A test question recommendation method and device, electronic equipment and storage medium

Abstract

This disclosure relates to a test question recommendation method, apparatus, electronic device, and storage medium. The test question recommendation method includes: obtaining the target object's answer to a first test question; diagnosing the target object's mastery of a first knowledge point corresponding to the first test question based on the answer result, and obtaining the mastery level of the first knowledge point; determining a second knowledge point based on a pre-built association rule base between knowledge point mastery levels and the mastery level of the first knowledge point; and recommending a second test question containing the second knowledge point to the target object. The test question recommendation method provided by this disclosure can recommend suitable test questions to users based on the association rules between knowledge point mastery levels and the current mastery level of the knowledge point.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and in particular to a test question recommendation method, apparatus, electronic device, and storage medium. Background Technology

[0002] With the development of computer technology, online learning through learning products has become a common learning method. Among many adaptive learning products, personalized question recommendation is one of the most crucial components. It can recommend suitable questions from a massive question bank based on students' knowledge mastery of key concepts, thereby improving their learning efficiency. However, existing question recommendation methods largely rely on large amounts of historical answer data as prior information. When the available data is limited, they cannot recommend appropriate questions. Summary of the Invention

[0003] To address the aforementioned technical issues, this disclosure provides a test question recommendation method, apparatus, electronic device, and storage medium, which can recommend suitable test questions to users based on the association rules between knowledge point mastery.

[0004] According to one aspect of this disclosure, a test item recommendation method is provided, including:

[0005] Obtain the target object's answer to the first question;

[0006] Based on the answer results, diagnose the target object's mastery of the first knowledge point corresponding to the first question, and obtain the mastery level of the first knowledge point;

[0007] The second knowledge point is determined based on the pre-built association rule base between the mastery of knowledge points and the mastery level of the first knowledge point;

[0008] The second test question, which contains the second knowledge point, will be recommended to the target audience.

[0009] According to another aspect of this disclosure, a test item recommendation device is provided, comprising:

[0010] The acquisition module is used to obtain the target object's answer to the first question;

[0011] The diagnostic module is used to diagnose the target object's mastery of the first knowledge point corresponding to the first question based on the answer results, and to obtain the mastery level of the first knowledge point.

[0012] The prediction module is used to determine the second knowledge point based on a pre-built rule base for associations between knowledge point mastery and the mastery level of the first knowledge point;

[0013] The recommendation module is used to recommend a second test question containing the second knowledge point to the target object.

[0014] According to another aspect of this disclosure, an electronic device is provided, the electronic device 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 method recommended in the above-described test questions.

[0015] According to another aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions for causing the computer to perform a method recommended based on test questions.

[0016] According to another aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the above-described test question recommendation method.

[0017] The technical solution provided in this disclosure has the following advantages compared with the prior art:

[0018] This method involves obtaining the target user's answer to a first test question; diagnosing the target user's mastery of the first knowledge point corresponding to the first test question based on the answer; determining the mastery level of the first knowledge point based on a pre-built association rule base between knowledge point mastery levels and the mastery level of the first knowledge point; and recommending a second test question containing the second knowledge point to the target user. The test question recommendation method provided in this disclosure can recommend suitable test questions to users based on the current mastery level of knowledge points according to the association rules between knowledge point mastery levels. Attached Figure Description

[0019] 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.

[0020] 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.

[0021] Figure 1 A flowchart of the test question recommendation method provided in this embodiment of the disclosure;

[0022] Figure 2 A flowchart of the test question recommendation method provided in this embodiment of the disclosure;

[0023] Figure 3 A flowchart of the test question recommendation method provided in this embodiment of the disclosure;

[0024] Figure 4 A schematic diagram of the test question recommendation device provided in this embodiment of the disclosure;

[0025] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this disclosure. Detailed Implementation

[0026] To better understand the above-described objects, features, and advantages of this disclosure, embodiments of the disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of the disclosure are shown in the drawings, it should be understood that the 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 the 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.

[0027] 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.

[0028] The term "comprising" and its variations as used herein 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., used in this disclosure are only used 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.

[0029] 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".

[0030] 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.

[0031] Before describing the embodiments of this disclosure, the key terms involved in the following embodiments are introduced, wherein:

[0032] Association rules: Reflect the relationships between objects. If multiple objects are related, an object can be predicted based on the relationships between other objects.

[0033] Apriori algorithm: a well-known method for mining association rules.

[0034] Item Response Theory (IRT): Also known as latent trait theory or latent trait model, it is a mathematical model that can be used to analyze test scores.

[0035] Knowledge tracking: A technique that models students' knowledge mastery based on their historical answer records to obtain a representation of their current knowledge status (mastery level).

[0036] Currently, among numerous adaptive learning products, personalized question recommendation is one of the most crucial elements. Learning products can select the most suitable questions from a massive question bank based on students' knowledge mastery, helping to improve their learning efficiency. However, the development of learning products typically faces the following challenges: 1. Cold start problem for new users: When a new user logs into the system, without sufficient prior information, it's difficult to quickly identify the student's weak knowledge points and recommend questions of matching difficulty. Prior information refers to the user's historical answer data; 2. Explainability of recommendations: In educational scenarios, providing reasons for recommending questions makes it easier for students to understand and accept the recommended questions; 3. Real-time nature of recommendations: It's necessary to promptly capture students' knowledge mastery based on their answers and dynamically adjust the matched questions.

