Knowledge mastery degree evaluation method, question recommendation method, device, equipment and medium

Abstract

The present disclosure provides a knowledge mastery evaluation method, a question recommendation method, a device, equipment and a medium, wherein the knowledge mastery evaluation method comprises: obtaining a question difficulty and discrimination of a question of a target knowledge point, and a correct or incorrect result of an answer of a user to the question of the target knowledge point; obtaining a plurality of preset simulation capability values and a preset weight corresponding to each simulation capability value; obtaining a maximum likelihood estimation value corresponding to each simulation capability value under the target knowledge point according to the question difficulty, the discrimination and the correct or incorrect result of the answer; performing weighted average processing on the plurality of simulation capability values based on the preset weight corresponding to each simulation capability value and the maximum likelihood estimation value corresponding to each simulation capability value under the target knowledge point, to obtain a capability average of the user; and the capability average is used to represent the mastery degree of the user to the target knowledge point. The present disclosure can effectively guarantee the accuracy and reliability of the mastery degree evaluation result, and help to further guarantee the rationality of the question recommendation on this basis.

Description

Technical Field

[0001] This disclosure relates to the field of artificial intelligence, and in particular to methods for assessing knowledge mastery, methods for recommending questions, devices, equipment, and media. Background Technology

[0002] Knowledge tracing technology assesses students' mastery of knowledge points based on their answers, enabling it to recommend questions or develop learning paths. However, these technologies typically require massive amounts of answer data for model training, and the trained model is then used to assess student mastery. During the cold start phase, both the number of participants and the number of questions answered are limited, making it difficult to effectively train the model and assess student mastery. Most technologies then rely on answer accuracy rates for a simple assessment, which suffers from low accuracy and reliability, leading to inappropriate question recommendations based on inaccurate assessments. Summary of the Invention

[0003] To solve the above-mentioned technical problems, or at least partially solve them, this disclosure provides a method for assessing knowledge mastery, a method for recommending topics, an apparatus, equipment, and a medium.

[0004] According to one aspect of this disclosure, a method for assessing knowledge mastery is provided, comprising: obtaining the difficulty and discrimination of a target question, and the correctness of a user's answers to the target question; wherein the target question contains target knowledge points; obtaining a plurality of preset simulated ability values ​​and a preset weight corresponding to each simulated ability value; obtaining a maximum likelihood estimate of each simulated ability value under the target question based on the question difficulty, the discrimination, and the correctness of the answers; and performing a weighted average of the plurality of simulated ability values ​​based on the preset weights corresponding to each simulated ability value and the maximum likelihood estimate of each simulated ability value under the target question to obtain the user's average ability value; wherein the average ability value is used to characterize the user's mastery of the target knowledge points.

[0005] According to another aspect of this disclosure, a question recommendation method is provided, comprising: obtaining a user's mastery level of a target knowledge point; wherein the mastery level is obtained based on the aforementioned knowledge mastery assessment method; determining a target difficulty level that matches the user's mastery level according to a pre-set correspondence between question difficulty levels and mastery levels; recalling candidate questions corresponding to the target difficulty level, wherein the candidate questions contain the target knowledge point; ranking the candidate questions based on preset ranking reference factors, and recommending questions to the user based on the ranking results.

[0006] According to another aspect of this disclosure, a knowledge mastery assessment device is provided, comprising: a first acquisition module, configured to acquire the difficulty and discrimination of a target question, and the correctness of a user's answers to the target question; wherein the target question contains a target knowledge point; a second acquisition module, configured to acquire a plurality of preset simulated ability values ​​and a preset weight corresponding to each simulated ability value; a maximum likelihood estimation module, configured to acquire a maximum likelihood estimate of each simulated ability value under the target question based on the question difficulty, the discrimination, and the correctness of the answers; and a mastery assessment module, configured to perform a weighted average of the plurality of simulated ability values ​​based on the preset weights corresponding to each simulated ability value and the maximum likelihood estimate of each simulated ability value under the target question, to obtain the user's average ability value; wherein the average ability value is used to characterize the user's mastery of the target knowledge point.

[0007] According to another aspect of this disclosure, a question recommendation device is provided, comprising: a mastery level acquisition module, configured to acquire a user's mastery level of a target knowledge point; wherein the mastery level is obtained based on the aforementioned knowledge mastery assessment method; a difficulty level determination module, configured to determine a target difficulty level matching the user's mastery level according to a pre-set correspondence between question difficulty levels and mastery levels; a candidate question recall module, configured to recall candidate questions corresponding to the target difficulty level, wherein the candidate questions contain the target knowledge point; and a question sorting and recommendation module, configured to sort the candidate questions based on preset sorting reference factors, and recommend questions to the user based on the sorting results.

[0008] According to another aspect of this disclosure, an electronic device is provided, comprising: a processor; and a memory storing a program, wherein the program includes instructions that, when executed by the processor, cause the processor to perform the knowledge mastery assessment method or the question recommendation method.

[0009] According to another aspect of this disclosure, a computer-readable storage medium is provided, the storage medium storing a computer program for performing the knowledge mastery assessment method, or performing the question recommendation method.

[0010] The knowledge mastery assessment method and apparatus provided in this embodiment do not require pre-training of a model using massive amounts of answer data. Instead, multiple simulated ability values ​​can be pre-set, and the maximum likelihood estimate of each simulated ability value can be determined by combining the difficulty of the questions, the discrimination, and the correctness of the answers. Based on the preset weights of each simulated ability value, a weighted average of the multiple simulated ability values ​​can be processed to obtain a more reasonable average ability value. Even in the cold start phase, the accuracy and reliability of the mastery assessment results can be effectively guaranteed.

