Academic level examination tutoring method and system

Abstract

The invention relates to the technical field of computer-aided teaching, and discloses an academic level examination tutoring method based on multi-dimensional evaluation. The method comprises the following steps: constructing a test paper composition material library containing horizontal test knowledge point information, test question information and test paper template information; based on the test paper composition material library, according to a test question screening rule containing multi-dimensional test question recommendation degree evaluation indexes, automatically extracting test questions to form test paper composition test paper; based on the test paper quality evaluation index, performing quality evaluation on the test paper, and determining whether to incorporate a review scheme according to an evaluation result; and summarizing the group test papers passing the quality evaluation to form a review scheme until the target knowledge point range is covered. Through a multi-dimensional interaction evaluation mechanism and a dynamic iteration optimization strategy, the problems of repeated knowledge points, low review efficiency and the like in the traditional question sea tactics are effectively solved, and the pertinence and effectiveness of examination tutoring are remarkably improved.

Description

Technical Field

[0001] The present application relates to the field of computer-assisted teaching technology, and in particular to an academic proficiency test tutoring technology based on multi-dimensional interactive evaluation. Background Art

[0002] With the rapid development of educational informatization, computer-assisted teaching has been widely used in the field of academic proficiency test tutoring. In various academic proficiency test scenarios such as CET-4, CET-6, and IELTS, test scores are the only criterion for measuring the level of candidates, and effective test tutoring methods are the key to improving test scores. At present, candidates usually adopt the "sea of ​​questions" tactic for test review, that is, to improve their test-taking ability by practicing a large number of test questions repeatedly.

[0003] However, this traditional review method has many technical problems: first, due to the lack of systematic analysis and planning of the distribution of knowledge points, candidates often repeat exercises on certain knowledge points, but insufficiently cover other knowledge points, resulting in a large degree of blindness in the review content; secondly, the difficulty and discrimination of the test questions are difficult to match the actual level of the candidates, and the test questions are not targeted, which reduces the review efficiency; in addition, in the process of formulating the review plan, due to the lack of an effective test paper quality assessment mechanism, the overall quality of the selected test paper cannot be guaranteed, affecting the final review effect.

[0004] In order to solve the above technical problems, there is an urgent need for an intelligent test tutoring method that can achieve accurate coverage of knowledge points, reasonable screening of test questions, and controllable test paper quality. Summary of the invention

[0005] The purpose of this application is to provide a method and system for tutoring for academic proficiency tests to solve the problems raised in the above-mentioned background technology.

[0006] The present application discloses a method for tutoring for academic proficiency tests, comprising the following steps:

[0007] Steps for constructing a test paper material library: constructing a target test paper material library containing level test knowledge point information, test question information and test paper template information;

[0008] Intelligent test paper composition step: based on the target test paper composition material library, according to the preset test paper screening rules, test questions are automatically extracted from the test question information to form a test paper composition; the test question screening rules include multiple dimensions of test question recommendation evaluation indicators and corresponding thresholds;

[0009] Test paper quality evaluation step: based on the preset test paper quality evaluation index, the test paper formed by the intelligent test paper generation step is evaluated for quality, and according to the evaluation result, it is determined whether the test paper is included in the review plan;

[0010] Review plan generation step: Summarize the test paper quality assessment step to determine the test papers to be included, and form a review plan until the review plan covers the scope of target knowledge points.

[0011] In a preferred example, the level test knowledge point information includes: knowledge points and corresponding level test stages, mastery requirements, and difficulty level information; the test question information includes: test questions and corresponding question types, test knowledge points, difficulty, and discrimination information; the test paper template information includes: question types, number of questions, and test question score information.

[0012] In a preferred example, the multiple dimensions of test question recommendation evaluation indicators include: content recommendation, difficulty recommendation and discrimination recommendation.

[0013] In a preferred example, the content recommendation degree F1 is calculated according to the following formula:

[0014]

[0015] Among them, Min(a,b) is the total number of test knowledge points contained in the test question, and Max(a,b) is the total number of knowledge points contained in the test knowledge point domain; X i , X j , X k are the assessment weights of the i-th, j-th, and k-th knowledge points in the test questions respectively; Y i , Y j , Y k are the requirements for mastering the i-th, j-th, and k-th knowledge points in the test knowledge point domain; Z i , Z j , Z k are the cognitive levels of the test groups on the ith, jth and kth knowledge points respectively; Match is the matching calculation function; α1, α2, α3, δ1, δ2, δ3, ε1, ε2 and ε3 are weight coefficients.

[0016] In a preferred example, the difficulty recommendation F2 is calculated according to the following formula:

[0017]

[0018]

[0019] Among them, R is the difficulty coefficient of the test question, R′ is the difficulty requirement of the question type to which the test question belongs; R i , R j , R k are the difficulty levels of the i-th, j-th, and k-th knowledge points in the test questions respectively; Match is the matching calculation function; β1, β2, β3, θ1, θ2, θ3, is the weight coefficient.

[0020] In a preferred embodiment, the discrimination recommendation degree F3 is calculated according to the following formula:

[0021]

[0022] Among them, S is the discrimination coefficient of the test question, S′ is the discrimination requirement of the question type to which the test question belongs; Z 1i , Z 1j , Z 1k are the first examinee’s cognitive levels of the i, j, and k knowledge points, respectively; H1 is the first examinee’s test score; Z qi , Z qj , Z qk are the cognitive levels of the qth examinee on the i-th, j-th, and k-th knowledge points, respectively, and H q is the test score of the qth examinee; Match is the matching calculation function; γ1, γ2, ω1, ω2, ..., ω m is the weight coefficient.

[0023] In a preferred embodiment, the final recommendation degree F of the test questions in the intelligent test paper composition step is calculated according to the following formula:

[0024] F=F1·L1+F2·L2+F3·L3

[0025] Among them, F1, F2, and F3 are content recommendation, difficulty recommendation, and discrimination recommendation, respectively; L1, L2, and L3 are weight coefficients of each recommendation, and L1+L2+L3=1.

[0026] In a preferred example, the test paper quality assessment indicators include at least two of the following: the ratio of the knowledge point spectrum quotient value of the test paper to the test target knowledge point spectrum quotient value G1; the proportion of the number of test target test knowledge points covered by the test paper G2; and the average value G3 of the ratio of the maximum difficulty value of the question type used by the knowledge point in the test paper to the corresponding maximum difficulty value in the question bank.

[0027] In a preferred embodiment, G1, G2, and G3 are calculated according to the following formula:

[0028] G1=(∑eu / ∑fv)

[0029] G2=(m / n)

[0030] G3=(∑max w / Max w ) / m

[0031] Among them, f and v are the frequency of knowledge points in the test knowledge domain and the test score, e and u are the frequency of knowledge points in the test paper and the test score, n is the total number of knowledge points in the test knowledge domain, n is the total number of knowledge points involved in the test paper, max wis the maximum difficulty value of the question type used in the test paper for the wth knowledge point, Max w is the maximum difficulty value of the question type used by the wth knowledge point in the question bank;

[0032] In addition, the test paper quality assessment result G in the test paper quality assessment step is calculated according to the following formula:

[0033]

[0034] in, is the weight coefficient.

