CN116595162B

CN116595162B - Question recommending method based on electronic student identity data and related equipment thereof

Info

Publication number: CN116595162B
Application number: CN202310874213.6A
Authority: CN
Inventors: 何全; 魏壮勇; 何潮华
Original assignee: Telepower Education Co ltd
Current assignee: Guangdong Suishou Elf Technology Co ltd
Priority date: 2023-07-17
Filing date: 2023-07-17
Publication date: 2023-09-15
Anticipated expiration: 2043-07-17
Also published as: CN116595162A

Abstract

The application belongs to the technical field of auxiliary teaching, and discloses a question recommending method based on electronic student identity data and related equipment thereof, wherein the method comprises the following steps: acquiring a historical question set and a corresponding wrong question set of a target user; determining the knowledge of the relevant knowledge points and the target user according to the historical question set and the wrong question set; according to the related knowledge points and the corresponding mastery degrees, determining other users as reference users of the target users, and screening recommendation questions from a wrong question set of the reference users according to the related knowledge points and the corresponding mastery degrees so as to recommend the recommendation questions to the target users; therefore, the topics can be recommended to the weak points of the knowledge of the user more effectively, and the exercise effect is improved.

Description

Question recommending method based on electronic student identity data and related equipment thereof

Technical Field

The application relates to the technical field of auxiliary teaching, in particular to a question recommending method based on electronic student identity data and related equipment thereof.

Background

At present, some electronic student certificates have the function of conducting online answering exercise through answering software, and wrong question sets can be formed according to answering conditions so as to record historical wrong questions of users, so that the users can conveniently conduct repeated exercise on related questions according to the wrong question sets, and the grasping degree of related knowledge points is enhanced.

However, the wrong question set is generated only according to questions made by the user, and in fact, some questions which are not made by the user can relate to knowledge weak points of the user, repeated exercises are performed only according to the wrong question set, more related questions matched with the user cannot be recommended to the user according to the knowledge blind points of the user, and the exercise effect is poor.

Disclosure of Invention

The application aims to provide a question recommending method based on electronic student identity data and related equipment thereof, which can more effectively recommend questions aiming at weak points of user knowledge and improve exercise effect.

In a first aspect, the present application provides a method for question recommendation based on electronic student status data, comprising the steps of:

A1. acquiring a historical question set and a corresponding wrong question set of a target user;

A2. determining the knowledge of the relevant knowledge points and the target user according to the historical question set and the wrong question set;

A3. according to the related knowledge points and the corresponding mastery degrees, other users are determined to serve as reference users of the target users;

A4. and screening recommended questions from the wrong question set of the reference user according to the related knowledge points and the corresponding mastery degree so as to recommend the recommended questions to the target user.

Firstly, determining the grasping degree of each relevant knowledge point by the target user according to the historical question set and the wrong question set of the target user, wherein the grasping degree is matched with other users with the grasping degree similar to the target user for each relevant knowledge point to serve as reference users, and finally, selecting recommended questions from the wrong question set of the reference users to be recommended to the target user.

Preferably, step A2 comprises:

acquiring knowledge points related to each wrong question in the wrong question set of the target user, and marking the knowledge points as related knowledge points;

counting the total occurrence times of each related knowledge point in the wrong question set of the target user, and recording the total occurrence times as first times;

counting the total occurrence times of each related knowledge point in the history topic collection, and recording the total occurrence times as second times;

And calculating the grasping degree of the target user on each relevant knowledge point according to the first times and the second times.

The grasping degree is calculated according to the occurrence times of the related knowledge points in the wrong question set and the occurrence times of the related knowledge points in the historical question set, so that the grasping degree of the target user on the related knowledge points can be reflected more accurately.

Preferably, the step of calculating the grasping degree of each of the related knowledge points by the target user according to the first number of times and the second number of times includes:

calculating the grasping degree of the target user on each relevant knowledge point according to the following formula:

；

wherein ,degree of mastery of the i-th relevant knowledge point for the target user, +.>For the first number of i-th relevant knowledge points, +.>For the second number of i-th relevant knowledge points, n is the total number of relevant knowledge points.