[0011] Based on obtaining accurate and reliable mastery assessment results, the question recommendation method and apparatus provided in this embodiment can further obtain a target difficulty level that matches the user's mastery level according to a pre-set correspondence between question difficulty level and mastery level, thereby realizing question recall, sorting and recommendation, so that the questions recommended to the user have a good match with the user, and effectively ensure the rationality of the questions recommended to the user.

[0012] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0013] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure.

[0014] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 A flowchart illustrating a method for assessing knowledge mastery provided in an embodiment of this disclosure;

[0016] Figure 2 A flowchart illustrating a topic recommendation method provided in an embodiment of this disclosure;

[0017] Figure 3 This is a schematic diagram of an adaptive recommendation method provided in an embodiment of the present disclosure;

[0018] Figure 4 A schematic diagram of the structure of a knowledge mastery assessment device provided in an embodiment of this disclosure;

[0019] Figure 5 This is a schematic diagram of the structure of a topic recommendation device provided in an embodiment of the present disclosure;

[0020] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present disclosure. Detailed Implementation

[0021] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0022] It should be understood that the steps described in the method embodiments of this disclosure may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of this disclosure is not limited in this respect.

[0023] The term "comprising" and its variations as used in this disclosure are open-ended, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below. It should be noted that the concepts of "first", "second", etc., mentioned in this disclosure are used only to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.

[0024] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0025] To better understand the above-mentioned objectives, features, and advantages of this disclosure, the solutions disclosed herein will be further described below. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0026] Figure 1This is a flowchart illustrating a method for assessing knowledge mastery according to an embodiment of this disclosure. This method can be executed by a knowledge mastery assessment device, which can be implemented using software and / or hardware, and is generally integrated into an electronic device. Figure 1 As shown, the method mainly includes the following steps S102 to S108:

[0027] Step S102: Obtain the difficulty and discrimination of the target question, as well as the correctness of the user's answers to the target question; wherein, the target question contains the target knowledge points.

[0028] This disclosure does not limit the type of the target question. In practical applications, questions in the question bank are marked with their difficulty level. Therefore, the difficulty information marked for the target question in the question bank can be directly obtained. Alternatively, the difficulty information recorded in the question bank can be preprocessed, such as through mapping, to convert it into a numerical range that is easier to process later. This is not limited here. In this disclosure, a discrimination score can be further obtained based on the question difficulty. This discrimination score can be used to distinguish user categories. For example, difficult questions have a higher discrimination score, which can distinguish excellent users who can answer difficult questions accurately. For example, the correctness of a user's answer to the target question can be marked using a method such as 0 or 1, where 1 indicates a correct answer and 0 indicates an incorrect answer, to facilitate subsequent processing.

[0029] Step S104: Obtain multiple preset simulation capability values ​​and the preset weights corresponding to each simulation capability value.

[0030] In this embodiment, multiple discrete simulated ability values ​​can be pre-set. For example, based on the distribution of question difficulty, multiple discrete ability values ​​can be simulated using the Guass-Hermite integral method to obtain multiple simulated ability values. These simulated ability values ​​exhibit a normal shoulder-shaped curve distribution, which better reflects the actual ability situation of users, i.e., it conforms to the ability distribution of the general user base. Furthermore, the preset weight corresponding to each simulated ability value also exhibits a normal shoulder-shaped curve distribution; that is, the weights corresponding to extremely low and extremely high simulated ability values ​​are typically small. The simulated ability values ​​and weights set in this way are more reasonable, which helps to further and more reasonably evaluate the user's ability value in subsequent assessments.

[0031] For example, the preset multiple simulation capability values ​​can be: [-7.85,-6.75,-5.83,-4.99,-4.21,-3.47,-2.75,-2.05,-1.36,-0.68,0,0.68,1.36,2.05,2.75,3.47,4.21,4.99,5.83,6.75,7.85], and each simulation capability value corresponds to a preset weight. Therefore, the weight corresponding to the above simulation capability values ​​can be: [2.1e-14,4]. [98e-11, 1.45e-08, 1.23e-06, 4.23e-05, 7.08e-04, 6.44e-03, 3.4e-02, 1.08e-01, 2.15e-01, 2.7e-01, 2.15e-01, 1.084e-01, 3.4e-02, 6.445e-03, 7.08e-04, 4.22e-05, 1.234e-06, 1.45e-08, 4.98e-11, 2.1e-14]. It should be noted that the aforementioned pre-set discrete distribution of multiple simulated ability values ​​is applicable to all users. That is, when assessing any user's knowledge of a particular topic, these multiple simulated ability values ​​can be obtained, and subsequent analysis can be conducted based on them.

[0032] Step S106: Based on the difficulty of the question, the discrimination, and the correctness of the answers, obtain the maximum likelihood estimate of each simulated ability value under the target question.

[0033] This embodiment of the disclosure can, based on the difficulty of the question, its discrimination, and the correctness of the answers, utilize the maximum likelihood estimation algorithm to obtain the maximum likelihood estimate of each simulated ability value under the target question. Maximum likelihood estimation is a statistical method based on the maximum likelihood principle, an application of probability theory in statistics. It provides a method for evaluating specified parameters given observed data; obtaining a parameter value that maximizes the probability of a sample occurring is called maximum likelihood estimation. Based on known question difficulty, discrimination, and the user's correctness of answers, this embodiment of the disclosure can assess the probability that the simulated ability value represents the user's actual ability value by calculating the maximum likelihood estimate of each simulated ability value under the target question. Therefore, based on the maximum likelihood estimates of each simulated ability value under the target question, the user's actual ability value can be reasonably inferred.