[0035] The present application also discloses a system for tutoring academic proficiency tests, comprising:

[0036] A test paper material library construction module is used to construct a target test paper material library containing level test knowledge point information, test question information and test paper template information;

[0037] An intelligent test paper composition module is used to automatically extract test questions from the test question information to form a test paper composition based on the target test paper composition material library and according to preset test question screening rules; the test question screening rules include multiple dimensions of test question recommendation evaluation indicators and corresponding thresholds;

[0038] The test paper quality evaluation module is used to evaluate the quality of the test paper formed by the intelligent test paper forming module based on the preset test paper quality evaluation index, and determine whether the test paper is included in the review plan according to the evaluation result;

[0039] The review plan generation module is used to summarize the test papers determined to be included by the test paper quality assessment module and form a review plan until the review plan covers the scope of target knowledge points.

[0040] The implementation methods of this application have the following technical effects:

[0041] By building a test material library covering knowledge points, test questions, and test paper templates, and storing and associating various types of information in a structured manner, the source of test questions has been effectively expanded, the adaptability and flexibility of test questions to exam review needs have been improved, and a reliable data foundation has been provided for subsequent intelligent test paper generation.

[0042] A parallel question recommendation and evaluation strategy is adopted based on the three dimensions of content recommendation, difficulty recommendation and discrimination recommendation. Through nonlinear weight allocation and adaptive matching operations among the dimensions, multiple factors such as knowledge point relevance, mastery degree matching, and discrimination ability are comprehensively considered, thereby improving the pertinence and effectiveness of the test papers and overcoming the problems of incomplete knowledge point coverage and inaccurate control of test difficulty in the traditional test paper composition method.

[0043] During the intelligent test paper compilation process, the knowledge points that have been included in the test paper are eliminated through the dynamic update mechanism of the real-time knowledge point domain, which ensures the mutual exclusivity and independence of the knowledge points in the test paper, effectively avoids the redundancy of review content caused by repeated compilation of knowledge points, and significantly improves the efficiency of test review.

[0044] A test paper quality assessment mechanism based on multiple indicators such as knowledge point spectrum quotient value ratio, knowledge point coverage, and knowledge point question type difficulty matching has been introduced. By objectively quantifying the test paper quality and combining it with preset assessment thresholds, dynamic screening of test papers is achieved, ensuring the superiority of the test papers included in the review plan in terms of comprehensive knowledge point coverage and reasonable difficulty distribution.

[0045] In the process of generating review plans, a dynamic iterative optimization strategy based on evaluation indicators was adopted. Through multiple rounds of test paper generation and feedback adjustment of quality assessment, the review plans were systematically screened and improved. The final review plan not only comprehensively covers the knowledge points, but also takes into account the systematicity and scientificity, which is of positive significance to improving the overall learning effect of exam review.

[0046] The method of the present invention has strong technical adaptability and promotion value through a configurable evaluation index system and a flexible weight adjustment mechanism. It can not only be used for English tests such as CET-4 and CET-6 and IELTS, but can also be promoted to other subject areas such as computer grade examinations and accounting qualification examinations through parameter adjustment. It has exemplary significance for promoting the intelligent and personalized development of examination review models.

[0047] A large number of technical features are recorded in the specification of this application, which are distributed in various technical solutions. If all possible combinations of technical features of this application (i.e., technical solutions) are to be listed, the specification will be too long. In order to avoid this problem, the various technical features disclosed in the above-mentioned invention content of this application, the various technical features disclosed in the various embodiments and examples below, and the various technical features disclosed in the accompanying drawings can be freely combined with each other to form various new technical solutions (these technical solutions are all deemed to have been recorded in this specification), unless the combination of such technical features is technically infeasible. For example, in one example, feature A+B+C is disclosed, and in another example, feature A+B+D+E is disclosed, and features C and D are equivalent technical means that play the same role. Technically, only one can be used, and it is impossible to use them at the same time. Feature E can be combined with feature C technically. Then, the solution of A+B+C+D should not be deemed to have been recorded because it is technically infeasible, and the solution of A+B+C+E should be deemed to have been recorded. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 and Figure 2It is a flowchart of the academic proficiency test tutoring method according to the first embodiment of the present application.

[0049] Figure 3 It is a flow chart of the evaluation of the test question recommendation degree in the academic proficiency test tutoring method according to the first embodiment of the present application, showing the entire process of calculating the final recommendation degree of the test question.

[0050] Figure 4 This is the structural intention of the academic proficiency test tutoring system according to the second embodiment of the present application. DETAILED DESCRIPTION

[0051] In the following description, many technical details are provided to help readers better understand the present application. However, those skilled in the art can understand that the technical solution claimed in the present application can be implemented even without these technical details and various changes and modifications based on the following embodiments.

[0052] Description of some concepts:

[0053] Academic proficiency test: refers to a standardized test conducted to test the learning effect and academic achievement of candidates, usually conducted at the end of a specific learning stage, such as CET-4 and CET-6, IELTS, and accounting level test.

[0054] Test material library: refers to a data set that contains knowledge point information, test questions and test paper template information pre-built for intelligent test composition. Among them, the knowledge point information of the level test includes the knowledge point and the corresponding level test stage, mastery requirements, difficulty level and other attributes; the test question information includes the test question and the corresponding question type, test knowledge point, difficulty, discrimination and other attributes; the test paper template information includes the question type, number of questions, test question score and other attributes.

[0055] Test knowledge point domain: refers to the collection of all knowledge points covering the scope of the academic proficiency test, representing the entire domain of knowledge points for test review.

[0056] Content recommendation degree: refers to the degree of match between the knowledge point composition of the test questions and the test knowledge point domain, reflecting the relevance and pertinence of the test questions content to the test review needs.

[0057] Recommended difficulty: refers to the degree of match between the difficulty of the test questions and the difficulty requirements of the examination objectives, reflecting the suitability of the difficulty of the test questions for the level of the examinees.

[0058] Recommended degree of discrimination: refers to the degree of match between the discriminating ability of the test questions and the requirements of the question type, reflecting the effectiveness of the test questions in identifying candidates of different levels.

[0059] Question recommendation threshold: refers to the lower limit of recommendation preset when screening questions, which is used to determine whether the questions should be included in the test paper.

[0060] Knowledge point spectrum quotient value ratio: refers to the ratio of the sum of the quotient values ​​of the knowledge point spectrum covered by the compiled test paper to the sum of the quotient values ​​of the target knowledge point spectrum in the examination target knowledge point spectrum, reflecting the degree to which the compiled test paper meets the requirements for mastering the examination target knowledge points.

[0061] Knowledge point coverage rate: refers to the ratio of the number of knowledge points contained in the test paper to the total number of knowledge points in the test knowledge point domain, reflecting the comprehensiveness of the test paper's coverage of the examination knowledge points.

[0062] Knowledge point question type difficulty matching degree: refers to the matching degree between the maximum difficulty value of the test questions appearing for each knowledge point in the test paper and the maximum difficulty value of the test questions corresponding to the knowledge point in the question bank, reflecting the accuracy of the test paper in controlling the difficulty of the test questions when it is refined to the knowledge point granularity.

