expert judgment results, use evaluation software to determine the weights of each dimension index of the analytic hierarchy process. This expert empowerment is a process in which experts make pairwise judgment comparisons on each first-level index and second-level index of the dimension. Here, it refers to filling out a questionnaire.

[0030] Using evaluation software to determine the weights of indicators in each dimension of the hierarchical analysis method includes: using the evaluation software to first calculate the weights of the three first-level indicators of discourse subject, discourse form and discourse function, and then calculate the weights of each second-level indicator separately.

[0031] Optional, such as Figure 4 As shown, the use of the hierarchical analysis method to determine the weights of indicators in each dimension and calculate the weights of classroom teaching evaluation indicators includes constructing a hierarchical model, wherein the construction of the hierarchical model includes: based on the constructed classroom teaching quality evaluation indicator system, using the hierarchical analysis method to construct a three-level hierarchical model in the evaluation environment of the evaluation software, the decision-making problem in the teaching evaluation problem is set to be decomposed into a target layer, a first-level indicator and a second-level indicator, and the logical relationship between the evaluation indicators is preset layer by layer; the target layer is the mathematics classroom teaching evaluation, the first-level indicator includes discourse form and discourse function, and each first-level indicator is divided into a number of second-level indicators.

[0032] Designing questionnaires and collecting quantifiable data include: based on the characteristics of mathematics classroom teaching, evaluating the relative importance of each dimension and sub-dimension, and determining the expert scoring standards for each indicator; designing a pairwise comparison questionnaire based on the nine-scale analytic hierarchy process, and constructing a judgment matrix of preset dimensions based on the collected questionnaire comparison results.

[0033] The consistency check and indicator weight calculation in the judgment matrix, the weight calculation of each dimension indicator includes: inputting the group decision of each expert in the judgment matrix of each dimension into the expert matrix in the group decision panel of the evaluation software, and the system automatically generates a judgment matrix for each expert weight result based on the algorithm for consistency check, and generates the weight results of each dimension indicator through consistency check. The weight results of each dimension indicator include: calculating the maximum eigenvalue of the judgment matrix and its corresponding eigenvector, normalizing the eigenvector, and calculating the dimensional weight to obtain the weight vector. The weight vector is the weight of each dimensional indicator.

[0034] The screening and correction of judgment matrix data include: according to the order of judgment matrix, query the average random consistency index to calculate the consistency ratio; when the consistency ratio is less than the set value, the consistency of the judgment matrix is ​​acceptable, otherwise the expert weight data is screened and corrected: abnormal scaling items are automatically fed back and re-evaluation is required; expert weight data that seriously deviates from the group consistency threshold are eliminated; the set value is 0.1.

[0035] In the evaluation software, select the geometric mean calculation result aggregation of the expert weights that pass the consistency test to generate the global weight; calculate the weights of each dimension and sub-dimension of the evaluation indicators and embed them into the comprehensive evaluation model. Among them, the weights of all indicators constitute the global weight. The expert weight refers to generating a judgment matrix for each expert weight result for consistency test and weight calculation, and calculating the obtained weight vector. Take the average value of the weight results as the final weight assignment.

[0036] Optionally, build a three-level hierarchical model in the evaluation environment of the evaluation software, including front-end data collection and back-end AI analysis, to realize the full-process automation of collection - analysis - feedback.

[0037] Optionally, as Figure 5 shown, use the weighted Euclidean distance model to calculate the score, calculate the weighted difference degree, and quantitatively compare the multi-dimensional classroom behavior data with the expert benchmark classroom to obtain the teaching quality evaluation results, including: Teaching data collection and processing: Through the multi-modal data collection device arranged in the environment, generate high-definition teaching audio and video, connect the video and voice after collection to the video library to realize the all-round multi-modal data collection of teacher teaching; perform machine training on the teaching evaluation model, and automatically transcribe, analyze and encode the audio and video of the all-round multi-modal data of teacher teaching uploaded according to the trained teaching evaluation model to generate the index frequency vector of the classroom to be evaluated; Dynamically match the reference set according to the index frequency statistical results of the classroom to be evaluated; Build a benchmark database: Collect the audio and video data of multiple expert demonstration classrooms, extract teaching behavior indicators and generate standardized frequency vectors, and store them in the benchmark database; Generate a scoring standard reference vector: Obtain the standard reference vector for scoring the same type of classroom from the collected expert vectors; Dynamically match the reference set: According to the subject type, teaching stage and student age level of the classroom to be evaluated, screen the set of expert classrooms of the same type from the benchmark database; Normalized scoring mapping: Obtain the maximum possible distance based on historical data and empirical values, output the final score, ensure that the final score does not exceed the expected range, and generate an evaluation analysis report.