Optionally, step A3 includes:

taking other users, which are related to the wrong question set and comprise all the related knowledge points, as first candidate users;

calculating a first matching degree of each first candidate user and the target user according to the grasping degree of each relevant knowledge point of the target user and the grasping degree of each first candidate user on each relevant knowledge point;

And selecting at least one first candidate user as the reference user according to the first matching degree.

The knowledge weak points and thinking habits of the reference users screened out in the mode are more similar to those of the target users, so that recommended topics have better exercise effects on the target users.

Optionally, step A3 includes:

according to the grasping degree of the target user on each relevant knowledge point, screening the relevant knowledge points to screen effective knowledge points;

taking other users, which are related to the wrong question set and comprise all the effective knowledge points, as second candidate users;

calculating a second matching degree of each second candidate user and the target user according to the grasping degree of each effective knowledge point of the target user and the grasping degree of each second candidate user on each effective knowledge point;

and selecting at least one second candidate user as the reference user according to the second matching degree.

Effective knowledge points are screened out firstly, and then second candidate users are selected according to the effective knowledge points, so that more second candidate users can be obtained, and the reference users can be screened out more reliably.

Preferably, step A4 comprises:

taking the relevant knowledge points with the grasping degree of the target user lower than a preset grasping degree threshold value as target knowledge points;

and screening topics which are related to the target knowledge point and are not included in the error topic set of the target user from the error topic set of the reference user as recommended topics so as to be recommended to the target user.

Preferably, step A4 comprises:

acquiring a related knowledge point combination and a corresponding error-causing association degree of error questions related to at least two related knowledge points in an error question set or a total error question set of the target user; the error-causing association degree characterizes the significance degree of the answering errors of the target user, which can be caused by the corresponding relevant knowledge point combination; the total error question set is a set formed by the error question sets of the target user and all the reference users;

and screening recommended topics from the error topic sets of the reference users according to the relevant knowledge point combinations and the corresponding error-causing association degrees so as to recommend the recommended topics to the target users.

In a second aspect, the present application provides a question recommending apparatus based on electronic student status data, including:

the first acquisition module is used for acquiring a historical question set and a corresponding wrong question set of the target user;

The grasping degree determining module is used for determining grasping degrees of related knowledge points and the target user on the related knowledge points according to the historical question set and the wrong question set;

the reference user determining module is used for determining other users as reference users of the target user according to the related knowledge points and the corresponding mastery degree;

and the screening module is used for screening recommended questions from the wrong question set of the reference user according to the related knowledge points and the corresponding mastery degree so as to recommend the recommended questions to the target user.

In a third aspect, the present application provides an electronic device comprising a processor and a memory, the memory storing a computer program executable by the processor, when executing the computer program, running the steps in the topic recommendation method based on electronic student status data as described above.

In a fourth aspect, the present application provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps in a method of topic recommendation based on electronic student status data as described hereinbefore.

The beneficial effects are that: according to the topic recommendation method based on the electronic student identity data and the related equipment thereof, the grasping degree of the target user on each related knowledge point is determined according to the historical topic set and the wrong topic set of the target user, the grasping degree is matched with other users with the grasping degree similar to the target user on each related knowledge point to serve as reference users, the recommended topics are finally screened out from the wrong topic set of the reference users to be recommended to the target user, and the reference users determined in the mode have similar knowledge weak points and similar thinking habits with the target users, so that the wrong topics of the reference users are also topics which are easy to be misplaced by the target users, the recommended topics are screened out from the wrong topic set of the reference users to be recommended to the target users, the knowledge weak points of the target users are more targeted, the training effect can be effectively improved, and the grasping degree of the target users on the knowledge weak points is improved.

Drawings

Fig. 1 is a flowchart of a method for recommending topics based on electronic student status data according to an embodiment of the present application.

Fig. 2 is a schematic structural diagram of a topic recommendation device based on electronic student status data according to an embodiment of the present application.

Fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Description of the reference numerals: 1. a first acquisition module; 2. a mastery degree determining module; 3. a reference user determination module; 4. a screening module; 301. a processor; 302. a memory; 303. a communication bus.

Detailed Description

The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. The components of the embodiments of the present application generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the application, as presented in the figures, is not intended to limit the scope of the application, as claimed, but is merely representative of selected embodiments of the application. All other embodiments, which can be made by a person skilled in the art without making any inventive effort, are intended to be within the scope of the present application.

It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further definition or explanation thereof is necessary in the following figures. Meanwhile, in the description of the present application, the terms "first", "second", and the like are used only to distinguish the description, and are not to be construed as indicating or implying relative importance.

Referring to fig. 1, fig. 1 is a method for recommending topics based on electronic student status data according to some embodiments of the present application, including the steps of:

A2. determining the knowledge of the related knowledge points and the target user according to the historical question set and the wrong question set;

A3. determining other users as reference users of the target users according to the related knowledge points and the corresponding mastery degree;

The method for recommending the questions based on the electronic student identity data can be applied to a server, the server is in communication connection with the electronic student identity of the target user, the target user logs in the server through answering software of the electronic student identity, and the server acquires a historical question set and a corresponding wrong question set of the target user according to account information sent when the electronic student identity logs in. The historical question set is a set of questions made by a user, and the historical question set records question information of the questions made by the user; the wrong question set is a set of missed questions made by the user, and the wrong question set records question information of the missed questions made by the user. The topic information comprises at least one item of index information such as topic number, topic stem, keywords and the like.

The error question set of the target user may be manually sorted by the target user (e.g., the target user manually removes the questions that have been mastered from the error question set) or periodically and automatically sorted by the server (e.g., the mastering degree of the target user on the error question is determined according to the accuracy of repeated answers of the target user on the same error question, so as to remove the questions that have been mastered by the target user from the error question set). Thus, in some embodiments, step A1 is preceded by the further step of:

The correct answer rate of each wrong question in the wrong question set of the target user is periodically obtained according to a preset period (which can be set according to actual needs), if the correct answer rate of the wrong questions exceeds a preset correct rate threshold value, the wrong questions are removed from the wrong question set.

In some embodiments, step A2 comprises:

acquiring knowledge points related to each wrong question in a wrong question set of a target user, and marking the knowledge points as related knowledge points;

counting the total number of occurrence of each related knowledge point in the wrong question set of the target user, and recording the total number as a first number;

counting the total number of occurrence of each related knowledge point in the historical question set, and recording the total number as a second number;

The knowledge points related to each question in the question bank can be predetermined and stored in the local database, and in step A2, the related knowledge points can be directly inquired from the local database according to index information such as the number of the wrong question, the question stem, the keywords and the like. The knowledge points can also be determined according to the pre-trained neural network model and the problem stems of the wrong questions, for example, the neural network model is a Bert (Bidirectional Encoder Representations from Transformers) model.

For example, if N questions in the questions set relate to the relevant knowledge point a, the first number of times of the relevant knowledge point a is N. If M topics in the history topic collection relate to the related knowledge point A, the second time of the related knowledge point A is M.

The step of calculating the grasping degree of each relevant knowledge point by the target user according to the first times and the second times comprises the following steps:

the grasping degree of each relevant knowledge point by the target user is calculated according to the following formula:

；

In other embodiments, step A2 comprises:

acquiring knowledge points related to each wrong question in a wrong question set of a target user, and recording the knowledge points as related knowledge points (refer to the previous for a specific process);

extracting the answer times and answer error times of questions related to each relevant knowledge point in the history question set;

calculating the total triggering times and the triggering error times of each relevant knowledge point according to the answer times and the answer error times;

and calculating the grasping degree of the target user on each relevant knowledge point according to the total triggering times and the triggering error times.

The question information further comprises the number of times of answering and the number of times of answering errors of the corresponding questions by the user, so that the number of times of answering and the number of times of answering errors of the corresponding questions can be directly extracted from the question information of the questions related to at least one relevant knowledge point in the history question set.