[0063] Test paper quality assessment threshold: refers to the preset lower limit of test paper quality when screening test papers, which is used to determine whether the test papers should be included in the review plan.

[0064] The following is a brief description of some of the innovative features of this application:

[0065] In view of the technical problems mentioned above, the inventor of this application has proposed a method for tutoring academic level examinations after extensive and in-depth research and analysis. By constructing a multi-dimensional interactive evaluation mechanism based on dynamic mapping of knowledge point domain, the contradiction between knowledge point repetitiveness and learning efficiency in existing academic level examination tutoring is specifically solved. This method innovatively performs adaptive matching operations on the evaluation indicators of the three dimensions of test content, difficulty and discrimination with the cognitive level and knowledge mastery requirements of the examinees, and realizes the synergy between the evaluation indicators through nonlinear weight distribution; at the same time, based on the ratio relationship between the knowledge point spectrum quotient value and the matching degree evaluation of the question type difficulty distribution, a test paper quality evaluation system with self-regulation ability is established, and the dynamic optimization and iteration of the review plan is realized. This intelligent test paper composition method based on multi-dimensional interaction uses complex mathematical models and evaluation algorithms to achieve the optimization of test question selection while ensuring the coverage of knowledge points, breaking through the technical bottleneck of mechanical repetition and low efficiency in the traditional "sea of ​​questions tactics".

[0066] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below in conjunction with the accompanying drawings.

[0067] The first embodiment of the present application relates to a method for tutoring for academic proficiency tests, the process of which is as follows: Figure 1 , Figure 2 and Figure 3 As shown, the method comprises the following steps:

[0068] Step 100: constructing a test paper material library, constructing a target test paper material library including level test knowledge point information, test question information and test paper template information;

[0069] Step 200: an intelligent test paper composition step, based on the target test paper composition material library, automatically extracting test questions from the test question information to form a test paper composition according to preset test question screening rules; the test question screening rules include multiple dimensions of test question recommendation evaluation indicators and corresponding thresholds;

[0070] Step 300: a test paper quality assessment step, based on a preset test paper quality assessment index, a quality assessment is performed on the test paper formed by the intelligent test paper forming step, and according to the assessment result, it is determined whether the test paper is included in the review plan;

[0071] Step 400: Review plan generation step, summarizing the test papers to be included in the test paper quality assessment step, and forming a review plan until the review plan covers the scope of target knowledge points.

[0072] Optionally, the level test knowledge point information includes: knowledge points and corresponding level test stages, mastery requirements, and difficulty level information; the test question information includes: test questions and corresponding question types, test knowledge points, difficulty, and discrimination information; the test paper template information includes: question types, number of questions, and test question score information.

[0073] Optionally, the multiple dimensions of test question recommendation evaluation indicators include: content recommendation, difficulty recommendation and discrimination recommendation.

[0074] More specifically, the core idea of ​​the academic proficiency test tutoring method of this embodiment is to realize the intelligence and optimization of the test tutoring process through four steps: building a test material library, intelligent test paper composition, test paper quality assessment and review plan generation.

[0075] First, this method constructs a test paper library containing knowledge point information, test question information and test paper template information. Among them, knowledge point information includes knowledge points and their corresponding test stage, mastery requirements and difficulty level; test question information includes test question content and its corresponding question type, test knowledge points, difficulty and discrimination; test paper template information includes question type, question quantity and score distribution. Through the structured storage and associative mapping of these three types of key information, the test paper library provides rich and targeted data support for subsequent intelligent test paper.

[0076] In the stage of intelligent test paper composition, this method automatically extracts test questions based on the test paper composition material library according to the preset test question screening rules to form a test paper. Its innovation lies in the introduction of multiple dimensions of test question recommendation evaluation indicators, including content recommendation, difficulty recommendation and discrimination recommendation. The content recommendation measures the degree of match between the test question content and the target knowledge points, the difficulty recommendation measures the degree of adaptation between the test question difficulty and the level of the examinees, and the discrimination recommendation measures the ability of the test question to distinguish between examinees of different levels. By comprehensively considering the recommendation of these three dimensions and setting the corresponding thresholds, the comprehensive coverage of knowledge points, the rationality of the difficulty gradient and the effectiveness of the discrimination ability of the test questions can be taken into account, thereby improving the overall quality of the test paper composition.

[0077] In the test paper quality assessment stage, this method evaluates the quality of the test paper based on the preset evaluation indicators, and decides whether to include the test paper in the review plan based on the evaluation results. This mechanism can filter out test papers that do not meet the quality standards by quantifying the quality of the test paper and setting the inclusion criteria, thus ensuring the scientificity and effectiveness of the review plan.

[0078] Finally, in the review plan generation stage, the method summarizes the test papers that have passed the quality assessment to form a review plan, and uses the knowledge point coverage as the termination condition to control the iterative process. This progressive and full-coverage plan generation strategy can make full use of the results of the test paper quality assessment and ultimately output a set of review plans with complete content and reasonable structure.

[0079] In general, this application builds a full-process intelligent test tutoring method starting from multiple links such as test material, test question screening, quality assessment and program generation. Its key innovations lie in the construction of the test material library, multi-dimensional test question recommendation and evaluation, closed-loop test paper quality control and progressive review program generation, which has positive significance in improving the quality and efficiency of test tutoring and optimizing the review experience of candidates.

[0080] Optionally, the content recommendation degree F1 is calculated according to the following formula:

[0081]

[0082] Among them, Min(a,b) is the total number of test knowledge points contained in the test question, and Max(a,b) is the total number of knowledge points contained in the test knowledge point domain; X i , X j , X k are the assessment weights of the i-th, j-th, and k-th knowledge points in the test questions respectively; Y i , Y j , Y k are the requirements for mastering the i-th, j-th, and k-th knowledge points in the test knowledge point domain; Z i ,

[0083] Zj , Z k are the cognitive levels of the test groups on the ith, jth and kth knowledge points respectively; Match is the matching calculation function; α1, α2, α3, δ1, δ2, δ3, ε1, ε2 and ε3 are weight coefficients.

[0084] More specifically, this embodiment introduces the evaluation index of content recommendation degree in the intelligent test paper stage, the purpose of which is to measure the degree of match between the test content and the target knowledge points. The calculation formula of content recommendation degree comprehensively considers the following factors:

[0085] The ratio of the number of test knowledge points contained in the test question to the number of knowledge points in the entire test knowledge point domain. This ratio reflects the breadth of coverage of the test questions on knowledge points. The higher the coverage, the higher the content recommendation of the test question.

[0086] The matching degree between the assessment weight of each knowledge point in the test question and its mastery requirement in the test knowledge point domain. The assessment weight reflects the proportion of the knowledge point in the test question, and the mastery requirement reflects the importance of the knowledge point in the test objective. The higher the matching degree between the two, the more consistent the test question's focus on the knowledge point is with the test requirements, and the higher the content recommendation of the test question. Among them, the Match function is used to quantify the matching degree between the two.

[0087] The matching degree between the assessment weight of each knowledge point in the test and the cognitive level of the test group on the knowledge point. The cognitive level reflects the actual mastery of the knowledge point by the examinee. The higher the matching degree with the assessment weight, the more targeted the test content is, and the higher the recommendation degree is. Similarly, the Match function is used to quantify the matching degree.