[0037] Existing adaptive methods for recommending test questions to users can be mainly categorized as follows: 1. Collaborative filtering-based methods: These calculate the similarity between students based on their answer records and predict student scores for candidate questions based on the scores of similar students on the recommended questions. 2. Matrix factorization-based methods: These generate latent vectors for students and questions based on the student-question score matrix and predict scores for questions the student has not yet answered. 3. Cognitive diagnostic-based recommendations: Common examples include Item Response Theory (IRT) models, which predict question scores based on student ability, question difficulty, discrimination, and guessing coefficient. 4. Knowledge tracing-based methods: Examples include Deep Knowledge Tracing (DKT) models, which predict scores based on students' historical answer sequences and select questions with predicted scores within a certain range for recommendation. In summary, the aforementioned methods, such as collaborative filtering and matrix factorization, which model based on the commonalities of similar students, struggle to guarantee the rationality and interpretability of recommended questions. The methods based on cognitive diagnosis and knowledge tracing require prediction or knowledge diagnosis based on students' historical answer data. When new users have no answer records or limited historical answer data, the prediction accuracy is poor, and they do not consider the sequential relationships between knowledge points, potentially leading to recommended questions exceeding the student's cognitive boundaries. Therefore, existing methods cannot solve the cold start problem and the interpretability problem in adaptive recommendation systems.

[0038] To address the aforementioned technical issues, this disclosure provides a test question recommendation method. By combining the sequential relationships between knowledge points in a knowledge graph and constructing a rule base for association between knowledge point mastery based on different students' historical answer data, when the recommendation system has limited historical answer data, it can infer possible intermediate knowledge points through frequent patterns. Then, based on the student's mastery of these intermediate knowledge points and the constructed rule base for association between knowledge point mastery, the recommended knowledge points are adjusted to provide students with suitable test questions in real time. This method solves the problem of mismatched difficulty levels when students have no historical answers or too few answer records within the tested knowledge point range. Secondly, the adaptive test question recommendation method provided in this disclosure uses knowledge points as the granularity, sets recommendation paths based on knowledge point mastery, and recommends test questions that match the student's ability within a single knowledge point, ensuring the rationality of the recommended test questions (the difficulty of the recommended test questions matches the student's cognitive level). It also provides students with reasons for recommending a particular test question, thus solving the problem of existing methods failing to provide interpretability for recommended test questions. Finally, the adaptive test recommendation method provided in this publication reduces the number of questions students need to answer by reasoning about weak knowledge points, avoiding the "sea of ​​questions" approach. It helps students quickly locate their weak knowledge points for learning with as few questions as possible, thereby improving their test-taking efficiency.

[0039] Figure 1 The flowchart of the test question recommendation method provided in this disclosure is applied to a terminal or server. In one feasible application scenario, a student and the terminal interact to generate an answer to a first test question. The server receives the answer from the terminal and sends a second test question to be recommended to the terminal based on the answer. In another feasible application scenario, the student and the terminal interact to generate an answer to a first test question, and then the terminal determines a second test question to recommend to the student based on the answer. It is understood that this disclosure also includes other possible application scenarios, which will not be elaborated here. The following embodiments use the terminal executing the test question recommendation method as an example for detailed explanation, specifically including... Figure 1 The following steps S110 to S140 are shown:

[0040] S110. Obtain the target object's answer to the first question.

[0041] Understandably, the terminal displays an answer interface, and the target object triggers the answer interface to generate an answer to the first question. The terminal then obtains this answer. The first question may be the target object's first attempt at answering a question. "First attempt" can be understood as the target object not having answered any related questions within the recommended knowledge point set, or it can be understood as the target object having answered a relatively small number of related questions within the recommended knowledge point set. The first question may also be a question recommended based on the question recommendation method provided in this disclosure.

[0042] Optionally, if the target audience is answering the question for the first time within the recommended knowledge point set, the first question can be recommended to the target audience through the following steps to obtain the answer result of the first question, wherein:

[0043] When recommending test questions to the target object for the first time within the recommended knowledge point set, the number of prerequisite knowledge points for each knowledge point in the recommended knowledge point set is calculated based on the constructed knowledge graph corresponding to the recommended knowledge point set.

[0044] Based on the number of prerequisite knowledge points for each knowledge point, intermediate knowledge points are determined in the recommended knowledge point set.

[0045] The first test question containing the intermediate knowledge points will be recommended to the target object.

[0046] Understandably, when a set of recommended knowledge points is selected, if it is determined that the target object is answering for the first time within the scope of this set of recommended knowledge points, the number of prerequisite knowledge points for each knowledge point in the set of recommended knowledge points is calculated based on the constructed knowledge graph. The number of knowledge points included in the knowledge graph is not less than the number of knowledge points included in the set of recommended knowledge points. A knowledge graph is a graph-based data structure that reflects the relationship network between knowledge points. The knowledge graph shows the state of knowledge points and the relationships between them. Arrows in the knowledge graph represent the logical order of learning knowledge points. Arrows point to subsequent knowledge points and lead to prerequisite knowledge points. That is, before learning a new knowledge point, you need to have the prerequisite knowledge points (prerequisite knowledge points). For example, before learning decimal addition and subtraction, you need to learn integer addition and subtraction. For the knowledge point of decimal addition and subtraction, integer addition and subtraction is one of the prerequisite knowledge points. Subsequently, based on the number of prerequisite knowledge points for each knowledge point, intermediate knowledge points are determined. After sorting all knowledge points by the number of prerequisite knowledge points, the knowledge point in the middle is selected as the intermediate knowledge point. If there are multiple intermediate knowledge points with the same number of prerequisite knowledge points, one of them can be randomly selected as the first knowledge point to be recommended. Then, the first test question containing the first knowledge point is recommended to the target user. This method can solve the problem of cold start for new users and can quickly locate weak knowledge points by the correlation between intermediate knowledge points and their prerequisite knowledge points.

[0047] S120. Based on the answer results, diagnose the target object's mastery of the first knowledge point corresponding to the first test question, and obtain the mastery level of the first knowledge point.