[0038] Optionally, calculate the weighted difference degree before the normalized scoring mapping, and calculate the weighted difference degree: Use the weighted Euclidean distance model to calculate the difference degree between the classroom to be evaluated and each expert benchmark classroom.

[0039] The specific usage method is as follows: As Figure 1As shown in the figure, for the classroom teaching quality evaluation index system, the standardization of classroom teaching quality evaluation index means collecting classroom teaching videos as input -> transcribing them into text -> encoding the text according to the index system -> outputting statistical charts (frequency, frequency, time series) separately according to dimensions -> obtaining the overall score. The following are the specific contents: 1. Discourse Subject The discourse subject is divided into two first-level index categories: teachers and students. Distinguish the contributions of teachers and students to classroom discourse.

[0040] (1) Teachers. The classroom discourse is teacher discourse initiated by teachers, including teachers' knowledge explanations, question raising, and giving students evaluation feedback, etc. (2) Students. The classroom discourse is student discourse initiated by students, including students' answering teachers' questions, students' question raising, etc. Describe the participation degree of students in the classroom.

[0041] 2. Discourse Forms include the first-level index of discourse forms and the corresponding second-level indexes of discourse forms. The first-level index of discourse forms includes questioning, responding, feedback, statement, and management. Based on the questioning - responding - feedback (IRF) structure, the first-level indexes of statement and management are added to ensure the full coverage and in-depth analysis of the coding for data.

[0042] Each discourse form is divided. Among them, the feedback shows the non-development (statement), horizontal (expansion) development, vertical (deepening) development, and loop (evaluation) development of the behavior chain. The discourse functions of the expansion and deepening categories are conducive to cultivating students' cooperation, critical thinking, and reflection abilities. Such discourses guide students to listen to others' speeches, collaborate with others, and supplement and deepen existing viewpoints from multiple angles such as horizontally and vertically.

[0043] (1) Statement: There is no second-level index.

[0044] Teachers or students give a detailed explanation and description of a certain mathematical concept, principle, or method.

[0045] (2) Questioning includes: second-level indexes of group-oriented and individual-oriented.

[0046] Group-oriented: Teachers or students initiate questions to a group or the whole class to stimulate group discussion, collect information, or promote learning.

[0047] Individual-oriented: Teachers or students initiate questions to an individual to stimulate personal potential or promote in-depth dynamic interaction.

[0048] (3) Responding includes: second-level indexes of group response and individual response.

[0049] Group response: A certain group (such as a group or the whole class) jointly answers a question.

[0050] Individual response: A teacher or student individually answers a question.

[0051] (4) Feedback includes the secondary indicators of restatement, deepening, expansion, and evaluation.

[0052] Restatement: A teacher or student repeats the words of others to emphasize or confirm the accuracy of information.

[0053] Deepening: Based on the previous content, through follow-up questions to the response, it guides deeper and more comprehensive thinking.

[0054] Expansion: Based on the previous content, different objects are invited to respond to promote diverse and multi-angle discussions.

[0055] Evaluation: A teacher or student criticizes, questions, accepts, or encourages a certain idea or view.

[0056] (5) Management includes the secondary indicators of issuing instructions and learning guidance. Issuing instructions: A teacher or student organizes classroom activities or maintains classroom order through relevant instructions to create a good learning environment.

[0057] Learning guidance: A teacher or student provides students with guidance on learning directions, methods, and strategies to help students learn effectively.

[0058] 3. Discourse functions include the first-level discourse function indicators and the corresponding second-level discourse function indicators. The first-level discourse function indicators include knowledge understanding, personal expression, analysis and argumentation, comparison and induction, and transfer and innovation. The first-level indicators are guided by the development of thinking, evolving from low-level thinking discourse types that focus on knowledge acquisition to discourse types that focus on the cultivation of core competencies, emotional attitudes, and values, and then to discourse types that focus on the cultivation of high-level thinking such as critical questioning and innovation and creation.