[0088] The weight coefficients α1, α2, and α3 are used to adjust the influence of the above three aspects in the recommendation calculation. The sum of the three coefficients is 1, and they can be adjusted according to actual needs.

[0089] In addition, the formula also takes into account the importance of knowledge points in the test questions, introduces the concepts of assessment knowledge points, secondary assessment knowledge points and general knowledge points, and uses Min(a,b)', Min(a,b)'', Min(a,b)''' to represent their numbers. δ1, δ2, δ3, ε1, ε2, ε3 are used to further distinguish the weights of the three types of knowledge points.

[0090] Through the above formula, the content recommendation index can comprehensively evaluate the degree of match between the test content and the test requirements and the level of the examinees. In the process of test screening, test questions with high content recommendation will be given priority in the test paper, thus ensuring the pertinence and effectiveness of the test paper content at the source. At the same time, the calculation idea of ​​this indicator has a certain universality, and the weight coefficient can be flexibly adjusted according to the characteristics of different test subjects, which is suitable for tutoring scenarios of various academic level tests.

[0091] Optionally, the difficulty recommendation F2 is calculated according to the following formula:

[0092]

[0093] Among them, R is the difficulty coefficient of the test question, R′ is the difficulty requirement of the question type to which the test question belongs; R i , R j , R k are the difficulty levels of the i-th, j-th, and k-th knowledge points in the test questions respectively; Match is the matching calculation function; β1, β2, β3, θ1, θ2, θ3, is the weight coefficient.

[0094] More specifically, in the difficulty recommendation evaluation index of intelligent test paper composition, this application proposes a calculation method that comprehensively considers the overall difficulty of the test questions, the matching degree between the difficulty of knowledge points and the test requirements, and the matching degree between the difficulty of knowledge points and the cognitive level of the examinees.

[0095] First, the first term in the formula, β1Match{(R),(R′)}, is used to measure the degree of match between the overall difficulty coefficient R of the test question and the difficulty requirement R′ of the question type to which the test question belongs. The difficulty coefficient is a quantitative representation of the overall difficulty of the test question, while the difficulty requirement of the question type is the difficulty standard pre-set in the test design. The higher the degree of match between the two, the more the difficulty of the test question meets the requirements of the question type, and the higher the difficulty recommendation. The Match function is used to quantify the degree of match, and β1 is the weight coefficient of this term.

[0096] Secondly, the second term of the formula It is used to measure the matching degree between the difficulty of each knowledge point in the test and the mastery requirements of the corresponding knowledge point. i , R j , R k Respectively represent the difficulty of the i, j, and k knowledge points in the test questions, Y i , Y j , Y k It represents the mastery requirements of these knowledge points in the exam syllabus, and the Match function is used to quantify the degree of match between the two. Similar to the content recommendation, the difference in the importance of knowledge points is also considered here, and the concepts of assessment knowledge points, secondary assessment knowledge points and general knowledge points are introduced, and different weights are assigned using θ1, θ2, and θ3. The ∑ symbol represents the weighted sum of the matching degree of each knowledge point, and Min(a,b)', Min(a,bn)'', and Min(a,b)''' respectively represent the number of three types of knowledge points. β2 is the weight coefficient of the second term.

[0097] Finally, the third term of the formula It is used to measure the matching degree between the difficulty of each knowledge point in the test and the cognitive level of the examinee. i , Z j , Z k It represents the examinee's cognitive level of the i, j, and k knowledge points, and the other symbols are similar to the second item. This item introduces the factor of the examinee's actual level to further improve the accuracy of the difficulty recommendation. β3 is the weight coefficient of the third item.

[0098] Through the weighted sum of the above three items, the difficulty recommendation index can comprehensively evaluate the appropriateness of the difficulty of the test questions, taking into account the degree of matching with the requirements of the question type, the requirements of the knowledge points, and the cognitive level of the examinees. In the screening of test questions, test questions with high recommendation are more likely to be selected into the test paper, making the test paper more scientific and reasonable. In addition, the calculation idea of ​​this indicator can also flexibly adjust the weight coefficient, which is suitable for different types of test tutoring.

[0099] Optionally, the discrimination recommendation degree F3 is calculated according to the following formula:

[0100]

[0101] Among them, S is the discrimination coefficient of the test question, S′ is the discrimination requirement of the question type to which the test question belongs; Z 1i , Z 1j , Z 1k are the first examinee’s cognitive levels of the i, j, and k knowledge points, respectively; H1 is the first examinee’s test score; Z qi , Z qj , Z qk are the cognitive levels of the qth examinee on the i-th, j-th, and k-th knowledge points, respectively, and H q is the test score of the qth examinee; Match is the matching calculation function; γ1, γ2, ω1, ω2, ..., ω m is the weight coefficient.

[0102] More specifically, in the recommended evaluation index of the discrimination of intelligent test paper composition, this application proposes a calculation method that comprehensively considers the overall discrimination of the test questions, the cognitive level of the examinees and the matching degree of the test questions. The discrimination reflects the ability of the test questions to distinguish examinees of different levels and is an important indicator for evaluating the quality of the test questions.

[0103] First, the first term in the formula, γ1Match{(S),(S′)}, is used to measure the degree of match between the overall discrimination coefficient S of the test question and the discrimination requirement S′ of the question type to which the test question belongs. The discrimination coefficient quantifies the discrimination ability of the test question, and the question type discrimination requirement is the discrimination standard preset in the test design. The higher the match between the two, the more the discrimination of the test question meets the question type requirements, the better the quality of the test question, and the higher the discrimination recommendation. The Match function is used to quantify the degree of match, and γ1 is the weight coefficient of this term.

[0104] Secondly, the second term of the formula

[0105]

[0106] It is used to measure the matching degree between the cognitive level of different examinees and the test scores. This item introduces the data of m examinees, among which Z qi , Z qj , Z qk They represent the qth examinee’s cognitive level of the i, j, and kth knowledge points, respectively. q Represents the test score of the qth candidate. The average cognitive level of the qth examinee on the knowledge points contained in the test questions is calculated, and the Match function is used to quantify the correlation between the cognitive level and the test score H q ω1, ω2, ..., ω m The weight coefficient for each candidate's matching degree.

[0107] By introducing data from multiple examinees, this item can comprehensively evaluate the test questions' ability to distinguish examinees of different levels. If examinees with higher cognitive levels also score higher on the test questions, and examinees with lower cognitive levels also score lower, it means that the test questions have better discrimination and a higher recommendation. On the contrary, if the scores of examinees of different levels on the test questions are not significantly different, it means that the test questions have poor discrimination and a lower recommendation.

[0108] Finally, the discrimination recommendation index is weighted and summed by the two weight coefficients γ1 and γ2 to obtain the comprehensive discrimination evaluation result of the test questions. In the test question screening, test questions with high recommendation are more likely to be selected into the test paper, thereby improving the overall discrimination ability of the test paper.