[0048] Understandably, based on the above S110, the target subject's mastery of the first knowledge point corresponding to the first question is diagnosed in real time according to the received answer results, and the mastery level of the first knowledge point is obtained. The mastery level is pre-divided, including the first level, the second level and the third level. The first level corresponds to a solid mastery, the second level corresponds to a moderate mastery, and the third level corresponds to a poor mastery. Knowledge points with poor or moderate mastery can be regarded as knowledge points with weak mastery by the target subject, which need to be strengthened.

[0049] Optionally, obtaining the mastery level of the first knowledge point in S120 above can be achieved through the following steps:

[0050] Using a pre-trained estimation model, the target object's mastery of the first knowledge point corresponding to the first test question is estimated based on the answer results, thereby obtaining the ability value of the first knowledge point.

[0051] The mastery level of the first knowledge point is obtained by calculating the quantile of the ability value of the first knowledge point according to the first threshold.

[0052] Understandably, a pre-trained estimation model (IRT model) is used to diagnose the target subject's understanding of the first knowledge point in real time, that is, to estimate the target subject's mastery of the first knowledge point and obtain an ability value for the first knowledge point, which is the IRT estimate. Subsequently, the ability value is quantified according to multiple first thresholds, mapping the continuous ability value to discrete values. The first thresholds can be 0.3, 0.4, and 0.3, thus mapping the ability value to three mastery levels: poor, average, and solid, to obtain the mastery level for the first knowledge point. The training process of the IRT model includes: acquiring historical answer data from multiple students, and for each student, labeling each historical answer data with a mastery level label for the relevant knowledge point. For example, student 1's mastery level for knowledge point 1 is labeled with label 1, indicating that student 1 has a solid mastery of knowledge point 1. The historical answer data and the knowledge point mastery level labels are then input into the IRT model to complete the training. The specific training process is not described in detail.

[0053] S130. Determine the second knowledge point based on the pre-built association rule base between the mastery of knowledge points and the mastery level of the first knowledge point.

[0054] Understandably, based on the above S120, the association rules between knowledge point mastery establish relationships between knowledge points and between knowledge point mastery levels. For example, before learning decimal addition and subtraction, one must first learn integer addition and subtraction; there is a relationship between these two knowledge points. If one's mastery of integer addition and subtraction is poor, then one is very likely to have poor mastery of decimal addition and subtraction as well. There is also a relationship between the mastery levels of these two knowledge points. Based on these two relationships, association rules between knowledge point mastery levels are constructed. Weak knowledge points associated with the first knowledge point are determined according to the mastery level of the first knowledge point, or intermediate knowledge points whose mastery level is unknown are determined according to the mastery level of the first knowledge point. Specifically, the second knowledge point can be determined based on the relationship between the mastery level of the first knowledge point and the first level.

[0055] Optionally, the association rule base between knowledge point mastery levels is implemented through the following steps:

[0056] Retrieve historical answer data for multiple objects, where each object's historical answer data corresponds to multiple knowledge points.

[0057] Based on each subject's historical answer data, estimate each subject's mastery of multiple knowledge points, and obtain the mastery level of multiple knowledge points for each subject.

[0058] The mastery levels of multiple knowledge points corresponding to each object are treated as a transaction, and the mastery level of one knowledge point is recorded as an item. The mastery levels of the target knowledge points corresponding to multiple objects are treated as an item set to obtain multiple item sets. The target knowledge point mastery level is at least one of the multiple knowledge point mastery levels.

[0059] A rule base for associations between knowledge point mastery is constructed based on all transactions and the multiple itemsets.

[0060] Understandably, historical response data for multiple subjects is obtained. This historical response data may partially overlap with the historical response data obtained during the training of the estimation model. Each subject's historical response data includes multiple response results, each corresponding to multiple knowledge points. These multiple knowledge points are at least partially identical across all subjects. Based on each subject's historical response data, the trained estimation model is used to estimate each subject's mastery of multiple knowledge points, determining the mastery level for each knowledge point. The method for determining the mastery level of each knowledge point is the same as in S120 above and will not be elaborated upon here. After determining the mastery levels of multiple knowledge points corresponding to each object, the set of mastery levels of multiple knowledge points for each object is regarded as a transaction. That is, one object corresponds to one transaction, and multiple objects have multiple transactions. Each mastery level of knowledge point is regarded as an item, and one knowledge point of an object corresponds to one item. The set of mastery levels of target knowledge points for multiple objects is regarded as an itemset. The target knowledge point mastery level is at least one of the mastery levels of multiple knowledge points. For example, object 1's mastery of knowledge point 1 is at level 1, object 2's mastery of knowledge point 1 is at level 2, and object 3's mastery of knowledge point 1 is at level 2. The mastery of knowledge point 1 is classified as Level 1. Knowledge point 1 is the target knowledge point. The set of one Level 1 and two Level 2 represents the mastery levels of different individuals for knowledge point 1, and can be considered as an itemset. The target knowledge point may be knowledge point 1 and knowledge point 2, meaning that question 1 involves two knowledge points. The historical answer data obtained from multiple individuals are not entirely the same. In this case, an individual may not have answered question 1, and therefore, the answer result for question 1 will not exist in the historical answer data of that individual. Thus, the number of items included in each itemset may be different, and the target knowledge points corresponding to each itemset are also not entirely the same.

[0061] Optionally, a rule base for association between knowledge point mastery can be constructed based on all transactions and the multiple itemsets. This can be achieved through the following steps:

[0062] Frequent itemsets are identified from the multiple itemsets, and multiple association rules are generated based on the frequent itemsets.

[0063] The confidence level of each association rule among the plurality of association rules is calculated based on all transactions and the frequent itemsets.