[0059] (1) Knowledge understanding Discourses of knowledge understanding include mathematical concepts and symbols, relationships and operations, mathematical historical materials, common sense knowledge, and other knowledge. According to the characteristics of a mathematics classroom, this knowledge understanding includes the secondary indicators of common sense knowledge, mathematics subject knowledge, mathematics history knowledge, and interdisciplinary knowledge. Common sense knowledge refers to the generally recognized knowledge acquired in life; mathematics subject knowledge includes basic concepts in mathematics, mathematical symbolic language, mathematical geometric figures, mathematical theorems, mathematical formulas, mathematical relationships or expressions, and the laws and mathematical operations reflected by mathematical knowledge; mathematics history knowledge is the knowledge that explores the origin, evolution, and its relationship with culture, science, and philosophy of mathematical content in different civilizations and historical stages. Interdisciplinary knowledge refers to the intersection of mathematics and other disciplines, such as the intersection of mathematics and kinematics knowledge in physics. This indicator refers to knowledge that can refer to textbooks or existing knowledge, can be judged as right or wrong, and can be answered or questioned only by memorization.

[0060] The discourse on basic knowledge is an important part of classroom discourse. The discourse on common sense knowledge enables students to experience the wide application of mathematics in life, which helps to enhance students' interest in learning mathematics. The discourse on mathematical subject knowledge is the basic task of a mathematics classroom, laying a solid foundation of subject knowledge for students. The discourse on the history of mathematics reveals mathematical concepts and thinking methods, allowing students to experience the thinking process of "mathematicians" and accumulate corresponding mathematical activity experience. The discourse on interdisciplinary knowledge is conducive to students' comprehensive application of knowledge from various disciplines to solve practical problems.

[0061] (2) Personal expression Personal expression discourse includes students' personal experiences, direct activity experiences, curiosity, thirst for knowledge, interests, imagination, discovery and raising of questions, etc. This first-level indicator of personal expression includes second-level indicators such as experience, subjective views, personal imagination, and emotional attitudes. Experience includes students' personal experiences, mathematical activity experiences, etc. Subjective views include the ideas and opinions expressed by students on a certain issue. Personal imagination includes ideas that do not actually exist and the raising of mathematical questions through imagination. Emotional attitudes include curiosity, interest in things, or the thirst for knowledge when facing new problems, difficult problems, and doubts.

[0062] This type of discourse is conducive to cultivating students' learning interests, curiosity, imagination, and establishing correct values of emotional attitudes. Discourse that is close to students' direct activity experiences helps students recognize the connection between mathematics and life and stimulates their learning interests.

[0063] (3) Analysis and argumentation Analysis and argumentation discourse includes splitting and extracting information, logical reasoning, analyzing and solving problems, giving examples to explain, arguing, evaluating, etc. This first-level indicator of analysis and argumentation includes second-level indicators such as examining and breaking through the problem, giving examples to illustrate, and explaining and arguing. Examining and breaking through the problem includes teachers' guidance or students' independent splitting and extracting of information from the question. Giving examples to illustrate includes understanding complex and abstract knowledge by giving specific examples and helping students understand the key points and difficulties of knowledge through organizing or counterexamples. Explaining and arguing conducts logical reasoning on the question or problem and solves the problem through clear explanations and arguments.

[0064] In this type of discourse, teachers guide students to deeply analyze problems and find breakthroughs in solving problems, which is conducive to enhancing the ability to analyze and solve problems. At the same time, in the process of argumentation and evaluation, students' logical reasoning ability is improved. Through giving examples to explain, students are cultivated to transform mathematical knowledge into specific and easily understandable knowledge and express it in their own language.

[0065] (4) Comparison and induction Comparative induction discourse includes generalizing and summarizing, finding patterns, establishing models, seeking logical connections, etc. It grasps the knowledge structure as a whole from the intuitive to the abstract. This type of discourse is conducive to cultivating students' ability to view problems comprehensively and integratively, enhancing their overall thinking, and discovering the laws of things through comparison and connection. This first-level indicator of comparative induction includes second-level indicators of comparative analysis, inductive summary, and mathematical modeling. This type of discourse is beneficial to cultivating students' inductive and modeling abilities. In this type of discourse, teachers use discourse and questions to guide students to summarize the essence of problems from complex knowledge, and find the basic methods and laws to solve problems.