[0034] Step S108: Based on the preset weight corresponding to each simulated ability value and the maximum likelihood estimate corresponding to each simulated ability value under the target question, perform a weighted average on multiple simulated ability values ​​to obtain the user's average ability value; wherein, the average ability value is used to characterize the user's mastery of the target knowledge point.

[0035] By using the weighted average method described above, the average ability score obtained can more accurately reflect the user's true ability score, that is, the user's mastery of the target knowledge points contained in the target question.

[0036] The knowledge mastery assessment method provided in this embodiment does not require pre-training of a model using massive amounts of answer data. Instead, it can directly pre-set multiple simulated ability values ​​and determine the maximum likelihood estimate of each simulated ability value by combining the difficulty of the questions, the discrimination, and the correctness of the answers. Based on the preset weights of each simulated ability value, a weighted average of the multiple simulated ability values ​​can be processed to obtain a more reasonable average ability value. Even in the cold start phase, it can effectively ensure the accuracy and reliability of the mastery assessment results.

[0037] In some implementations, step S102, namely, obtaining the difficulty and discrimination of the target question, can be implemented with reference to the following steps (1) to (3):

[0038] Step (1) Obtain the difficulty level marked on the target question. For example, the difficulty level can be an integer value from 1 to 5, that is, the difficulty level includes the first level to the fifth level.

[0039] Step (2) involves mapping the difficulty levels to obtain the corresponding mapping values, which are then used as the difficulty level of the target question. Considering that difficulty levels are integer values, for ease of subsequent processing, the difficulty levels can be mapped to a specified range of values, and these mapped values ​​are used as the difficulty level of the target question. For example, the range of the mapping values ​​is [-M, M], meaning the question difficulty can be a value between [-M, M]. It should be noted that the question difficulty is not limited to integers; it can also be a decimal value obtained from the mapping, without restriction.

[0040] For example, we can first obtain the maximum level value (such as 5 mentioned above) and the minimum level value (such as 1 mentioned above) in the difficulty level, as well as the maximum difficulty value (such as M mentioned above) and the minimum difficulty value (such as -M mentioned above) for easier processing later. We can then set a mapping relationship between the level range and the difficulty range as needed. Based on this mapping relationship, we can obtain the question difficulty corresponding to the difficulty level of the target question. For example, the mapping relationship can be expressed as:

[0041]

[0042] Where b represents the mapped difficulty of the target question, that is, b is the mapped value corresponding to the difficulty level of the target question, which can be directly regarded as the question difficulty below. diff represents the difficulty level of the target question, diff max This represents the maximum level value, such as 5, diff.min This represents the minimum level value, such as 1; b max This represents the maximum difficulty level, such as M=4, b min This represents the minimum difficulty level, such as -M = -4. Where b max and b min It can be preset.

[0043] It should be noted that the above is merely an example, and the mapping relationship can be flexibly changed in practical applications; no restrictions are imposed here. Using the above method, difficulty levels can be flexibly converted into difficulty values ​​that are easier to process later.

[0044] Step (3) determines the discrimination index of the target question based on its difficulty. In this embodiment, the correlation between discrimination index and question difficulty is fully considered; therefore, the discrimination index of the target question is determined based on its difficulty. In some specific implementation examples, the discrimination index is positively correlated with the square of the question difficulty.

[0045] For example, assuming the discrimination index is 'a', the relationship between the discrimination index 'a' and the question difficulty 'b' can be:

[0046]

[0047] By using the above methods, we can effectively ensure that simple and difficult questions have a high degree of differentiation, while medium-difficulty questions have a relatively linear degree of differentiation, which is more in line with the actual situation.

[0048] In some implementations, step S106, namely, obtaining the maximum likelihood estimate of each simulated ability value under the target question based on the question difficulty, discrimination, and the correctness of the answers, can be performed with reference to steps A and B below:

[0049] Step A: For each simulated ability value, the simulated ability value, the question difficulty, and the discrimination index are used as parameters of the IRT model, so that the IRT model outputs the predicted probability of the user answering the target question correctly based on the parameters.

[0050] Item Response Theory (IRT) is a data model, specifically a parametric model. It primarily attempts to parameterize the questions (such as question difficulty, discrimination, and probability of correct answering) and the students (such as ability scores and knowledge mastery), and then uses a mathematical model to predict the probability of students answering questions correctly. In this embodiment, the IRT model used can be a two-parameter model, and its expression is as follows:

[0051]

[0052] Where θ represents the ability value, a represents the discrimination index, and b represents the difficulty of the question.

[0053] In practical applications, simulated ability values, question difficulty, and discrimination can be substituted into the IRT model to obtain predicted probabilities. Given a fixed question difficulty and discrimination, each simulated ability value will correspond to a predicted probability.

[0054] Step B: Based on the correctness of the answer and the predicted probability corresponding to the simulated ability value, obtain the maximum likelihood estimate of the simulated ability value under the target question.