The total triggering times of a related knowledge point are the sum of the answering times of all the questions related to the related knowledge point in the history question set, and the triggering error times of a related knowledge point are the sum of the answering error times of all the questions related to the related knowledge point in the history question set. For example, if the questions related to the relevant knowledge point B in the history question set have the questions Q1 and Q2, the number of questions answered by the question Q1 and the number of questions answered by the answer error are respectively t1 and f1, the number of questions answered by the question Q2 and the number of questions answered by the answer error are respectively t2 and f2, and the total trigger number of the relevant knowledge point B is t1+t2, and the trigger number of errors is f1+f2.

The step of calculating the grasping degree of the target user on each relevant knowledge point according to the total triggering times and the triggering error times comprises the following steps:

；

wherein ,for the total number of triggers of the ith relevant knowledge point, +. >The number of trigger errors for the i-th relevant knowledge point.

As the target user repeatedly exercises each question, the grasping degree of the target user on the relevant knowledge points is improved, the grasping degree is calculated through the total triggering times and the triggering error times of each relevant knowledge point, the grasping degree is correspondingly changed along with the increasing times of the repeated questions of the target user on the corresponding questions, and compared with the grasping degree calculated through the occurrence times only, the grasping degree of the target user on the relevant knowledge points can be more accurately reflected.

In a first embodiment, step A3 comprises:

taking other users of which knowledge points related to the wrong question set comprise all relevant knowledge points as first candidate users;

calculating a first matching degree of each first candidate user and the target user according to the grasping degree of each target user on each relevant knowledge point and the grasping degree of each first candidate user on each relevant knowledge point;

and selecting at least one first candidate user as a reference user according to the first matching degree.

Wherein the other users are other than the target user. For example, if the relevant knowledge points include knowledge point a and knowledge point B, the knowledge points related to the wrong set of user a1 include knowledge point a, knowledge point B, and knowledge point C, and the knowledge points related to the wrong set of user a2 include knowledge point a, knowledge point C, and knowledge point D, then since the knowledge points related to the wrong set of user a1 include knowledge point a and knowledge point B, and the knowledge points related to the wrong set of user a2 do not include knowledge point B, user a1 is a first candidate user, and user a2 is not a first candidate user.

The calculation mode of the grasping degree of each relevant knowledge point by the first candidate user is the same as that of the grasping degree of each relevant knowledge point by the target user.

The first grasping degree arrays are formed by grasping degrees of the target users on the relevant knowledge points, the second grasping degree arrays are formed by grasping degrees of the first candidate users on the relevant knowledge points, and then the similarity between the second grasping degree arrays and the first grasping degree arrays is calculated to be used as the first matching degree between the first candidate users and the target users. Or after the similarity is calculated, calculating the first matching degree of each first candidate user and the target user according to the difference value between the number of knowledge points related to the wrong question set of each first candidate user and the number of related knowledge points and the corresponding similarity.

For example, according to the difference between the number of knowledge points related to the mistopic set of each first candidate user and the number of related knowledge points and the corresponding similarity, the step of calculating the first matching degree between each first candidate user and the target user includes:

carrying out normalization processing on the difference value between the number of knowledge points related to the wrong question set of each first candidate user and the number of related knowledge points to obtain a normalized difference value;

obtaining correction parameters of each first candidate user according to the normalized difference value;

and calculating the first matching degree of each first candidate user and the target user by using the correction parameters and the corresponding similarity of each first candidate user.

And dividing the difference value between the number of knowledge points related to the wrong question set of each first candidate user and the number of related knowledge points by the maximum value in the differences when normalization processing is carried out, so as to obtain a corresponding normalized difference value.

When obtaining the correction parameters of each first candidate user according to the normalized difference value, the corresponding preset correction parameters may be obtained according to the normalized difference value range in which the normalized difference value falls (corresponding preset correction parameters are set for different normalized difference value ranges in advance), or the normalized difference value is substituted into a preset correction parameter calculation formula to calculate the corresponding correction parameters (the correction parameter calculation formula may be set according to actual needs, for example, w=k (1- Δe) +l, where W is the correction parameter, k and L are preset parameters, and Δe is the normalized difference value, but the correction parameter calculation formula is not limited thereto).