[0066] (5)Transfer and innovation Transfer and innovation discourse refers to using existing knowledge and information to explore the unknown, infer and solve problems, guess conclusions, and predict the development direction of things based on evidence. On the basis of existing knowledge, transfer mathematical knowledge, skills, and even mathematical thinking methods, and put forward ideas, thoughts, or viewpoints different from the conventional or others, cultivating students' awareness and ability of innovation and creation. This first-level indicator of transfer and innovation includes second-level indicators of inference and conjecture, transfer and application, and innovation and creation. This type of discourse is conducive to cultivating students' transfer and innovation abilities.

[0067] Determine the weights of the Analytic Hierarchy Process (AHP), including using the Analytic Hierarchy Process (AHP), through expert weighting and consistency testing, mathematical modeling, and using the established evaluation software (yaahp) to determine the weights of each dimension and index, so as to achieve the scientific and objective evaluation results. Ensure the rationality of the weights of each dimension and index. It can quantitatively analyze the quality of classroom discourse and construct a standardized and quantified tool in the classroom teaching quality evaluation system. Among them, yaahp is a software assisting in the Analytic Hierarchy Process, which provides help in aspects such as model construction, calculation, and analysis for the decision-making process using the Analytic Hierarchy Process. The group decision-making panel of Yaahp provides group decision-making support, manages the information of experts participating in the decision-making and the decision-making data provided. Group decision-making is the overall process in which multiple experts jointly participate in decision-making analysis and make decisions. The participants in the decision-making process are all experts who participate in filling out the questionnaire, and the final decision result is determined according to the data provided by all experts.

[0068] (1)Construct a hierarchical structure model. According to the established classroom teaching quality evaluation index system, use the Analytic Hierarchy Process (AHP) to construct a three-level hierarchical model in the evaluation environment of the evaluation software. Decompose the teaching evaluation problem into the goal layer, first-level indicators, and second-level indicators, and preset the logical relationships between the evaluation indicators layer by layer. The goal layer is defined as "Mathematics Classroom Teaching Evaluation", which is the ultimate goal of the evaluation. The two major dimensions of the classroom teaching quality evaluation index system include discourse form and discourse function. The first-level indicators include five dimensions of discourse form and five dimensions of discourse function. The first-level indicator of discourse form includes statement, question, response, feedback, and management; the five dimensions of the first-level indicator of discourse function include knowledge understanding, personal expression, analysis and argumentation, comparison and induction, and transfer and innovation. To achieve the evaluation of mathematics classroom quality. The classroom teaching quality evaluation index system is the mathematics classroom teaching quality evaluation index system.

[0069] (2)Design a questionnaire and collect data Based on the characteristics of mathematics classroom teaching, evaluate the relative importance of each dimension and sub-dimension, and determine the scoring criteria for experts on each indicator; design a pairwise comparison questionnaire based on the nine-scale of the Analytic Hierarchy Process. According to the comparison results of the collected questionnaires, construct a judgment matrix for the preset dimensions. The questionnaire includes the comparison of the relative importance of the first-level index criterion layer (discourse subject, discourse form, and discourse function dimensions), and the comparison between the second-level indicators (such as question vs feedback). The designed questionnaire is as follows:

[0070]

[0071]

[0072]

[0073] When filling out the questionnaire, it is necessary to evaluate the relative importance of each dimension and sub-dimension based on the characteristics of mathematics classroom teaching, and determine the scoring criteria for experts on each indicator according to the nine-scale method of the Analytic Hierarchy Process. As shown in Table 1, the meaning of each scale of the judgment matrix, where a ij represents the relative importance degree of factor i and j.

[0074] Table 1

[0075] This judgment matrix means that any system analysis is based on certain information, and different-dimensional judgment matrices are constructed according to the comparison results of the questionnaire. The judgment matrix of the Analytic Hierarchy Process gives judgments on the relative importance of each factor at each level, and the judgments are represented by numerical values and made into a matrix form result. Mathematics classroom teaching evaluation A, discourse subject B1, discourse form B2, discourse function B3.