[0055] In some specific implementation examples, for a certain simulated capability value θ, its maximum likelihood estimate is:

[0056] L(θ)=P score (1-P) (1-score)

[0057] The P-value can be obtained directly from the aforementioned IRT model, and the score is obtained based on the correctness of the answer. When the answer is correct, the score = 1, and when the answer is incorrect, the score = 0.

[0058] In related technologies, massive amounts of data are typically used for model training to obtain model parameters such as ability values. Alternatively, based on known model parameters such as ability values, discrimination, and question difficulty, the probability of a student answering a question accurately is predicted by solving for P. Taking the IRT model as an example, from the perspective of questions, it needs to have been answered by at least 500 people; from the perspective of students, at least 5 questions are needed for each knowledge point to achieve model convergence and obtain model parameter values. However, in the cold start phase, the amount of data available for training the model is limited. Therefore, related technologies cannot use parameterized models in the cold start phase. Instead, they simply initialize students' mastery level based on their answer accuracy. However, this approach is ineffective, as it completely ignores the difficulty information of the questions themselves. The mastery level assessed for a student answering an easy question accurately is the same as that for a student answering a difficult question accurately, which is unrealistic. Therefore, in this embodiment, question difficulty is introduced to assess students' mastery level. Instead of using massive amounts of data to train IRT model parameters, the discrimination index can be determined based on question difficulty. Multiple preset discrete simulated ability values ​​are used to first determine the predicted probability corresponding to each simulated ability value. Based on the known correct / incorrect answers, the maximum likelihood estimate of each simulated ability value under the target question is determined by combining the predicted probability. This effectively assesses the probability that the simulated ability value represents the user's actual ability value, allowing for a comprehensive inference of the user's actual ability value based on the maximum likelihood estimates of each simulated ability value under the target question. Since the above process incorporates information such as question difficulty, discrimination index, and the user's correct / incorrect answers, compared to related technologies that directly assess user knowledge mastery based on the user's correct / incorrect answers during the cold start phase, the method provided in this embodiment effectively ensures the reliability of assessing user knowledge mastery during the cold start phase.

[0059] In practical applications, there can be multiple target questions, meaning that multiple target questions containing the same target knowledge point can be analyzed simultaneously, making the assessment results of the user's mastery of the target knowledge point more accurate and reliable. Therefore, step S108, which involves weighted averaging of multiple simulated ability values ​​based on the preset weights corresponding to each simulated ability value and the maximum likelihood estimate of each simulated ability value under the target question, can be performed as follows:

[0060] Step a: For each target question, the actual weight of each simulated ability value under the target question is obtained by multiplying the preset weight corresponding to each simulated ability value and the maximum likelihood estimate corresponding to each simulated ability value under the target question.

[0061] Step b: Based on the actual weight of each simulated ability value under each target question, perform a weighted average of the multiple simulated ability values ​​corresponding to each target question.

[0062] For example, the final obtained mean ability (weighted average result) The following formula can be used to obtain the result:

[0063]

[0064] Where m represents the total number of preset simulated ability values, n represents the total number of target questions containing the target knowledge points, j represents the j-th target question, and i represents the i-th simulated ability value. The preset weight W(θ) corresponding to the i-th simulated ability value is used. ij And the maximum likelihood estimate L(θ) corresponding to the i-th simulated ability value under the j-th target question. ij The product of these values ​​yields the actual weight W(θ) of the i-th simulated ability value α under the j-th target question. ij L(θ) ij In this context, regardless of the target question, i.e., regardless of the value of j, the preset weight W(θ) corresponding to the i-th simulated ability value is... ij The value of is fixed and is a preset value. However, when the problem is different, that is, when j is different, the corresponding maximum likelihood estimate L(θ) will vary. ij The actual weights W(θ) will change. ij L(θ) ij This will also change. Based on this, by taking the weighted average of the multiple simulated ability values ​​corresponding to each target question according to the actual weight of each simulated ability value under each target question, the above-mentioned ability mean is obtained. This allows for a relatively accurate and objective assessment of the user's actual capabilities.

[0065] Based on the foregoing, this disclosure further provides a method for recommending topics. Figure 2 This is a flowchart illustrating a question recommendation method provided in an embodiment of this disclosure. This method can be executed by a question recommendation device, which can be implemented in software and / or hardware, and is generally integrated into an electronic device. Figure 2 As shown, the method mainly includes the following steps S202 to S208:

[0066] Step S202: Obtain the user's level of mastery over the target knowledge point; wherein, the level of mastery is obtained based on the aforementioned knowledge mastery assessment method, such as using the average ability of the user over the target knowledge point obtained in the final step. The specific methods for obtaining the representation will not be elaborated here.

[0067] Step S204: Based on the pre-set correspondence between question difficulty levels and mastery levels, determine the target difficulty level that matches the user's mastery level.

[0068] In practical applications, the relationship between different question difficulty levels and mastery levels can be pre-defined. In implementation, the mastery level can be normalized, for example, by normalizing it to a range of 0 to 1. This is not only more intuitive but also easier to correlate with question difficulty levels. For instance, the mastery level can be linearly scaled to normalize it to the range [0, 1]. Specifically, refer to the following formula:

[0069]

[0070] Here, θ′ represents the normalized level of mastery. This normalized level of mastery can then be correlated with the difficulty level of the questions. For example: θ′∈[0,0.4) corresponds to difficulty level 1; θ′∈[0.4,0.55) corresponds to difficulty level 2; θ′∈[0.4,0.7) corresponds to difficulty level 3; θ′∈[0.7,0.85) corresponds to difficulty level 4; and θ′∈[0.85,1] corresponds to difficulty level 5. It should be noted that the above is merely an example and should not be considered a limitation. In practical applications, the correspondence between question difficulty levels and mastery levels can be flexibly set. Based on the above correspondence, given the user's known level of mastery, the target difficulty level matching the user's mastery level can be easily and quickly determined.