[0109] In general, the discrimination recommendation index of this application evaluates the discrimination effect of test questions from both the overall and individual levels, fully considers the actual performance of candidates, and improves the scientificity and accuracy of the index. At the same time, this idea also provides new ideas for the in-depth mining and application of test data.

[0110] Optionally, the final recommendation degree F of the test questions in the intelligent test paper composition step is calculated according to the following formula:

[0111] F=F1·L1+F2·L2+F3·L3

[0112] Among them, F1, F2, and F3 are content recommendation, difficulty recommendation, and discrimination recommendation, respectively; L1, L2, and L3 are weight coefficients of each recommendation, and L1+L2+L3=1.

[0113] Optionally, the test paper quality assessment indicators include at least two of the following: the ratio of the knowledge point spectrum quotient value of the test paper to the examination target knowledge point spectrum quotient value G1; the proportion of the number of examination target test knowledge points covered by the test paper G2; and the average value G3 of the ratio of the maximum difficulty value of the question type used by the knowledge point in the test paper to the corresponding maximum difficulty value in the question bank.

[0114] Optionally, G1, G2, and G3 are calculated as follows:

[0115] G1=(∑eu / ∑fv)

[0116] G2=(m / n)

[0117] G3=(∑max w / Max w ) / m

[0118] Among them, f and v are the frequency of knowledge points in the test knowledge domain and the test score, e and u are the frequency of knowledge points in the test paper and the test score, n is the total number of knowledge points in the test knowledge domain, m is the total number of knowledge points involved in the test paper, max w is the maximum difficulty value of the question type used in the test paper for the wth knowledge point, Max w is the maximum difficulty value of the question type used in the question bank for the wth knowledge point.

[0119] Optionally, the test paper quality assessment result G in the test paper quality assessment step is calculated according to the following formula:

[0120]

[0121] in, is the weight coefficient.

[0122] More specifically, in the intelligent test paper composition step, this embodiment introduces the final recommendation degree F of the test questions as a comprehensive evaluation indicator for test question screening. F is obtained by weighted summation of the three sub-indicators of content recommendation degree F1, difficulty recommendation degree F2 and discrimination recommendation degree F3, where L1, L2 and L3 are the weight coefficients of the three indicators, and L1+L2+L3=1 is satisfied. Through this weighted combination method, F can more comprehensively evaluate the overall quality of the test questions, taking into account multiple dimensions such as test question content, difficulty and discrimination, thereby providing a reliable reference for test question screening.

[0123] Furthermore, in the test paper quality assessment step, this embodiment proposes three optional assessment indicators:

[0124] G1 = ∑eu / ∑fv, which represents the ratio of the knowledge point spectrum quotient of the test paper to the knowledge point spectrum quotient of the exam target. Among them, e and u are the frequency and score of the knowledge point in the test paper, and f and v are the frequency and score of the knowledge point in the knowledge point spectrum of the exam target. G1 reflects the coverage of the key knowledge points in the exam target by the test paper.

[0125] G2 = m / n, which means the ratio of the number of knowledge points covered by the test paper to the total number of knowledge points in the test target. Among them, m is the number of knowledge points included in the test paper, and n is the total number of knowledge points in the test target. G2 reflects the overall comprehensiveness of the test paper's coverage of the test knowledge points.

[0126] G3=(∑max w / Max w ) / m, which represents the average value of the ratio of the maximum difficulty of the corresponding question type of all knowledge points in the test paper to the maximum difficulty of the corresponding question type in the question bank. w and Max w are the maximum difficulty values ​​of the wth knowledge point in the test paper and the question bank. G3 reflects the overall matching degree of the test paper difficulty relative to the question bank.

[0127] Finally, the test paper quality evaluation result G is obtained by weighted summation of the above three indicators, namely: in is the weight coefficient, which can be adjusted according to actual needs.

[0128] Through the above indicator system, this application can comprehensively evaluate the overall quality of the test paper, such as content coverage and difficulty matching, and provide a quantitative basis for the selection of test papers, thereby improving the scientificity and effectiveness of the review plan. At the same time, the design ideas of each indicator have a certain universality and can be appropriately adjusted according to the characteristics of different examination subjects, which has broad application prospects in the field of intelligent test paper generation.

[0129] It should be pointed out that this embodiment proposes an innovative technical solution in the field of academic proficiency test tutoring. In view of the problems existing in the prior art of using the tactics of the sea of ​​questions for test review, such as lack of pertinence in review, incomplete coverage of knowledge points, and low review efficiency, this embodiment constructs a target test paper material library containing level test knowledge point information, test question information, and test paper template information, and based on the material library, by comprehensively considering the test question recommendation evaluation indicators of multiple dimensions such as test question content, difficulty, and discrimination, an intelligent test paper model is established to automatically screen test questions to form test papers. In the process of intelligent test paper composition, this embodiment adopts a strategy of parallel evaluation of three evaluation indicators, namely, content recommendation, difficulty recommendation, and discrimination recommendation. The three indicators quantify the pros and cons of test questions from the perspectives of knowledge point relevance, mastery degree matching, and discrimination ability, respectively, and obtain the comprehensive recommendation of the test questions through weighted summation, and filter and screen the test questions based on the comparison result of the comprehensive recommendation and the preset threshold, making full use of the complementary advantages and synergy of each evaluation indicator to ensure the pertinence of the content of the test paper and the rationality of the score. In addition, this embodiment also introduces a test paper quality evaluation mechanism based on multiple evaluation indicators such as knowledge point spectrum quotient value ratio, knowledge point coverage, knowledge point question type difficulty matching, etc. Through the iterative test paper composition and evaluation process, the dynamic optimization and closed-loop feedback of the test paper composition quality are realized, and finally a review plan that takes into account both comprehensiveness and effectiveness is output, which improves the learning effect of the test review. In short, the technical creativity of this embodiment is reflected in three aspects: first, the construction of the material library enhances the adaptability of the question source; second, the intelligent test paper composition model with multi-indicator fusion improves the scientific nature of the test paper composition; third, the iterative optimization quality evaluation mechanism ensures the effectiveness of the output review plan. By comprehensively applying the above technical means, this embodiment has achieved remarkable technical effects in improving the efficiency of test review and optimizing the learning process, and has substantial technical contributions and significant practical value.

[0130] In order to better understand the technical solution of the present application, a specific example is provided below for illustration. The details listed in the example are mainly for ease of understanding and are not intended to limit the scope of protection of the present application.

[0131] In this example, a method and system for tutoring academic proficiency tests are proposed. The method first constructs a test paper material library that includes various proficiency test knowledge point libraries, test question libraries, and test paper template libraries for intelligent test paper composition; secondly, by judging the three dimensions of content recommendation, difficulty recommendation, and discrimination recommendation of each extracted test question, it is determined whether the test question is extracted, thereby realizing intelligent test paper composition; then, the quality of each test paper is intelligently evaluated based on evaluation factors such as effectiveness, knowledge point coverage, and matching, to determine whether the test paper should be retained, and finally output a review plan.

[0132] The specific implementation steps are as follows Figure 2The specific implementation of each step is as follows:

[0133] (I) Constructing a test material library

[0134] Build a test paper material library that includes various level test knowledge point libraries, test question libraries and test paper template libraries to store and call test knowledge points, knowledge point-related test questions and test paper templates.