[0076] For example, the questionnaire results of the first-level indicators filled in by A are as follows:

[0077] Then, the corresponding judgment matrix A is obtained:

[0078] The questionnaire results of the comparison of the discourse subject dimension of the second-level indicators are as follows:

[0079] The corresponding judgment matrix B1 is obtained:

[0080] The questionnaire results of the comparison of the discourse form dimension of the second-level indicators are as follows:

[0081] The corresponding judgment matrix B2 is obtained:

[0082] The questionnaire results of the comparison of the discourse function dimension of the second-level indicators are as follows:

[0083] The corresponding judgment matrix B3 is obtained:

[0084] (3) Consistency test and weight calculation in the judgment matrix Weight calculation: Input the judgment matrix of each expert in each preset dimension into the group decision-making panel of the evaluation software. The system automatically conducts a consistency test based on the following algorithm. If the consistency test passes, the weight results of each dimension index will be generated: Calculate the maximum eigenvalue of the judgment matrix , and its corresponding eigenvector . Normalize the eigenvector to obtain the weight vector .

[0085] The consistency test in the judgment matrix refers to evaluating the logical consistency of the decision maker when constructing the judgment matrix in the analytic hierarchy process. Using the evaluation software yaahp, when inputting the judgment matrix data, the consistency ratio of the judgment matrix is displayed in real time according to the change of the judgment matrix data. Moreover, for inconsistent judgment matrices, the element that has the greatest impact on the consistency is also displayed in real time, which is convenient for users to master the situation and make adjustments.

[0086]

[0087]

[0088]

[0089]

[0090] Consistency test and judgment expert matrix data screening and correction: According to the order of the judgment matrix, query the average random consistency index RI (Table 2) to calculate the consistency ratio . When , it is considered that the consistency of the judgment matrix is acceptable; otherwise, screen and correct the data: automatically feedback abnormal scaling items and require re-evaluation; eliminate expert data that seriously deviates from the group consistency threshold to ensure the logical rationality of the weights. The consistency test of the evaluation software is calculated through the consistency index . Among them, is the maximum eigenvalue of the judgment matrix.

[0091] In the evaluation software, select the calculation result aggregation of the geometric mean of the weights of experts who pass the consistency test to generate the global weight set , satisfying and . The calculated weights of each dimension and sub-dimension will be embedded into the comprehensive evaluation model for quantitative analysis of classroom dialogue. Selecting the calculation result aggregation - geometric mean is to perform geometric mean and then normalization on the sorted weights Wi of each expert for this judgment matrix to obtain the final global weight.

[0092] Table 2

[0093] 2. Adopt the scoring calculation of the weighted Euclidean distance model to realize the quantitative comparison and scoring of multi-dimensional classroom behavior data and the expert benchmark classroom.

[0094] Audio and video data collection and processing: Through multi-modal data collection devices such as teacher tracking cameras, student tracking cameras, switcher systems, and omnidirectional microphones arranged in the environment, generate high-definition teaching videos (encoded with H264), audio (AAC encoded), and after collecting the video and voice, connect them to the video library to realize the all-round multi-modal information collection of teacher teaching.

[0095] Machine training is carried out on the teaching evaluation model (AI large model related to education). The materials required for the training of the AI large model include curriculum standards and multiple versions of mathematics textbooks. Based on the content and requirements of the curriculum standards and textbooks and the mathematics classroom teaching quality evaluation index system, AI coding annotation is carried out; according to the trained teaching evaluation model, the uploaded audio and video are automatically transcribed, analyzed, and encoded to generate the index frequency vector of the classroom to be evaluated ; The curriculum standard is the mathematics curriculum standard.

[0096] Dynamically match the reference set according to the index frequency statistical results of the classroom to be evaluated; Construct a benchmark database: Collect the audio and video data of multiple expert demonstration classes, extract teaching behavior indicators and generate standardized frequency vectors , and store them in the benchmark database; Generate a reference vector for the scoring standard: Obtain the reference vector for the scoring of the same type of classroom from the collected expert vectors .

[0097] Dynamically match the reference set: The system filters the set of expert classes of the same type from the benchmark database according to the subject type, teaching stage, and student age level of the classroom to be evaluated; Calculate the weighted difference degree: Use the weighted Euclidean distance: Calculate the difference degree between the classroom to be evaluated and each expert benchmark classroom, where is the reference vector value of the i-th index, is the corresponding index value of the classroom to be evaluated, and is the preset index weight coefficient.

[0098] Normalized scoring mapping: Obtain the maximum possible distance based on historical data and empirical values , ensuring that the final score does not exceed the expected range. Through Output the final score, and the score range is [0, 100].

[0099] Generate an evaluation and analysis report: Output relevant statistics and present visual charts, display each index and the gap from the expert classroom, give feedback and suggestions by combining AI evaluation and manual evaluation, and finally output the analysis and evaluation report of the classroom discourse.