[0071] Step S206: Recall candidate questions that correspond to the target difficulty level and contain the target knowledge points. In practical applications, questions in the question bank are marked with corresponding difficulty levels. Therefore, questions marked with the target difficulty level and containing the target knowledge points can be directly recalled as candidate questions. Furthermore, questions that the user has already answered can be filtered out, that is, the candidate questions do not include questions that the user has already answered.

[0072] Step S208: Sort the candidate questions based on preset sorting reference factors, and recommend questions to the user based on the sorting results.

[0073] For example, the ranking reference factors include one or more of the following: the number of times the question is cited, the frequency of the question being tested, the matching degree between the question and the test questions in the user's region, and the occasion in which the question appears. The number of times the question is cited refers to the number of times the question is cited in various scenarios. Different regions may use different textbooks, and the focus of knowledge may also differ, so the matching degree between the question and the test questions in the user's region can be examined. The occasions in which the question appears could be midterm exams, final exams, classroom quizzes, etc. The above is only an illustrative example; other ranking reference factors can be selected in practical applications, and there are no limitations here. Compared to the method of recommending questions solely based on similar text in related technologies, the method provided by this disclosure provides users with more reasonable and reliable questions that better meet their actual needs.

[0074] In summary, based on obtaining accurate and reliable mastery assessment results, the question recommendation method provided in this embodiment can obtain a target difficulty level that matches the user's mastery level according to the pre-set correspondence between question difficulty level and mastery level, thereby realizing question recall, sorting and recommendation, so that the questions recommended to the user have a good match with the user, and effectively ensure the rationality of the questions recommended to the user.

[0075] In some implementation examples, there are multiple ranking reference factors; in some implementations, step S208 above, that is, the step of ranking candidate questions based on preset ranking reference factors, can be performed with reference to steps 1 to 3 below:

[0076] Step 1: Obtain the weight corresponding to each ranking reference factor, and the reference value for each ranking reference factor for each candidate question. The reference value of the ranking reference factor is the quantified value of the candidate question under each ranking reference factor. Factors such as the number of times the question is cited, the frequency of the question being tested, and the matching degree between the question and test questions in the user's region are easily quantified, and the reference value for these ranking reference factors can be obtained directly. For ranking reference factors such as the occasion in which the question appears, values ​​corresponding to various occasions can be pre-set. For example, the reference value can be set according to the importance of the occasion in which the question appears; for example, the reference value for a classroom test occasion is 1, the reference value for a midterm exam occasion is 2, and the reference value for a final exam occasion is 3, etc. The above is only an illustrative example and should not be considered a limitation. Through the above method, the various ranking reference factors can be quantified, making subsequent ranking processing based on the quantified reference values ​​easier.

[0077] Step 2: For each candidate topic, perform a weighted summation based on the weight of each ranking reference factor and the reference value of each ranking reference factor corresponding to the candidate topic to obtain the comprehensive reference value corresponding to the candidate topic.

[0078] In practical applications, the weight of each ranking reference factor can be flexibly set according to requirements, and the comprehensive reference value Order can be obtained in the following way:

[0079] Order=w1f1+w2f2+w3f3……+w n f n

[0080] Among them, w i f represents the weight corresponding to the i-th ranking reference factor. i This represents the reference value corresponding to the i-th ranking factor. Using the above method, the reference values ​​of multiple ranking factors can be combined to reasonably evaluate candidate questions.

[0081] Step 3: Sort the multiple candidate topics according to the comprehensive reference value of each candidate topic.

[0082] In practical applications, multiple candidate questions can be sorted in descending order of comprehensive reference value. This allows the top n candidate questions with the highest comprehensive reference value to be recommended to the user. These questions not only match the user's level of mastery but are also selected by considering multiple sorting factors, making them more reasonable and reliable.

[0083] Based on the foregoing, this disclosure provides an application scenario for the aforementioned knowledge mastery assessment method and question recommendation method. For example, see [link to relevant documentation]. Figure 3The diagram illustrates an adaptive recommendation method, showing the business layer and the diagnosis + recommendation layer. The business layer is primarily located on the user terminal, while the diagnosis + recommendation layer is primarily located on the server side. Specifically, users can complete book exercises offline and upload their complete answers to the server using methods such as taking photos for feedback. The server can use a knowledge diagnosis module to assess the user's mastery of the knowledge points in the answered questions based on the uploaded answer records. The assessment method can employ the aforementioned knowledge mastery assessment method. The server can display the user's knowledge mastery level on the user terminal and ask if the user needs to engage in adaptive practice. If the user confirms the need for adaptive practice, the server can use the aforementioned question recommendation method to recall, sort, and recommend questions, generating practice questions for the user to complete. The server can then further retrieve the user's answer records for the recommended practice questions and analyze the user's knowledge mastery level through the knowledge diagnosis module, recommending questions again. This process continues until the user meets the exit criteria and exits the practice. If the exit criteria are not met, the server continues to display the user's current knowledge mastery level and asks if the user wants to engage in adaptive practice. For example, the exit condition could be that the user has completed N recommended practice questions, or that the user's mastery level is higher than a preset threshold. Additionally, when a user uploads a full page of answer data, it may contain n questions. The knowledge points corresponding to different questions may be the same or different. Therefore, it can be categorized according to the main knowledge points corresponding to each question: K = {K1, K2, K3…Kn}, where K represents the set of all knowledge points in the full page of practice questions uploaded by the user, and Kn represents the nth category of knowledge points. For each category of knowledge points, questions containing that knowledge point can be used as target questions, and the correctness of the answers to the target questions under that knowledge point can be statistically analyzed, using a method such as K... score =[0,1,0,…,1] etc., briefly represent the answer results of each question under this knowledge point, where 1 indicates a correct answer and 0 indicates an incorrect answer; based on the above information and combined with the knowledge mastery assessment method provided in this embodiment, the user's mastery of each type of knowledge point can be reasonably assessed.