[0064] Based on the target association rules with a confidence level greater than the third threshold among the multiple association rules, an association rule base for knowledge point mastery is constructed.

[0065] Understandably, the Apriori algorithm is used to analyze and mine multiple itemsets to generate frequent itemsets. The frequency of an itemset is the number of transactions containing that itemset. If an itemset satisfies the minimum support requirement, it is called a frequent itemset. After identifying frequent itemsets, all possible association rules related to them are generated. Itemset X => itemset Y represents an association rule, indicating that Y can be derived from X. Here, X and Y are called the antecedent (or left-hand side, LHS) and consequent (or right-hand side, RHS) of the association rule, respectively. For example, if the itemsets of knowledge point 1 and knowledge point 2 are identified as frequent itemsets, there are possible association rules between them. Subsequently, the confidence of each association rule is calculated based on the two itemsets involved and all transactions included in those two itemsets. All target association rules with a confidence greater than a third threshold are stored, forming the association rule base for knowledge point mastery. The third threshold is the given minimum confidence.

[0066] Optionally, the calculation of the confidence score for each association rule can be achieved through the following steps:

[0067] Count the first number of all transactions, and count the second number of transactions that include the target knowledge point mastery level of the first frequent itemset.

[0068] Calculate the third number of mastery levels of all target knowledge points included in the first and second frequent itemsets.

[0069] The support of the target association rule is calculated based on the third quantity and the first quantity, wherein the target association rule is the association rule between the knowledge point mastery levels in the first sub-frequent itemset and the second sub-frequent itemset.

[0070] The support of the first sub-frequent itemset is calculated based on the first quantity and the second quantity, and the confidence of the target association rule is calculated based on the support of the first sub-frequent itemset and the support of the target association rule.

[0071] Understandably, we first count the total number of transactions, denoted as the first quantity, calculated using the formula `count(D)`, where `D` represents all transactions and `count` is used for counting. Then, we count the number of transactions that include the target knowledge point mastery level of the first frequent itemset, denoted as the second quantity. The second quantity refers to the total number of transactions that simultaneously include the target knowledge point, which must include at least one knowledge point. The formula for the second quantity is `count(x)`, where `x` represents the first frequent itemset. Finally, we count the total number of target knowledge point mastery levels included in both the first and second frequent itemsets, denoted as the third quantity. This is the union of all target knowledge points included in both the first and second frequent itemsets, calculated using the formula `count(x∪)`. y ),in, y Let be the second frequent itemset. The support of the first frequent itemset is calculated based on the first and second quantities, as shown in formula (1).

[0072] support(x)=count(x) / count(D) (1)

[0073] In the formula, su pp ort(x) represents the support of itemset x. When determining frequent itemsets in multiple itemsets, the support of each itemset is also determined by formula (1). If the support of an itemset is greater than the given minimum support, it means that the itemset is a frequent itemset. If support(x) is greater than the given minimum support, then itemset x is determined to be a frequent itemset, and then itemset x is recorded as the first sub-frequent itemset.

[0074] Understandably, the support of the target association rule is calculated based on the third quantity and the first quantity. The target association rule is the association rule between the knowledge mastery of the first and second sub-frequent itemsets, as shown in formula (2).

[0075] support(x=>y)=support(x∪y)=count(x∪y) / count(D) (2)

[0076] In the formula, x=>y represents the target association rule, and support(x=>y) represents the support of the target association rule.

[0077] Understandably, the confidence of the target association rule is calculated based on the support of the first frequent itemset support(x) and the support of the target association rule support(x=>y), as shown in formula (3).

[0078] confidence(x∪y)=support(x∪y) / support(x) (3)

[0079] In the formula, confidence(x∪y) represents the confidence level of the target association rule.

[0080] S140. Recommend the second test question containing the second knowledge point to the target object.

[0081] Understandably, based on the above S130, a preset number of second questions containing the second knowledge point are selected from the question database, and the preset number of second questions are recommended to the target audience.

[0082] Optionally, after identifying the second knowledge point, you can also generate a recommendation reason for the second question, which includes the following steps:

[0083] Determine a preset number of second test questions that contain the second knowledge point, and generate recommendation reasons for the second test questions.

[0084] While recommending the preset number of second test questions to the target audience, the reasons for the recommendation are also explained to the target audience.

[0085] Understandably, after identifying the second knowledge point to be recommended, a preset number of second test questions containing the second knowledge point are recommended to assess the target audience's mastery of the second knowledge point. Simultaneously, a recommendation reason is generated based on the mastery of the first knowledge point and the second knowledge point. For example, if the first knowledge point is poorly mastered, and the second knowledge point is a prerequisite or dependent knowledge point of the first, the target audience's mastery of the second knowledge point may also be poor. In this case, the generated recommendation reason could be, "Based on the mastery of the first knowledge point, it is estimated that the mastery of its dependent knowledge points is poor; therefore, practice questions containing dependent knowledge points are recommended." Conversely, if the first knowledge point is well mastered, and a second knowledge point with unknown mastery is needed, the recommendation reason could be, "The first knowledge point is well mastered; therefore, practice questions containing second knowledge points with unknown mastery are recommended." Simultaneously with recommending the preset number of second test questions to the target audience—that is, displaying the second test questions on the terminal—the recommendation reason is also displayed, making it easier for the target audience to understand and accept the recommended test questions, thus solving the problem of interpretability of the recommended test questions.

[0086] Understandably, after obtaining the answer to the first question and determining the mastery of the first knowledge point corresponding to the first question, the association rules between the mastery of the knowledge point are updated based on the answer and the mastery level of the first knowledge point. In other words, the association rule library is updated in real time according to the answer of the target object, and self-learning is carried out to improve the rationality of personalized question recommendations.