[0100] The present invention takes a large number of normal mathematics classes as research samples, and deeply analyzes and compares the classroom discourse characteristics by using frequency analysis, sequence data mining, and sequence frequent pattern mining technologies. It provides a theoretical basis for classroom discourse evaluation and improvement. In activities such as teacher teaching research and teaching competitions, on the one hand, this evaluation index system provides guidance for pre-class preparation, and on the other hand, it provides a data-supported and visual analysis report for teachers' classroom teaching, which is conducive to helping teachers clarify the improvement direction and formulate targeted improvement strategies, so as to achieve the integration of teaching-learning-evaluation. This evaluation system lays a foundation for the future realization of intelligent evaluation of classroom teaching. Through the manual coding and machine training of tens of thousands of discourses, the automatic annotation of some indicators has been initially realized, and in the future, the effect of machine annotation will be improved by increasing the quantity and quality of the training set. It can provide a technical basis for the development of large-scale classroom teaching evaluation and improvement work. The classroom teaching quality evaluation index system is shown in Table 4.

[0101] Table 4

[0102] In the description of this specification, the descriptions with reference to terms such as "in one embodiment", "in another embodiment", "exemplary" or "in a specific embodiment" mean that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0103] Although the present invention has been described in detail above with general descriptions, specific embodiments and experiments, modifications or improvements can be made on the basis of the present invention, which are obvious to those skilled in the art. Therefore, these modifications or improvements made without departing from the spirit of the present invention all fall within the scope of protection required by the present invention.

Claims

1. A classroom teaching quality evaluation index system, the dimensions of the classroom teaching quality evaluation index system include the discourse subject, discourse form, and discourse function, and it is characterized in that: The discourse subjects include teachers and students; the discourse forms include first-level discourse form indicators and corresponding second-level discourse form indicators; the discourse functions include first-level discourse function indicators and corresponding second-level discourse function indicators; the first-level discourse form indicators include statements, questions, responses, feedback, and management; the first-level discourse function indicators include knowledge understanding, personal expression, analysis and argumentation, comparison and induction, and transfer and innovation.

2. A method for using the evaluation index of classroom teaching quality, characterized in that: including: Based on the preset subject literature, construct classroom teaching quality evaluation indicators, then collect classroom teaching practice data according to the classroom teaching quality evaluation indicators, and construct a classroom teaching quality evaluation index system based on the subject literature and classroom teaching practice data; Divide the classroom teaching quality evaluation index system into multiple dimension indicators according to the evaluation criteria, use the analytic hierarchy process to determine the weights of each dimension indicator and calculate the weights of the classroom teaching evaluation indicators, conduct quantitative analysis on the quality of classroom discourse, and construct a comprehensive evaluation model based on the index system; Adopt the scoring calculation of the weighted Euclidean distance model, calculate the weighted difference degree, quantitatively compare the multi-dimensional classroom behavior data with the expert benchmark classroom, and obtain the teaching quality evaluation result.

3. The usage method of the classroom teaching quality evaluation index according to claim 2, characterized in that, The preset subject literature is the evaluation indicators in the teaching evaluation model.

4. The usage method of the classroom teaching quality evaluation index according to claim 2, wherein The classroom teaching quality evaluation indicators refer to extracting classroom teaching quality evaluation indicators from the text and the existing evaluation system by using the measurement research method based on the teaching curriculum standards; demonstrating the classroom teaching quality evaluation indicators by an expert team and conducting reliability and validity tests; Then, code the classroom teaching according to the classroom teaching quality evaluation indicators, construct a quantitative comprehensive evaluation model, transform the quality of classroom discourse into quantifiable data, and define the operability of each indicator.

5. The method for using the classroom teaching quality evaluation index according to claim 4, characterized in that The transformation of the quality of classroom discourse into quantifiable data means that after coding the classroom discourse according to the classroom teaching quality evaluation indicators, the data is presented dimensionally based on the frequency of each evaluation indicator and the time sequence between the evaluation indicators, and the weighted distance is calculated according to the frequency of the classroom teaching quality evaluation indicators to obtain the overall score of the teaching quality of a class.