[0135] ① Level test knowledge point database: According to the knowledge point information in the examination syllabus of different types of level tests, the test knowledge point domain of each level test is obtained, and the knowledge point database of each level test is constructed to obtain the test knowledge points in the test requirements when compiling the test paper. The knowledge point database includes knowledge points and knowledge point attribute information. The knowledge point attribute information includes the level test stage, mastery requirements, and difficulty level.

[0136] ② Question bank: According to the requirements of different types of level tests, a massive question bank is built to extract test questions when compiling the test paper. The question bank includes test questions and question attribute information, which includes question type, test knowledge points, difficulty, discrimination, etc.

[0137] ③Exam paper template library: According to different types of proficiency test application scenarios, an exam paper template library is constructed to recommend and select exam paper templates when compiling exam papers. Each set of exam papers contains clear information on question types, number of questions, and score values.

[0138] (II) Intelligent test paper generation

[0139] Based on the level test knowledge point domain and test paper template required for test tutoring, the computer system repeats the following steps until all test questions in the test paper template are extracted: First, select a question type according to the test paper template; second, extract test questions containing the selected knowledge points from the preset test question bank according to the selected question type, and judge the three dimensions of content recommendation, difficulty recommendation, and discrimination recommendation of the extracted test questions to determine whether the test questions are extracted; finally, remove the knowledge points contained in the extracted test questions from the test knowledge point domain. The detailed process is as follows: Figure 3 shown.

[0140] Step 1: Determine the assessment weight of each knowledge point in the extracted test questions

[0141] Determine the number of assessment knowledge points, secondary assessment knowledge points, and general knowledge points in the test questions, and assign assessment weights to each knowledge point in the test questions according to the assessment weights. Among them, the assessment knowledge point weight value is assigned as μ1, the secondary assessment knowledge point weight value is assigned as μ2, and the general knowledge point weight value is assigned as μ3, and μ1>μ2>μ3 (according to the assessment characteristics of the test questions themselves and the actual experimental data, μ1, μ2, and μ3 are set to 0.6, 0.3, and 0.1 respectively).

[0142] Step 2: Calculate the recommended results of the extracted test content

[0143] According to the number of test knowledge points contained in the test questions, the matching degree between the assessment weight of each test knowledge point and the mastery requirements of each knowledge point in the test knowledge point domain, and the matching degree between the assessment weight of each test knowledge point and the cognitive level of the test group on each test knowledge point, the recommendation result F1 of the test question content is obtained. The specific calculation is:

[0144]

[0145] Among them, F1 is the result of the recommendation degree of the test content; Min(a,b) is the total number of test knowledge points contained in the test question, Max(a,b) is the knowledge points contained in the test knowledge point domain; Min(a,b)' is the number of assessment knowledge points in the test question, Min(a,b)'' is the number of secondary assessment knowledge points in the test question, and Min(a,b)''' is the number of general knowledge points in the test question; X i , X j , X k are the assessment weights of the i-th, j-th, and k-th knowledge points in the test questions respectively; Y i , Y j , Y k are the mastery requirements of the i, j, and k knowledge points in the test knowledge point domain, respectively. The values ​​are divided into y′, y″, and y′′′ according to the three levels of mastery requirements, and y′>y″>y′′′; Match(X j ,Y i )、Match(X j ,Y j )、Match(X k ,Y k ) are the matching indexes of the assessment weights and mastery requirements of the i, j, and k knowledge points in the test questions; Z i , Z j , Z k are the cognitive levels of the test group for the i-th, j-th, and k-th knowledge points in the test questions; Match(X i ,Z i )、Match(X j ,Z j )、Match(X k ,Z k ) are the matching indexes of the assessment weights of the ith, jth and kth knowledge points in the test questions and the cognitive level of the test group; α1, α2, α3, δ1, δ2, δ3, ε1, ε2 and ε3 are the weight coefficients respectively.

[0146] Step 3: Calculate the recommended results of the difficulty of the extracted test questions

[0147] According to the matching degree between the difficulty coefficient of the extracted test question and the difficulty requirement of the question type, the matching degree between the difficulty level of each test knowledge point in the test knowledge point domain and the mastering requirement, and the matching degree between the difficulty level of each test knowledge point and the cognitive level of the teaching group on each test knowledge point, the recommended degree result F2 of the test question difficulty is obtained. The specific calculation is:

[0148]

[0149] Among them, F2 is the result of the recommended difficulty of the test question; R is the difficulty coefficient of the test question, R′ is the difficulty requirement of the question type to which the test question belongs; Min(a,b)' is the number of knowledge points in the test question, Min(a,b)'' is the number of secondary knowledge points in the test question, and Min(a,b)''' is the number of general knowledge points in the test question; R i , R j , R k are the difficulty levels of the i, j, and k knowledge points in the test questions respectively. The values ​​are divided into r′, r″, and r''' according to the three levels of difficulty (i.e., high difficulty, medium difficulty, and low difficulty), and r′>r″>r'''; Y i , Y j , Y k are the mastery requirements of the i, j, and k knowledge points in the test knowledge point domain, respectively. The values ​​are divided into y′, y″, and y′′′ according to the three levels of mastery requirements, and y′>y″>y′′′; Match(R i ,Y i )、Match(R j ,Y j )、Match(R k ,Y k ) are the matching indexes of the difficulty and mastery requirements of the i, j, and k knowledge points in the test questions; Z i , Z j , Z k are the cognitive levels of the test group for the i, j, and k knowledge points in the test questions respectively; Match(R i ,Z i )、Match(R j ,Z j )、Match(R k ,Z k ) are the matching indexes of the difficulty of the i-th, j-th, and k-th knowledge points in the test questions and the cognitive level of the test group; β1, β2, β3, θ1, θ2, θ3, are weight coefficients respectively.

[0150] Step 4: Calculate the recommended result of the discrimination of the extracted test questions

[0151] According to the matching degree between the discrimination coefficient of the extracted test question and the discrimination requirement of the test type, and the matching degree between the cognitive level of each knowledge point of the test group and the test score, the recommended result F3 of the discrimination of the test question is obtained. The specific calculation is:

[0152]

[0153]

[0154] Among them, F3 is the recommended result of the test question discrimination; S is the discrimination coefficient of the test question, Min′ is the discrimination requirement of the question type; Min(a,b) is the total number of test knowledge points contained in the test question, Min(a,b)' is the number of assessment knowledge points in the test question, Min(a,b)'' is the number of secondary assessment knowledge points in the test question, and Min(a,b)''' is the number of general knowledge points in the test question; Z 1i , Z 1j , Z 1k are the first examinee’s cognitive levels of the i, j, and k knowledge points, respectively; H1 is the first examinee’s test score; Z 2i , Z 2j , Z 2k are the second student’s cognitive level of knowledge points u, j, and k respectively, H2 is the test score of the second examinee; Z qi , Z qj , Z qk are the cognitive levels of the qth examinee on the i-th, j-th, and k-th knowledge points, respectively, and H q is the test score of the qth student; γ1, γ2, ω1, ω2, ω3 are weight coefficients respectively.