[0087] This disclosure provides a test question recommendation method that acquires a large number of students' historical answer results, labels each historical answer result with a mastery level tag for the knowledge points they have mastered, and mines the correlations between knowledge point mastery levels to construct a rule base for these correlations. This leverages the commonalities among students to address the cold start problem for certain knowledge points within the assessment scope, and can quickly diagnose a student's mastery of all knowledge points within the recommended knowledge point range, recommending test questions with difficulty matching the student's cognitive level. Furthermore, it recommends weak knowledge points based on the students' mastery levels and the rule base, reducing the amount of test questions students need to answer while improving their efficiency. The method also adjusts the next recommended test questions in real time based on the current knowledge point mastery, providing good interpretability; the generated recommendation reasons can guide students in understanding their learning progress.

[0088] Based on the above embodiments, Figure 2 Optionally, in the flowchart of the test question recommendation method provided in this embodiment, in the above-described S130, a second knowledge point is determined based on a pre-built association rule base between knowledge point mastery levels and the mastery level of the first knowledge point. The second knowledge point includes two scenarios: first, the second knowledge point may be a knowledge point where the target object has weak mastery; second, the second knowledge point may be an intermediate knowledge point in the recommended knowledge point set where the target object's mastery level is unknown. Specifically, this is achieved through methods such as... Figure 2 The following steps S210 to S230 show how to determine the second knowledge point in the first case, wherein:

[0089] S210. If the mastery level of the first knowledge point is lower than the first level, then the set of prerequisite knowledge points for the first knowledge point is determined from the corresponding set of recommended knowledge points based on the constructed knowledge graph.

[0090] Understandably, the determination of whether the mastery level of the first knowledge point is lower than the first level is crucial. The first level indicates a solid grasp of the knowledge point, while a level lower than the first level suggests that the target audience's mastery level of the first knowledge point may be the second or third level, meaning that their mastery of the first knowledge point is weak or average. In this case, based on the constructed knowledge graph, the prerequisite knowledge points for the first knowledge point are determined from the recommended knowledge point set, and a prerequisite knowledge point set for the first knowledge point is generated.

[0091] S220. Based on a pre-built association rule base between knowledge point mastery levels, determine at least one third knowledge point in the set of prerequisite knowledge points that is associated with the first knowledge point and whose mastery level is also lower than the first level.

[0092] Understandably, based on the above S210, and using a pre-built association rule base for knowledge point mastery, at least one third knowledge point is retrieved from the combination of prerequisite knowledge points that is associated with the first knowledge point and whose mastery level is also lower than the first level. For example, the set of prerequisite knowledge points for the first knowledge point includes knowledge point 1, knowledge point 2, knowledge point 3, and knowledge point 4. In the association rule base, there is an association rule between the first knowledge point and each of knowledge point 2, knowledge point 3, and knowledge point 4. In the association rule between the mastery of the first knowledge point and knowledge point 3, the mastery level of knowledge point 3 is also lower than the first level, indicating that if the mastery of knowledge point 3 is weak, the mastery of the first knowledge point is also weak. Therefore, knowledge point 3 is recorded as the third knowledge point. For another example, in the association rule between the mastery of the first knowledge point and knowledge point 4, the mastery level of knowledge point 4 is the first level, indicating that if the mastery of knowledge point 4 is solid, the mastery of the first knowledge point may also be weak. In this case, knowledge point 4 is not the third knowledge point.

[0093] S230. Based on the at least one third knowledge point, determine the second knowledge point that the target object has a weak grasp of.

[0094] Optionally, the second knowledge point that the target object has a weak grasp of, as identified in S230 above, can be achieved through the following steps:

[0095] Determine the confidence level of each of the at least one third knowledge points, and sort the at least one third knowledge points according to the confidence level.

[0096] The third knowledge points with a confidence level greater than the second threshold among at least one of the sorted third knowledge points are identified as the second knowledge points that the target object has a weak grasp of.

[0097] Understandably, based on the above S220, the confidence level of each third knowledge point in at least one third knowledge point is determined. The confidence level of the third knowledge point can be the confidence level of the association rule. The association rule between the mastery of the first knowledge point and the third knowledge point corresponds to a confidence level, that is, one association rule corresponds to one confidence level. Then, at least one third knowledge point is sorted according to the confidence level. The third knowledge point with a confidence level higher than the second threshold is identified as the second knowledge point where the target object has a weak mastery. The second knowledge point is also the knowledge point included in the next recommended test questions. If there are multiple third knowledge points with a confidence level higher than the second threshold, the third knowledge point with the highest confidence level is identified as the weak knowledge point.

[0098] Optionally, the second knowledge point in the second situation can be determined through the following steps, wherein:

[0099] If the mastery level of the first knowledge point is Level 1, then based on the pre-built association rule base between the mastery levels of knowledge points, at least one fourth knowledge point that is associated with the first knowledge point and whose mastery level is also Level 1 is determined.

[0100] Remove the first knowledge point and the at least one fourth knowledge point from the recommended knowledge point set, and determine the second knowledge point from the removed recommended knowledge point set.