6. The method for using the classroom teaching quality evaluation index according to claim 2, characterized in that Using the analytic hierarchy process to determine the weights of each dimension indicator and calculate the weights of the classroom teaching evaluation indicators includes: constructing a judgment matrix on the classroom teaching quality evaluation index system, and conducting expert empowerment and consistency practice tests in combination with goal-oriented analysis and expert team demonstration. According to the expert judgment results, use the evaluation software to determine the weights of each dimension indicator of the analytic hierarchy process; Using the evaluation software to determine the weights of each dimension indicator of the analytic hierarchy process includes: using the evaluation software to first calculate the weights of the three first-level indicators of discourse subject, discourse form, and discourse function, and then calculate the weights of each second-level indicator respectively.

7. The method for using the classroom teaching quality evaluation index according to claim 2, characterized in that, The use of the analytic hierarchy process to determine the weights of each dimension indicator and calculate the weights of the classroom teaching evaluation indicators includes: constructing a hierarchical structure model; Designing a questionnaire and collecting quantifiable data includes: evaluating the relative importance of each dimension and sub-dimension based on the characteristics of mathematics classroom teaching, and determining the scoring criteria for experts on each indicator; designing a questionnaire based on the pairwise comparison of the nine scales of the analytic hierarchy process, and constructing a judgment matrix of the preset dimension according to the comparison results of the collected questionnaires; The consistency check and indicator weight calculation in the judgment matrix. The weight calculation of each dimension indicator includes: inputting the group decision of each expert in the judgment matrix of each dimension into the expert matrix in the group decision panel of the evaluation software. The system automatically generates a judgment matrix for each expert weight result based on the algorithm, and generates the weight result of each dimension indicator through the consistency check. The weight result of each dimension indicator is generated, which includes: calculating the maximum eigenvalue of the judgment matrix and its corresponding eigenvector, normalizing the eigenvector, and calculating the dimensional weight to obtain the weight vector; The screening and correction of judgment matrix data include: according to the order of judgment matrix, query the average random consistency index to calculate the consistency ratio; when the consistency ratio is less than the set value, the consistency of the judgment matrix is ​​acceptable, otherwise the expert weight data is screened and corrected: abnormal scale items are automatically fed back and re-evaluation is required; expert weight data that seriously deviates from the group consistency threshold is eliminated; In the evaluation software, the geometric mean calculation results of the expert weights that have passed the consistency test are aggregated to generate global weights; the weights of each dimension and sub-dimension of the evaluation index are calculated and embedded in the comprehensive evaluation model.

8. The method for using the classroom teaching quality evaluation index according to claim 7, characterized in that, The construction of the hierarchical model includes: based on the constructed classroom teaching quality evaluation index system, using the hierarchical analysis method to construct a three-level hierarchical model in the evaluation environment of the evaluation software, decomposing the decision-making problem in the teaching evaluation problem into a target layer, a first-level index and a second-level index, and presetting the logical relationship between the evaluation indicators layer by layer.

9. The method for using the classroom teaching quality evaluation index according to claim 2, wherein The weighted Euclidean distance model is used to calculate the score, calculate the weighted difference, and quantitatively compare the multi-dimensional classroom behavior data with the expert benchmark classroom. The teaching quality evaluation results include: Teaching data collection and processing: Generate high-definition teaching audio and video through the multimodal data collection devices arranged in the environment, collect the video and voice and connect them to the video library to realize the comprehensive multimodal data collection of teachers' teaching; Machine training is performed on the teaching evaluation model. The uploaded audio and video of the teacher's comprehensive multimodal teaching data is automatically transcribed, analyzed and encoded based on the trained teaching evaluation model to generate the indicator frequency vector of the classroom to be evaluated. Dynamically match the reference set based on the frequency statistics of the indicators in the classroom to be evaluated; Build a benchmark database: collect audio and video data from multiple expert demonstration classes, extract teaching behavior indicators and generate standardized frequency vectors, which are stored in the benchmark database; Generate a scoring standard reference vector: Get the standard reference vector for the same type of classroom scoring from the collected expert vectors; Dynamic matching reference set: According to the subject type, teaching stage and student age level of the class to be evaluated, a set of expert classes of the same type is selected from the benchmark database; Normalized score mapping: Get the maximum possible distance based on historical data and experience, output the final score, and generate an evaluation analysis report.

10. The method for using the classroom teaching quality evaluation index according to claim 9, characterized in that, The weighted difference calculation is performed before the normalized score mapping, and the weighted difference calculation includes: using a weighted Euclidean distance model to calculate the difference between the class to be evaluated and each expert benchmark class.

Images