The correction parameters of the first candidate user can be multiplied by the corresponding similarity to obtain a first matching degree of the first candidate user and the target user; or, calculating the weighted sum of the correction parameters and the corresponding similarity of the first candidate user to obtain the first matching degree of the first candidate user and the target user.

Specifically, when at least one first candidate user is selected as a reference user according to the first matching degree, firstly, sorting all the first candidate users in a descending order according to the first matching degree, selecting the first candidate users m before sorting as the reference users, wherein m is a preset positive integer (m is more than or equal to 1, and can be specifically set according to actual needs). Or when at least one first candidate user is selected as a reference user according to the first matching degree, selecting the first candidate user with the first matching degree larger than a preset matching degree threshold (which can be set according to actual needs) as the reference user.

In practical application, the number of relevant knowledge points may be more (i.e. the number of knowledge points related to the wrong question set of the target user is more), if other users, including all relevant knowledge points, of the knowledge points related to the wrong question set are used as first candidate users and then reference users are selected from the first candidate users, the first candidate users without symbol conditions are easy to cause, or the number of first candidate users meeting the conditions is too small, and further the selection of the reference users fails or is insufficient, so that effective question recommendation cannot be performed subsequently.

For this purpose, in a second embodiment, step A3 comprises:

screening the relevant knowledge points according to the grasping degree of the target user on the relevant knowledge points so as to screen effective knowledge points;

taking other users of which knowledge points related to the wrong question set comprise all effective knowledge points as second candidate users;

calculating a second matching degree of each second candidate user and the target user according to the grasping degree of each effective knowledge point by the target user and the grasping degree of each effective knowledge point by each second candidate user;

and selecting at least one second candidate user as a reference user according to the second matching degree.

Effective knowledge points are screened out firstly, and then second candidate users are selected according to the effective knowledge points, so that more second candidate users can be obtained, and the reference users can be screened out more reliably. In addition, the calculated amount can be reduced, and the processing efficiency can be improved.

For example, when screening for effective knowledge points, relevant knowledge points with a grasping degree smaller than a preset grasping degree threshold (which can be set according to actual needs) can be used as effective knowledge points. When the grasping degree of a certain relevant knowledge point exceeds the grasping degree threshold, it is considered that the target user has already learned about the relevant knowledge point, and thus it is possible not to be a valid knowledge point.

Or for example, when screening the effective knowledge points, the relevant knowledge points can be sorted in a descending order according to the grasping degree, then the relevant knowledge points of x before sorting are selected as the effective knowledge points, and x is a preset positive integer greater than 1 and can be set according to the implementation requirement. That is, a plurality of relevant knowledge points with low mastery degree are selected as effective knowledge points.

For example, when screening the effective knowledge points, calculating the comprehensive influence degree of each relevant knowledge point according to the ratio of the number of the wrong questions related to each relevant knowledge point in the wrong question set of the target user to the total number of the wrong questions in the wrong question set of the target user and the grasping degree of each relevant knowledge point, and then selecting the relevant knowledge points with the comprehensive influence degree exceeding a preset influence degree threshold (which can be set according to actual needs) or with the comprehensive influence degree being larger than y (namely, after descending order according to the comprehensive influence degree, arranging y values of y, wherein y is a preset positive integer larger than 1, which can be set according to implementation needs) as the effective knowledge points. The larger the duty ratio, the more significant the question answering error of the target user is caused by the relevant knowledge point, and the more the relevant knowledge point is supposed to be used as the effective knowledge point.

For example, if the number of questions related to the relevant knowledge point a in the set of questions related to the target user is c1 and the total number of questions related to the relevant knowledge point a in the set of questions related to the target user is z1, the ratio of the number of questions related to the relevant knowledge point a in the set of questions related to the target user to the total number of questions related to the set of questions related to the target user is c1/z1.

At the time of calculating the degree of the composite influence, a weighted sum of the duty ratio and the grasping degree may be calculated as the degree of the composite influence.