[0101] Understandably, if the mastery level of the first knowledge point is determined to be Level 1, meaning the target audience has a solid grasp of the first knowledge point, then based on a pre-built association rule base for knowledge point mastery, at least one fourth knowledge point is identified that is associated with the first knowledge point and also has a Level 1 mastery. The fourth knowledge point is a knowledge point that has an association rule with the first knowledge point and is firmly mastered. For example, there might be an association rule between a solid grasp of decimal addition and subtraction and a solid grasp of integer addition and subtraction. The fourth knowledge point can be a predecessor, successor, or other knowledge point in the knowledge graph that has an established relationship with the first knowledge point. After identifying at least one fourth knowledge point, the first knowledge point and at least one fourth knowledge point are removed from the recommended knowledge point set. In other words, there is no need to recommend firmly mastered knowledge points to the target audience. Specifically, users can decide whether to remove firmly mastered knowledge points from the recommended knowledge point set based on their needs. Alternatively, a knowledge point can be removed only after it has been repeatedly confirmed as firmly mastered. Another approach is to reduce the probability of recommending a knowledge point or the number of recommended test questions after each knowledge point is confirmed as firmly mastered. Other possible removal methods are not limited. After removing the knowledge points that are firmly grasped from the recommended knowledge point set, determine the second knowledge point to be recommended from the removed recommended knowledge point set. In this case, the second knowledge point is also the intermediate knowledge point. The steps to determine the second knowledge point are the same as the steps to determine the first knowledge point to be recommended when the target object answers for the first time, and will not be repeated here.

[0102] The test question recommendation method provided in this embodiment allows for the following steps: If a student has a solid grasp of the first knowledge point, then based on an association rule base, at least one fourth knowledge point associated with the first knowledge point and with a solid grasp level is retrieved from the recommended knowledge point set. The first knowledge point and at least one fourth knowledge point with a solid grasp are removed from the recommended knowledge point set. The remaining recommended knowledge points, after removal, contain unanswered or unknown knowledge points. This method can quickly diagnose a student's grasp of all knowledge points within the recommended knowledge point range using only a small number of test questions. If a student has a poor or average grasp of the first knowledge point, then based on an association rule base, at least one third knowledge point associated with the first knowledge point and with a moderate or weak grasp is retrieved from the set of prerequisite knowledge points for the first knowledge point. These third knowledge points are then ranked according to their confidence level, and the third knowledge point with the highest confidence level is selected as the knowledge point for the next recommended test question. This method avoids the "sea of ​​questions" approach, reduces the number of questions students need to answer, and helps students quickly identify and address their weak knowledge points with a minimal number of questions.

[0103] Based on the above embodiments, Figure 3 The flowchart of the test question recommendation method provided in this embodiment of the disclosure specifically includes, as follows: Figure 3 The following steps S310 to S360 are shown:

[0104] S310. If the target object answers a question for the first time within the recommended knowledge point set, the first question containing the first knowledge point is recommended to the target object, where the first knowledge point is an intermediate knowledge point in the recommended knowledge point set.

[0105] S320. Based on the received answer data of the first test question, diagnose the target's mastery of the first knowledge point in real time and obtain the mastery level of the first knowledge point.

[0106] S330. Determine whether the mastery level of the first knowledge point is lower than the first level.

[0107] Understandably, based on the above S320, it is determined whether the mastery level of the first knowledge point is lower than the first level, that is, whether the target object has a general or poor mastery of the first knowledge point. If yes, it means that the target object has a weak or general mastery of the first knowledge point, and then S340 is executed; if no, it means that the target object has a solid mastery of the first knowledge point, and then S341 is executed.

[0108] S340. Based on the association rule base between the knowledge point mastery status, determine at least one third knowledge point that is associated with the first knowledge point and whose mastery level is also lower than the first level, and determine the second knowledge point among the at least one third knowledge point.

[0109] Understandably, based on the above S330, the second knowledge point that the target object to be recommended has weak knowledge is determined based on the association rule base.

[0110] S341. Based on the constructed association rule base between knowledge point mastery, determine at least one fourth knowledge point that is associated with the first knowledge point and has a mastery level of the first level, and remove at least one fourth knowledge point and the first knowledge point from the recommended knowledge point set.

[0111] Understandably, based on the above S330, at least one fourth knowledge point that is associated with and firmly grasped by the first knowledge point is determined according to the association rules, and at least one fourth knowledge point and the first knowledge point are removed from the recommended knowledge point set, that is, no longer recommended knowledge points that are firmly grasped to the target object, and then S350 is executed.

[0112] S350. Determine the second knowledge point from the set of recommended knowledge points after removal.

[0113] Understandably, based on the above S341, the same method as above S310 is used to determine the intermediate knowledge points in the removed set of recommended knowledge points as the second knowledge points.

[0114] S360. Recommend a preset number of second test questions containing the second knowledge point to the target audience.

[0115] Understandably, the specific implementation steps of S310 to S360 described above are explained in the above embodiments and will not be repeated here.

[0116] Based on the above embodiments, Figure 4 This is a schematic diagram of the structure of the test item recommendation device provided in this embodiment of the disclosure. The test item recommendation device 400 includes an acquisition module 410, a diagnosis module 420, a prediction module 430, and a recommendation module 440, wherein:

[0117] The acquisition module 410 is used to acquire the target object's answer to the first test question;

[0118] The diagnostic module 420 is used to diagnose the target object’s mastery of the first knowledge point corresponding to the first question based on the answer results, and to obtain the mastery level of the first knowledge point.

[0119] Prediction module 430 is used to determine the second knowledge point based on a pre-built association rule base between knowledge point mastery and the mastery level of the first knowledge point;

[0120] The recommendation module 440 is used to recommend a second test question containing the second knowledge point to the target object.

[0121] Optionally, the diagnostic module 420 is specifically used for:

[0122] Using a pre-trained estimation model, the target object's mastery of the first knowledge point corresponding to the first question is estimated based on the answer results, and the ability value of the first knowledge point is obtained.

[0123] The mastery level of the first knowledge point is obtained by calculating the quantile of the ability value of the first knowledge point according to the first threshold.

[0124] Optionally, the prediction module 430 is specifically used for:

[0125] If the mastery level of the first knowledge point is lower than the first level, then the set of prerequisite knowledge points for the first knowledge point is determined from the corresponding set of recommended knowledge points based on the constructed knowledge graph.