[0084] It should be noted that the above is only an application example. In practical applications, the knowledge mastery assessment method and question recommendation method provided in this disclosure embodiment can be used in any scenario that requires knowledge mastery assessment or question recommendation.

[0085] In summary, the knowledge mastery assessment method and adaptive recommendation method provided in this disclosure do not require pre-training of models using massive amounts of answer data. Even during the cold start phase, they can effectively ensure the accuracy and reliability of the mastery assessment results. Furthermore, based on obtaining accurate and reliable mastery assessment results, the questions recommended to users have a good match with the users, effectively ensuring the rationality of the recommended questions.

[0086] Corresponding to the aforementioned method for assessing knowledge mastery, this disclosure also provides a device for assessing knowledge mastery. Figure 4 This is a schematic diagram of a knowledge mastery assessment device provided in an embodiment of this disclosure. The device can be implemented by software and / or hardware and is generally integrated into an electronic device. Figure 4 As shown, the knowledge mastery assessment device 400 includes:

[0087] The first acquisition module 402 is used to acquire the difficulty and discrimination of the target question, as well as the correctness of the user's answers to the target question; wherein, the target question contains target knowledge points;

[0088] The second acquisition module 404 is used to acquire multiple preset simulation capability values ​​and preset weights corresponding to each simulation capability value;

[0089] The maximum likelihood estimation module 406 is used to obtain the maximum likelihood estimate of each simulated ability value under the target question based on the question difficulty, discrimination, and the correctness of the answer.

[0090] The mastery assessment module 408 is used to perform weighted averaging of multiple simulated ability values ​​based on the preset weights corresponding to each simulated ability value and the maximum likelihood estimate of each simulated ability value under the target question, so as to obtain the user's average ability value; wherein, the average ability value is used to characterize the user's mastery of the target knowledge point.

[0091] The knowledge mastery assessment device provided in this embodiment does not require pre-training of a model using massive amounts of answer data. Instead, it can directly pre-set multiple simulated ability values ​​and determine the maximum likelihood estimate of each simulated ability value by combining the difficulty of the questions, the discrimination, and the correctness of the answers. Based on the preset weights of each simulated ability value, a weighted average of the multiple simulated ability values ​​can be processed to obtain a more reasonable average ability value. Even in the cold start phase, it can effectively ensure the accuracy and reliability of the mastery assessment results.

[0092] In some implementations, the first acquisition module 402 is specifically used to: acquire the difficulty level marked on the target question; perform mapping processing on the difficulty level to obtain the mapping value corresponding to the difficulty level, and use the mapping value as the question difficulty of the target question; and determine the discrimination of the target question based on the question difficulty of the target question.

[0093] In some implementations, the mapping value ranges from [-M, M]; the discrimination is positively correlated with the square of the question difficulty.

[0094] In some implementations, the maximum likelihood estimation module 406 is specifically used to: for each simulated ability value, use the simulated ability value, the question difficulty, and the discrimination as parameters of the IRT model, so that the IRT model outputs a predicted probability that the user correctly answers the target question based on the parameters; and obtain the maximum likelihood estimate of the simulated ability value under the target question based on the correctness of the answer and the predicted probability corresponding to the simulated ability value.

[0095] In some implementations, the number of target questions is multiple, and the mastery assessment module 408 is specifically used to: for each target question, obtain the actual weight of each simulated ability value under the target question based on the product between the preset weight corresponding to each simulated ability value and the maximum likelihood estimate corresponding to each simulated ability value under the target question; and perform weighted averaging of the multiple simulated ability values ​​corresponding to each of the target questions based on the actual weight of each simulated ability value under each target question.

[0096] In some implementations, the plurality of simulated capability values ​​are distributed in a normal shoulder-shaped curve, and the preset weight corresponding to each simulated capability value is distributed in a normal shoulder-shaped curve.

[0097] The knowledge mastery assessment device provided in this disclosure can execute the knowledge mastery assessment method provided in any embodiment of this disclosure, and has the corresponding functional modules and beneficial effects for executing the method.

[0098] Corresponding to the aforementioned question recommendation method, this disclosure also provides a question recommendation device. Figure 5 This is a schematic diagram of a topic recommendation device provided in an embodiment of this disclosure. The device can be implemented by software and / or hardware and is generally integrated into an electronic device. Figure 5 As shown, the recommended device 500 includes:

[0099] The mastery level acquisition module 502 is used to acquire the user's mastery level for the target knowledge point; wherein, the mastery level is obtained based on the aforementioned knowledge mastery assessment method;

[0100] The difficulty level determination module 504 is used to determine the target difficulty level that matches the user's level of mastery based on the pre-set correspondence between the difficulty level of the questions and the level of mastery.