Assuming that the valid knowledge points include a knowledge point a and a knowledge point B, the knowledge points related to the wrong question set of the user B1 include a knowledge point a, a knowledge point B and a knowledge point C, and the knowledge points related to the wrong question set of the user B2 include a knowledge point a, a knowledge point C and a knowledge point D, since the knowledge points related to the wrong question set of the user B1 include a knowledge point a and a knowledge point B, and the knowledge points related to the wrong question set of the user B2 do not include a knowledge point B, the user B1 is a second candidate user, and the user B2 is not a second candidate user.

The process of calculating the second matching degree can refer to the process of calculating the first matching degree, and the difference between the two processes is that the grasping degree of each relevant knowledge point is used for calculation when the first matching degree is calculated, and the grasping degree of each effective knowledge point is used for calculation when the second matching degree is calculated.

The process of selecting at least one second candidate user as the reference user according to the second matching degree may refer to the process of selecting at least one first candidate user as the reference user according to the first matching degree, which is not described herein.

In practice, one of the two embodiments may be selected to be used to screen the reference user according to the number of relevant knowledge points (e.g., when the number of relevant knowledge points is less than a preset number threshold, the first embodiment is selected, otherwise the second embodiment is selected); the first embodiment may be used to screen the reference users, and if the screening fails (i.e. the reference users meeting the condition are not screened) or the number of the screened reference users is insufficient, the second embodiment is used to screen the reference users, and then the results of the two screening are integrated to be the final screening result of the reference users.

In step A4, all questions not included in the error question set of the target user in the error question set of the reference user may be used as recommended questions to be recommended to the target user. However, the number of recommended topics is too large, which can reduce the training effect of the target user on weak knowledge points.

To this end, in some embodiments, step A4 comprises:

A401. taking a relevant knowledge point with the grasping degree of the target user lower than a preset grasping degree threshold value (which can be set according to actual needs) as a target knowledge point;

A402. and screening topics which relate to the target knowledge point and are not included in the error topic set of the target user from the error topic set of the reference user as recommended topics so as to be recommended to the target user.

The relevant knowledge points are screened by setting the mastery degree threshold value, so that relevant knowledge points with lower mastery degree are screened out to serve as target knowledge points, the target knowledge points are weak knowledge points which are urgently required to be subjected to intensive training, and the subjects are recommended aiming at the target knowledge points, so that pertinence to the weak knowledge points can be improved, and the training effect is improved.

In some embodiments, step A4 comprises:

A403. acquiring a related knowledge point combination and a corresponding error-causing association degree of error questions related to at least two related knowledge points in an error question set or a total error question set of a target user; the error-induced association represents the significance degree of the answer errors of the target user, which can be caused by the corresponding relevant knowledge point combination; the total error question set is a set formed by error question sets of the target user and all reference users;

A404. And screening recommended topics from the wrong topic set of the reference user according to the related knowledge point combination and the corresponding error-causing association degree so as to recommend the recommended topics to the target user.

Sometimes, when a question involves only a single knowledge point, the target user may easily answer correctly, but when the knowledge point and other knowledge points appear in one question at the same time, the target user may not answer correctly. The error-causing association degree corresponding to the relevant knowledge point combination represents the significance degree (probability) of the answering error of the target user when a plurality of relevant knowledge points in the relevant knowledge point combination are simultaneously presented in one question. And performing question recommendation according to the error-causing association degree of various combinations of the related knowledge points, so that training of a target user aiming at the combinations of different weak knowledge points can be enhanced, and the training effect is further improved.

Wherein the combination of related knowledge points comprises a combination of two related knowledge points or further comprises a combination of more related knowledge points. For example, when the relevant knowledge points related to a topic include the relevant knowledge point a and the relevant knowledge point B, the relevant knowledge points of the topic are combined into [ relevant knowledge point a, relevant knowledge point B ], when the relevant knowledge points related to a topic include the relevant knowledge point a, relevant knowledge point B and relevant knowledge point C, the relevant knowledge points of the topic are combined into [ relevant knowledge point a, relevant knowledge point B, relevant knowledge point C ], and so on.