[0155] Step 5: Comprehensively calculate the test question recommendation results and determine the test questions to be extracted

[0156] ① Based on the recommendation results of the above three dimensions and the weight distribution values ​​of each dimension, the test question recommendation result F is calculated comprehensively. The specific calculation is:

[0157] F=F1·L1+F2·L2+F3·L3

[0158] Among them, F is the comprehensive recommendation result of the test question; F1, F2, and F3 are the content recommendation result, difficulty recommendation result, and discrimination recommendation result of the test question respectively; L1, L2, and L3 are the weight values ​​of the three dimensions respectively, and L1+L2+L3=1.

[0159] ② Determine whether to extract the question based on the judgment result between the question recommendation result F and the question recommendation threshold F′: if F is greater than F′, the question is determined to be extracted; if F is less than F′, the question is not extracted.

[0160] (III) Output review plan

[0161] Based on the level test knowledge point domain and intelligent test paper generation results required by test tutoring, the computer system repeatedly executes the following steps until a review plan is generated: first, the quality of the generated test paper is intelligently evaluated by analyzing evaluation factors such as effectiveness, knowledge point coverage, and matching degree; second, based on the test paper quality evaluation results, it is determined whether the test paper is retained in the review plan; then, the number of test knowledge points included in the review plan is analyzed, and intelligent test paper generation continues until all test knowledge points are included in the review plan, and the review plan is output.

[0162] The specific implementation process is as follows:

[0163] Step 1: Obtain test paper quality assessment results

[0164] According to the dimensions of test paper validity, knowledge point coverage, matching degree, etc., the quality assessment results of test paper composition are obtained. Specifically, the test paper quality assessment is as follows:

[0165]

[0166] Where G is the quality assessment result of the test paper; f and v are the frequency of occurrence and test score of each knowledge point in the level test knowledge domain, e and u are the frequency of occurrence and test score of all tested knowledge points of the level test target covered in the test paper; n is the total number of tested knowledge points of the test target covered in the level test knowledge domain, m is the total number of tested knowledge points of the test target covered in the test paper; max w is the maximum difficulty value of the test question type used in the test paper for the wth knowledge point, Max w is the maximum difficulty value of the test question type used by the wth knowledge point in the question bank, are weight coefficients respectively.

[0167] Step 2: Determine whether to keep the test paper

[0168] Whether the test paper should be retained is determined based on the judgment result between the test paper quality assessment result G and the test paper quality assessment threshold G′: if G is greater than G′, the test paper is retained in the review plan and the knowledge points of the test paper are removed from the test knowledge point domain; if G is less than G′, the test paper is not retained.

[0169] Step 3: Output review plan

[0170] Based on all the test papers retained in the review plan, analyze whether the review plan contains all the knowledge points in the knowledge point domain of the level test. If not, intelligently compile the test papers based on the knowledge points that have not been eliminated in the knowledge point domain until all the test knowledge points are included in the review plan, and output the review plan.

[0171] The above embodiments have the following technical effects:

[0172] By building a test material library covering knowledge points, test questions, and test paper templates, and storing and associating various types of information in a structured manner, the source of test questions has been effectively expanded, the adaptability and flexibility of test questions to exam review needs have been improved, and a reliable data foundation has been provided for subsequent intelligent test paper generation.

[0173] A parallel question recommendation and evaluation strategy is adopted based on the three dimensions of content recommendation, difficulty recommendation and discrimination recommendation. Through nonlinear weight allocation and adaptive matching operations among the dimensions, multiple factors such as knowledge point relevance, mastery degree matching, and discrimination ability are comprehensively considered, thereby improving the pertinence and effectiveness of the test papers and overcoming the problems of incomplete knowledge point coverage and inaccurate control of test difficulty in the traditional test paper composition method.

[0174] During the intelligent test paper compilation process, the knowledge points that have been included in the test paper are eliminated through the dynamic update mechanism of the real-time knowledge point domain, which ensures the mutual exclusivity and independence of the knowledge points in the test paper, effectively avoids the redundancy of review content caused by repeated compilation of knowledge points, and significantly improves the efficiency of test review.

[0175] A test paper quality assessment mechanism based on multiple indicators such as knowledge point spectrum quotient value ratio, knowledge point coverage, and knowledge point question type difficulty matching has been introduced. By objectively quantifying the test paper quality and combining it with preset assessment thresholds, dynamic screening of test papers is achieved, ensuring the superiority of the test papers included in the review plan in terms of comprehensive knowledge point coverage and reasonable difficulty distribution.

[0176] In the process of generating review plans, a dynamic iterative optimization strategy based on evaluation indicators was adopted. Through multiple rounds of test paper generation and feedback adjustment of quality assessment, the review plans were systematically screened and improved. The final review plan not only comprehensively covers the knowledge points, but also takes into account the systematicity and scientificity, which is of positive significance to improving the overall learning effect of exam review.

[0177] The above-mentioned embodiment has strong technical adaptability and promotion value through a configurable evaluation index system and a flexible weight adjustment mechanism. It can not only be used for English tests such as CET-4 and CET-6 and IELTS, but can also be promoted to other subject areas such as computer grade examinations and accounting qualification examinations through parameter adjustment. It has exemplary significance for promoting the intelligent and personalized development of examination review models.

[0178] The second embodiment of the present application relates to an academic level test tutoring system, the structure of which is as follows: Figure 4 As shown, the academic level test tutoring system includes:

[0179] A test paper material library construction module is used to construct a target test paper material library containing level test knowledge point information, test question information and test paper template information;

[0180] An intelligent test paper composition module is used to automatically extract test questions from the test question information to form a test paper composition based on the target test paper composition material library and according to preset test question screening rules; the test question screening rules include multiple dimensions of test question recommendation evaluation indicators and corresponding thresholds;

[0181] The test paper quality evaluation module is used to evaluate the quality of the test paper formed by the intelligent test paper forming module based on the preset test paper quality evaluation index, and determine whether the test paper is included in the review plan according to the evaluation result;

[0182] The review plan generation module is used to summarize the test papers determined to be included by the test paper quality assessment module and form a review plan until the review plan covers the scope of target knowledge points.

[0183] The first implementation manner is a method implementation manner corresponding to the present implementation manner. The technical details in the first implementation manner can be applied to the present implementation manner, and the technical details in the present implementation manner can also be applied to the first implementation manner.

[0184] It should be noted that those skilled in the art should understand that the implementation functions of each module shown in the implementation of the above-mentioned academic level test tutoring system can be understood with reference to the relevant description of the above-mentioned academic level test tutoring method. The functions of each module shown in the implementation of the above-mentioned academic level test tutoring system can be realized by a program (executable instruction) running on a processor, or by a specific logic circuit. If the above-mentioned academic level test tutoring system of the present application embodiment is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the embodiment of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory (ROM, Read Only Memory), a disk or an optical disk. In this way, the embodiment of the present application is not limited to any specific combination of hardware and software.

[0185] Accordingly, an embodiment of the present application further provides a computer storage medium, in which computer executable instructions are stored. When the computer executable instructions are executed by a processor, the various method embodiments of the present application are implemented.