[0126] Based on a pre-built rule base for associations between knowledge point mastery levels, at least one third knowledge point is identified in the set of prerequisite knowledge points that is associated with the first knowledge point and whose mastery level is also lower than the first level.

[0127] Based on the at least one third knowledge point, determine the second knowledge point that the target object has a weak grasp of.

[0128] Optionally, the prediction module 430 is specifically used for:

[0129] Determine the confidence level of each of the at least one third knowledge points, and sort the at least one third knowledge points according to the confidence level;

[0130] The third knowledge points with a confidence level greater than the second threshold among at least one of the sorted third knowledge points are identified as the second knowledge points that the target object has a weak grasp of.

[0131] Optionally, the prediction module 430 is specifically used for:

[0132] If the mastery level of the first knowledge point is the first level, then based on the pre-built association rule base between the mastery status of knowledge points, at least one fourth knowledge point that is associated with the first knowledge point and whose mastery level is also the first level is determined.

[0133] Remove the first knowledge point and the at least one fourth knowledge point from the recommended knowledge point set, and determine the second knowledge point from the removed recommended knowledge point set.

[0134] Optionally, device 400 is also used for:

[0135] When recommending test questions to the target object for the first time within the scope of the recommended knowledge point set, the number of prerequisite knowledge points for each knowledge point in the recommended knowledge point set is calculated based on the constructed knowledge graph corresponding to the recommended knowledge point set.

[0136] Based on the number of prerequisite knowledge points for each knowledge point, intermediate knowledge points are determined in the recommended knowledge point set;

[0137] The first test question containing the intermediate knowledge points will be recommended to the target object.

[0138] Optionally, module 440 is recommended for:

[0139] Determine a preset number of second test questions that contain the second knowledge point, and generate recommendation reasons for the second test questions;

[0140] While recommending the preset number of second test questions to the target audience, the reasons for the recommendation are also explained to the target audience.

[0141] Optionally, device 400 is also used to build an association rule base, wherein:

[0142] Retrieve historical answer data for multiple objects, where each object's historical answer data corresponds to multiple knowledge points;

[0143] Based on the historical answer data of each object, estimate the mastery of each object on multiple knowledge points, and obtain the mastery level of multiple knowledge points for each object;

[0144] The mastery levels of multiple knowledge points corresponding to each object are treated as a transaction, and the mastery level of one knowledge point is recorded as an item. The mastery levels of the target knowledge points corresponding to the multiple objects are treated as an item set to obtain multiple item sets. The target knowledge point mastery level is at least one of the multiple knowledge point mastery levels.

[0145] A rule base for associations between knowledge point mastery is constructed based on all transactions and the multiple itemsets.

[0146] Optionally, device 400 is also used for:

[0147] Frequent itemsets are identified from the multiple itemsets, and multiple association rules are generated based on the frequent itemsets;

[0148] The confidence level of each association rule among the plurality of association rules is calculated based on all transactions and the frequent itemsets.

[0149] Based on the target association rules with a confidence level greater than the third threshold among the multiple association rules, an association rule base for knowledge point mastery is constructed.

[0150] Optionally, the frequent itemset in device 400 includes a first sub-frequent itemset and a second sub-frequent itemset.

[0151] Optionally, device 400 is also used for:

[0152] Count the first number of all transactions, and count the second number of all transactions that include the target knowledge point mastery level of the first frequent itemset.

[0153] Count the third number of mastery levels of all target knowledge points included in the first and second frequent itemsets;

[0154] The support of the target association rule is calculated based on the third quantity and the first quantity, wherein the target association rule is the association rule between the knowledge point mastery levels in the first sub-frequent itemset and the second sub-frequent itemset;

[0155] The support of the first sub-frequent itemset is calculated based on the first quantity and the second quantity, and the confidence of the target association rule is calculated based on the support of the first sub-frequent itemset and the support of the target association rule.

[0156] The device provided in this embodiment has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0157] 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.

[0158] 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.

[0159] refer to Figure 5 The present invention describes a structural block diagram of an electronic device 500 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.

[0160] like Figure 5As shown, the electronic device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. The RAM 503 may also store various programs and data required for the operation of the device 500. The computing unit 501, ROM 502, and RAM 503 are interconnected via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0161] Multiple components in electronic device 500 are connected to I / O interface 505, including: input unit 506, output unit 507, storage unit 508, and communication unit 509. Input unit 506 can be any type of device capable of inputting information to electronic device 500. Input unit 506 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 507 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 504 may include, but is not limited to, disks and optical discs. Communication unit 509 allows electronic device 500 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.

[0162] The computing unit 501 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 501 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 501 performs the various methods and processes described above. For example, in some embodiments, the test item recommendation method or the training method of the recognition network can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as storage unit 508. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 500 via ROM 502 and / or communication unit 509. In some embodiments, the computing unit 501 can be configured to perform the test item recommendation method or the training method of the recognition network by any other suitable means (e.g., by means of firmware).

[0163] 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.

[0164] 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.

[0165] 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.

[0166] 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).

[0167] 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.

[0168] 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.

[0169] 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 test item recommendation method, characterized in that, include: Obtain the target object's answer to the first question; Based on the answer results, diagnose the target object's mastery of the first knowledge point corresponding to the first question, and obtain the mastery level of the first knowledge point; The second knowledge point is determined based on the pre-built association rule base between the mastery of knowledge points and the mastery level of the first knowledge point; The second test question containing the second knowledge point will be recommended to the target object; The step of determining the second knowledge point based on a pre-built association rule base between knowledge point mastery levels and the mastery level of the first knowledge point includes: If the mastery level of the first knowledge point is the first level, then based on the pre-built association rule base between the mastery status of knowledge points, at least one fourth knowledge point that is associated with the first knowledge point and whose mastery level is also the first level is determined. Remove the first knowledge point and the at least one fourth knowledge point from the recommended knowledge point set, and determine the second knowledge point from the removed recommended knowledge point set.