[0101] The candidate question recall module 506 is used to recall candidate questions that correspond to the target difficulty level, and the candidate questions contain the target knowledge points;

[0102] The question sorting and recommendation module 508 is used to sort candidate questions based on preset sorting reference factors and recommend questions to users based on the sorting results.

[0103] Based on obtaining accurate and reliable mastery assessment results, the question recommendation device provided in this embodiment can obtain a target difficulty level that matches the user's mastery level according to the pre-set correspondence between question difficulty level and mastery level, thereby realizing question recall, sorting and recommendation, so that the questions recommended to the user have a good match with the user, and effectively ensure the rationality of the questions recommended to the user.

[0104] In some implementations, there are multiple ranking reference factors; the question ranking and recommendation module 508 is specifically used to: obtain the weight corresponding to each ranking reference factor, and the reference value of each ranking reference factor corresponding to each candidate question; for each candidate question, perform weighted summation based on the weight corresponding to each ranking reference factor and the reference value of each ranking reference factor corresponding to the candidate question to obtain the comprehensive reference value corresponding to the candidate question; and rank the multiple candidate questions according to the comprehensive reference value of each candidate question.

[0105] In some implementations, the sorting reference factors include one or more of the following factors: the number of times the question is cited, the frequency of the question being tested, the matching degree between the question and the test questions in the user's region, and the occasion in which the question appears.

[0106] The question recommendation device provided in this disclosure can execute the question recommendation method provided in any embodiment of this disclosure, and has the corresponding functional modules and beneficial effects of the execution method.

[0107] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described device embodiments can be referred to the corresponding process in the method embodiments, and will not be repeated here.

[0108] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0109] Exemplary embodiments of this disclosure also provide an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor. The memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to cause the electronic device to perform a method according to an embodiment of this disclosure.

[0110] Exemplary embodiments of this disclosure also provide a non-transitory computer-readable storage medium storing a computer program, wherein the computer program, when executed by a computer's processor, is used to cause the computer to perform a method according to embodiments of this disclosure.

[0111] Exemplary embodiments of this disclosure also provide a computer program product, including a computer program, wherein, when executed by a processor of a computer, the computer program is used to cause the computer to perform a method according to an embodiment of this disclosure.

[0112] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this disclosure. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on a user's computing device, partially on a user's computing device, as a standalone software package, partially on a user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0113] Furthermore, embodiments of this disclosure can also be computer-readable storage media storing computer program instructions that, when executed by a processor, cause the processor to perform the XYZ method provided in embodiments of this disclosure. The computer-readable storage medium can be any combination of one or more readable media. A readable medium can be a readable signal medium or a readable storage medium. A readable storage medium can be, for example, including but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0114] refer to Figure 6The present invention describes a structural block diagram of an electronic device 600 that can serve as a server or client of the present disclosure, which is an example of a hardware device that can be applied to various aspects of the present disclosure. The electronic device is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0115] like Figure 6 As shown, the electronic device 600 includes a computing unit 601, which can perform various appropriate actions and processes based on a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 into a random access memory (RAM) 603. The RAM 603 may also store various programs and data required for the operation of the device 600. The computing unit 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0116] Multiple components in electronic device 600 are connected to I / O interface 605, including: input unit 606, output unit 607, storage unit 608, and communication unit 609. Input unit 606 can be any type of device capable of inputting information to electronic device 600. Input unit 606 can receive input digital or character information and generate key signal inputs related to user settings and / or function control of electronic device. Output unit 607 can be any type of device capable of presenting information and may include, but is not limited to, a display, speaker, video / audio output terminal, vibrator, and / or printer. Storage unit 608 may include, but is not limited to, disks and optical discs. Communication unit 609 allows electronic device 600 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks, and may include, but is not limited to, modems, network cards, infrared communication devices, wireless communication transceivers, and / or chipsets, such as Bluetooth™ devices, WiFi devices, WiMax devices, cellular communication devices, and / or the like.

[0117] The computing unit 601 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 601 performs the various methods and processes described above. For example, in some embodiments, the knowledge mastery assessment method or the question recommendation method can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 800 via ROM 802 and / or communication unit 809. In some embodiments, the computing unit 801 can be configured to perform the knowledge mastery assessment method or the question recommendation method by any other suitable means (e.g., by means of firmware).

[0118] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0119] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0120] As used in this disclosure, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, device, and / or apparatus (e.g., disk, optical disk, memory, programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including machine-readable media that receive machine instructions as machine-readable signals. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0121] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0122] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0123] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other.

[0124] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0125] The above description is merely a specific embodiment of this disclosure, enabling those skilled in the art to understand or implement it. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not to be limited to the embodiments described herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for assessing knowledge mastery, comprising: The difficulty and discrimination of the target question, as well as the correctness of the user's answers to the target question, are obtained; wherein, the target question contains the target knowledge point; Obtain multiple preset simulation capability values ​​and a preset weight corresponding to each of the simulation capability values; Based on the question difficulty, the discrimination index, and the correctness of the answers, obtain the maximum likelihood estimate of each simulated ability value under the target question; Based on the preset weight corresponding to each simulated ability value and the maximum likelihood estimate corresponding to each simulated ability value under the target question, the multiple simulated ability values ​​are weighted and averaged to obtain the user's ability mean; wherein, the ability mean is used to characterize the user's mastery of the target knowledge point; The step of obtaining the maximum likelihood estimate of each simulated ability value under the target question based on the question difficulty, the discrimination index, and the correctness of the answers includes: For each simulated ability value, the simulated ability value, the question difficulty, and the discrimination index are used as parameters of the IRT model, so that the IRT model outputs a predicted probability that the user will answer the target question correctly based on the parameters; Based on the correctness of the answer and the predicted probability corresponding to the simulated ability value, the maximum likelihood estimate of the simulated ability value under the target question is obtained.