[0186] In addition, the embodiment of the present application also provides an academic proficiency test tutoring system, which includes a memory for storing computer executable instructions, and a processor; the processor is used to implement the steps in the above-mentioned method implementation when executing the computer executable instructions in the memory. Among them, the processor can be a central processing unit (Central Processing Unit, referred to as "CPU"), or other general-purpose processors, digital signal processors (Digital Signal Processor, referred to as "DSP"), Application Specific Integrated Circuit (Application Specific Integrated Circuit, referred to as "ASIC"), etc. The aforementioned memory can be a read-only memory (read-only memory, referred to as "ROM"), a random access memory (random access memory, referred to as "RAM"), a flash memory (Flash), a hard disk or a solid-state hard disk, etc. The steps of the method disclosed in each embodiment of the present invention can be directly embodied as a hardware processor to be executed, or a combination of hardware and software modules in the processor can be executed.

[0187] It should be noted that in the application documents of this patent, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "including one" do not exclude the existence of other identical elements in the process, method, article or device including the elements. In the application documents of this patent, if it is mentioned that an action is performed according to an element, it means that the action is performed at least according to the element, which includes two situations: performing the action only according to the element, and performing the action according to the element and other elements. Expressions such as multiple, multiple, and multiple include 2, 2 times, 2 kinds, and more than 2, more than 2 times, and more than 2 kinds.

[0188] All documents mentioned in this application are considered to be included in the disclosure of this application as a whole, so that they can be used as the basis for modification when necessary. In addition, it should be understood that after reading the above disclosure of this application, those skilled in the art can make various changes or modifications to this application, and these equivalent forms also fall within the scope of protection claimed in this application.

Claims

1. A method for tutoring for academic proficiency tests, characterized in that: The following steps are involved: Steps for constructing a test paper material library: constructing a target test paper material library containing level test knowledge point information, test question information and test paper template information; Intelligent test paper composition step: based on the target test paper composition material library, according to the preset test paper screening rules, test questions are automatically extracted from the test question information to form a test paper composition; the test question screening rules include multiple dimensions of test question recommendation evaluation indicators and corresponding thresholds; Test paper quality evaluation step: based on the preset test paper quality evaluation index, the test paper formed by the intelligent test paper generation step is evaluated for quality, and according to the evaluation result, it is determined whether the test paper is included in the review plan; Review plan generation step: Summarize the test paper quality assessment step to determine the test papers to be included, and form a review plan until the review plan covers the scope of target knowledge points.

2. The method for tutoring for academic proficiency tests according to claim 1, characterized in that: The level test knowledge point information includes: knowledge points and corresponding level test stages, mastery requirements, and difficulty level information; the test question information includes: test questions and corresponding question types, test knowledge points, difficulty, and discrimination information; the test paper template information includes: question types, number of questions, and test question score information.

3. The method for tutoring for academic proficiency tests according to claim 1, characterized in that: The multiple dimensions of test question recommendation evaluation indicators include: content recommendation, difficulty recommendation and discrimination recommendation.

4. The method for tutoring for academic proficiency tests according to claim 3, characterized in that: The content recommendation degree F1 is calculated according to the following formula: Among them, Min(a,b) is the total number of test knowledge points contained in the test question, and Max(a,b) is the total number of knowledge points contained in the test knowledge point domain; X i , X j , X k are the assessment weights of the i-th, j-th, and k-th knowledge points in the test questions respectively; Y i , Y j , Y k are the requirements for mastering the i-th, j-th, and k-th knowledge points in the test knowledge point domain; Z i , Z j , Z k are the cognitive levels of the test groups on the ith, jth and kth knowledge points respectively; Match is the matching calculation function; α1, α2, α3, δ1, δ2, δ3, ε1, ε2 and ε3 are weight coefficients.

5. The method for tutoring for academic proficiency tests according to claim 3, characterized in that: The difficulty recommendation F2 is calculated according to the following formula: Among them, R is the difficulty coefficient of the test question, R ′ R is the difficulty requirement of the question type; i , R j , R k are the difficulty levels of the i-th, j-th, and k-th knowledge points in the test questions respectively; Match is the matching calculation function; β1, β2, β3, θ1, θ2, θ3, is the weight coefficient.

6. The method for tutoring for academic proficiency tests according to claim 3, characterized in that: The discrimination recommendation F3 is calculated according to the following formula: Among them, S is the discrimination coefficient of the test question, S ′ Z is the discrimination requirement of the question type; 1i , Z 1j , Z 1k are the first examinee’s cognitive levels of the i, j, and k knowledge points, respectively; H1 is the first examinee’s test score; Z qi , Z qj , Z qk are the cognitive levels of the qth examinee on the i-th, j-th, and k-th knowledge points, respectively, and H q is the test score of the qth examinee; Match is the matching calculation function; γ1, γ2, ω1, ω2, ..., ω m is the weight coefficient.

7. The method for tutoring for academic proficiency tests according to claim 3, characterized in that: The final recommendation degree F of the test questions in the intelligent test paper composition step is calculated according to the following formula: F=F1·L1+F2·L2+F3·L3 Among them, F1, F2, and F3 are content recommendation, difficulty recommendation, and discrimination recommendation, respectively; L1, L2, and L3 are weight coefficients of each recommendation, and L1+L2+L3=1.

8. The method for tutoring for academic proficiency tests according to claim 1, characterized in that: The test paper quality evaluation index includes at least two of the following: the ratio of the knowledge point spectrum quotient value of the test paper to the test target knowledge point spectrum quotient value G1; the proportion of the number of test target test knowledge points covered by the test paper G2; The average value G3 of the ratio of the maximum difficulty value of the question types used in the test paper to the corresponding maximum difficulty value in the question bank.

9. The method for tutoring for academic proficiency tests according to claims 1 and 8, characterized in that: G1, G2, and G3 are calculated using the following formula: G1=(∑eu / ∑fv) G2=(m / n) G3=(∑max w / Max w ) / m Among them, f and v are the frequency of knowledge points in the test knowledge domain and the test score, e and u are the frequency of knowledge points in the test paper and the test score, n is the total number of knowledge points in the test knowledge domain, m is the total number of knowledge points involved in the test paper, max w is the maximum difficulty value of the question type used in the test paper for the wth knowledge point, Max w is the maximum difficulty value of the question type used by the wth knowledge point in the question bank; In addition, the test paper quality assessment result G in the test paper quality assessment step is calculated according to the following formula: in, is the weight coefficient.

10. An academic proficiency test tutoring system, characterized in that: include: A test paper material library construction module is used to construct a target test paper material library containing level test knowledge point information, test question information and test paper template information; An intelligent test paper composition module is used to automatically extract test questions from the test question information to form a test paper composition based on the target test paper composition material library and according to preset test question screening rules; the test question screening rules include multiple dimensions of test question recommendation evaluation indicators and corresponding thresholds; The test paper quality evaluation module is used to evaluate the quality of the test paper formed by the intelligent test paper forming module based on the preset test paper quality evaluation index, and determine whether the test paper is included in the review plan according to the evaluation result; The review plan generation module is used to summarize the test papers determined to be included by the test paper quality assessment module and form a review plan until the review plan covers the scope of target knowledge points.

Images