2. The method according to claim 1, characterized in that, The step of diagnosing the target object's mastery of the first knowledge point corresponding to the first test question based on the answer result, and obtaining the mastery level of the first knowledge point, includes: Using a pre-trained estimation model, the target object's mastery of the first knowledge point corresponding to the first question is estimated based on the answer results, and the ability value of the first knowledge point is obtained. The mastery level of the first knowledge point is obtained by calculating the quantile of the ability value of the first knowledge point according to the first threshold.

3. The method according to claim 1, characterized in that, The step of determining the second knowledge point based on a pre-built association rule base between knowledge point mastery levels and the mastery level of the first knowledge point includes: If the mastery level of the first knowledge point is lower than the first level, then the set of prerequisite knowledge points for the first knowledge point is determined from the corresponding set of recommended knowledge points based on the constructed knowledge graph. Based on a pre-built association rule base between knowledge point mastery levels, at least one third knowledge point is identified in the set of prerequisite knowledge points that is associated with the first knowledge point and whose mastery level is also lower than the first level. Based on the at least one third knowledge point, determine the second knowledge point that the target object has a weak grasp of.

4. The method according to claim 3, characterized in that, The step of determining the second knowledge point where the target object has a weak grasp based on the at least one third knowledge point includes: Determine the confidence level of each of the at least one third knowledge points, and sort the at least one third knowledge points according to the confidence level; The third knowledge points with a confidence level greater than the second threshold among at least one of the sorted third knowledge points are identified as the second knowledge points that the target object has a weak grasp of.

5. The method according to claim 1, characterized in that, Before obtaining the target object's answer to the first test question, the method includes: When recommending test questions to the target object for the first time within the scope of the recommended knowledge point set, the number of prerequisite knowledge points for each knowledge point in the recommended knowledge point set is calculated based on the constructed knowledge graph corresponding to the recommended knowledge point set. Based on the number of prerequisite knowledge points for each knowledge point, intermediate knowledge points are determined from the recommended knowledge point set; The first test question containing the intermediate knowledge points will be recommended to the target object.

6. The method according to claim 1, characterized in that, The method of recommending a second test question containing the second knowledge point to the target object further includes: Determine a preset number of second test questions that contain the second knowledge point, and generate recommendation reasons for the second test questions; While recommending the preset number of second test questions to the target audience, the reasons for the recommendation are also explained to the target audience.

7. The method according to claim 1, characterized in that, The rule base for linking knowledge point mastery is constructed in the following way: Retrieve historical answer data for multiple objects, where each object's historical answer data corresponds to multiple knowledge points; Based on each subject's historical answer data, estimate each subject's mastery of multiple knowledge points to obtain the mastery level of multiple knowledge points for each subject; The mastery levels of multiple knowledge points corresponding to each object are treated as a transaction, and the mastery level of one knowledge point is recorded as an item. The mastery levels of the target knowledge points corresponding to multiple objects are treated as an item set to obtain multiple item sets. The target knowledge point mastery level is at least one of the multiple knowledge point mastery levels. A rule base for associations between knowledge point mastery is constructed based on all transactions and the multiple itemsets.

8. The method according to claim 7, characterized in that, The step of constructing an association rule base between knowledge point mastery based on all transactions and the multiple itemsets includes: Frequent itemsets are identified from the multiple itemsets, and multiple association rules are generated based on the frequent itemsets; The confidence level of each association rule among the plurality of association rules is calculated based on all transactions and the frequent itemsets. Based on the target association rules with a confidence level greater than the third threshold among the multiple association rules, an association rule base for knowledge point mastery is constructed.

9. The method according to claim 8, characterized in that, The frequent itemset includes a first sub-frequent itemset and a second sub-frequent itemset. The step of calculating the confidence level of each association rule among the multiple association rules based on all transactions and the frequent itemset includes: Count the first number of all transactions, and count the second number of all transactions that include the target knowledge point mastery level of the first frequent itemset. Count the third number of mastery levels of all target knowledge points included in the first and second frequent itemsets; The support of the target association rule is calculated based on the third quantity and the first quantity, wherein the target association rule is the association rule between the knowledge point mastery levels in the first sub-frequent itemset and the second sub-frequent itemset; The support of the first sub-frequent itemset is calculated based on the first quantity and the second quantity, and the confidence of the target association rule is calculated based on the support of the first sub-frequent itemset and the support of the target association rule.

10. A test question recommendation device, characterized in that, include: The acquisition module is used to obtain the target object's answer to the first question; The diagnostic module is used to diagnose the target object's mastery of the first knowledge point corresponding to the first question based on the answer results, and to obtain the mastery level of the first knowledge point. The prediction module is used to determine the second knowledge point based on a pre-built rule base for associations between knowledge point mastery and the mastery level of the first knowledge point; The recommendation module is used to recommend a second test question containing the second knowledge point to the target object; The prediction module is specifically used for: If the mastery level of the first knowledge point is the first level, then based on the pre-built association rule base between the mastery status of knowledge points, at least one fourth knowledge point that is associated with the first knowledge point and whose mastery level is also the first level is determined. Remove the first knowledge point and the at least one fourth knowledge point from the recommended knowledge point set, and determine the second knowledge point from the removed recommended knowledge point set.

11. An electronic device, characterized in that, The electronic device includes: Processor; and Stored program memory, The program includes instructions that, when executed by the processor, cause the processor to perform the test item recommendation method according to any one of claims 1-9.

12. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to execute the test item recommendation method according to any one of claims 1-9.

Images