2. The method for assessing knowledge mastery as described in claim 1, wherein, The steps for obtaining the difficulty and discrimination of the target question include: Obtain the difficulty level indicated on the target question; The difficulty level is mapped to obtain the mapping value corresponding to the difficulty level, and the mapping value is used as the difficulty of the target question. The discrimination factor of the target question is determined based on the difficulty level of the target question.

3. The method for assessing knowledge mastery as described in claim 2, wherein, The range of the mapping value is [-M, M]; the discrimination is positively correlated with the square of the difficulty of the question.

4. The method for assessing knowledge mastery as described in claim 1, wherein, The number of target questions is multiple. The step of performing a weighted average of the multiple simulated ability values ​​based on the preset weight corresponding to each simulated ability value and the maximum likelihood estimate corresponding to each simulated ability value under the target questions includes: For each target question, the actual weight of each simulated ability value under the target question is obtained by multiplying the preset weight corresponding to each simulated ability value and the maximum likelihood estimate corresponding to each simulated ability value under the target question. Based on the actual weight of each simulated ability value under each target question, a weighted average is performed on the multiple simulated ability values ​​corresponding to each of the target questions.

5. The method for assessing knowledge mastery as described in claim 1, wherein, The multiple simulated capability values ​​are distributed in a normal shoulder-shaped curve, and the preset weight corresponding to each simulated capability value is also distributed in a normal shoulder-shaped curve.

6. A method for recommending questions, including: The user's level of mastery of a target knowledge point is obtained; wherein the level of mastery is obtained based on the knowledge mastery assessment method according to any one of claims 1 to 5; Based on the pre-set correspondence between question difficulty levels and mastery levels, a target difficulty level matching the user's mastery level is determined; Recall candidate questions that correspond to the target difficulty level, and the candidate questions contain the target knowledge points; The candidate questions are ranked based on preset ranking reference factors, and questions are recommended to the user based on the ranking results.

7. The question recommendation method as described in claim 6, wherein, The number of ranking reference factors is multiple; the step of ranking the candidate questions based on the preset ranking reference factors includes: Obtain the weight corresponding to each of the ranking reference factors, and the reference value of each of the ranking reference factors corresponding to each of the candidate questions; For each candidate question, a weighted summation is performed based on the weight corresponding to each ranking reference factor and the reference value of each ranking reference factor corresponding to the candidate question to obtain the comprehensive reference value corresponding to the candidate question. The candidate topics are ranked according to the comprehensive reference value of each candidate topic.

8. The question recommendation method as described in claim 6, wherein, The sorting reference factors include one or more of the following factors: the number of times the question is cited, the frequency of the question being tested, the matching degree between the question and the test questions in the user's region, and the occasion in which the question appears.

9. A device for assessing knowledge mastery, comprising: The first acquisition module is used to acquire the difficulty and discrimination of the target question, as well as the correctness of the user's answers to the target question; wherein, the target question contains target knowledge points; The second acquisition module is used to acquire a plurality of preset simulation capability values ​​and a preset weight corresponding to each of the simulation capability values; The maximum likelihood estimation module is used to obtain the maximum likelihood estimate of each simulated ability value under the target question based on the question difficulty, the discrimination, and the correctness of the answer. The mastery assessment module is used to perform a weighted average of the multiple simulated ability values ​​based on the preset weight corresponding to each simulated ability value and the maximum likelihood estimate corresponding to each simulated ability value under the target question, so as to obtain the user's average ability value; wherein, the average ability value is used to characterize the user's mastery of the target knowledge point; The step of obtaining the maximum likelihood estimate of each simulated ability value under the target question based on the question difficulty, the discrimination index, and the correctness of the answers includes: For each simulated ability value, the simulated ability value, the question difficulty, and the discrimination index are used as parameters of the IRT model, so that the IRT model outputs a predicted probability that the user will answer the target question correctly based on the parameters; Based on the correctness of the answer and the predicted probability corresponding to the simulated ability value, the maximum likelihood estimate of the simulated ability value under the target question is obtained.

10. A question recommendation device, comprising: The mastery level acquisition module is used to acquire the user's mastery level for the target knowledge point; wherein, the mastery level is obtained based on the knowledge mastery assessment method according to any one of claims 1 to 5; The difficulty level determination module is used to determine a target difficulty level that matches the user's level of mastery based on a pre-set correspondence between question difficulty levels and mastery levels. The candidate question recall module is used to recall candidate questions that correspond to the target difficulty level, and the candidate questions contain the target knowledge points; The question sorting and recommendation module is used to sort the candidate questions based on preset sorting reference factors, and recommend questions to the user based on the sorting results.

11. An electronic device, comprising: processor; as well as Stored program memory, The program includes instructions that, when executed by the processor, cause the processor to perform the knowledge mastery assessment method according to any one of claims 1-5, or to perform the question recommendation method according to any one of claims 6-8.

12. A computer-readable storage medium storing a computer program for performing a knowledge mastery assessment method according to any one of claims 1-5, or for performing a question recommendation method according to any one of claims 6-8.

Images