Wherein, step a403 may include:

s1, extracting all related knowledge point combinations of wrong questions related to at least two related knowledge points in a wrong question set or a total wrong question set of a target user;

s2, eliminating repeated relevant knowledge point combinations from the extracted relevant knowledge point combinations (specifically, if a plurality of same relevant knowledge point combinations exist, one of the relevant knowledge point combinations is reserved, and other same relevant knowledge point combinations are eliminated);

s3, sequentially taking the rest relevant knowledge point combinations as target knowledge point combinations, counting the total number of questions related to the target knowledge point combinations from a historical question set or a total historical question set of a target user as a first total number, and counting the total number of questions related to the target knowledge point combinations from an error question set or a total error question set of the target user as a second total number; the total historical topic collection is a collection formed by the historical topic collection of the target user and all the reference users;

s4, calculating the ratio of the second total number in the first total number, and taking the ratio as the error-causing association degree of the target knowledge point combination.

When the present relevant knowledge point combination contains other relevant knowledge point combinations (for example, all relevant knowledge points of the relevant knowledge point combination Z1 belong to relevant knowledge points of the relevant knowledge point combination Z2, the relevant knowledge point combination Z2 is considered to contain the relevant knowledge point combination Z1), the other relevant knowledge point combinations contained in the present relevant knowledge point combination are referred to as sub-combinations of the present relevant knowledge point combination, and a sub-combination containing the closest relevant knowledge point number to the present relevant knowledge point combination among all sub-combinations of the present relevant knowledge point combination is referred to as a lower-level combination of the present relevant knowledge point combination (when there is only one sub-combination, the sub-combination is a lower-level combination of the present relevant knowledge point combination, and when there is a plurality of sub-combinations, one sub-combination containing the closest relevant knowledge point number to the present relevant knowledge point combination is referred to as a lower-level combination of the present relevant knowledge point combination). In general, questions involving a larger number of related knowledge points are more prone to answer questions errors, so that the error-causing association degree of a related knowledge point combination should be greater than that of a next-level combination, otherwise, it is indicated that the influence of related knowledge points except for the next-level combination in the related knowledge point combination on the target user is smaller, and the root cause of the error-causing is the next-level combination, and at this time, after step S4, the method further includes the steps of:

S5, if the error-causing association degree of the related knowledge point combination is not greater than that of the lower-level combination, taking the related knowledge point combination as a removing object and taking the lower-level combination of the removing object as a correction object;

s6, correcting the error incidence of the correction object according to the first total number and the second total number corresponding to the removal object and the correction object;

s7, eliminating the eliminating object and error-caused association degree thereof.

For example, in step S6, the error-causing correlation of the correction object is corrected by the following formula:

；

wherein ,error-causing association degree after correction for correction object, +.>For correcting the second total number corresponding to the object, +.>For eliminating the second total number corresponding to the object, +.>For correcting the first total number corresponding to the object, +.>And eliminating the first total number corresponding to the object. However, the specific correction method is not limited thereto. By correcting the error-causing association degree of the correction object, the accuracy of the error-causing association degree can be improved. The steps S5-S7 can be repeatedly executed until the error-generating association degree of the combination without the relevant knowledge points is not greater than the error-generating association degree of the combination at the lower level.

In step a404, the relevant knowledge point combination with the error association degree greater than the preset association degree threshold (which can be set according to actual needs) may be used as an effective relevant knowledge point combination, and the questions related to the effective relevant knowledge point combination and not included in the error question set of the target user may be selected from the error question set of the reference user as recommended questions to be recommended to the target user.

It should be noted that one of the two embodiments may be selected to filter the recommended topics for recommendation to the target user (i.e. only performing steps a401-a402 or only performing steps a403-a 404); both of the above embodiments may also be used to screen recommended topics (i.e., steps a401-a404 are performed).

When the method for recommending the topics based on the electronic student identity data is applied to the server, recommendation information is sent to the electronic student identity of the target user to achieve recommendation of the recommended topics, and the recommendation information comprises at least one item of index information such as a number, a stem and a keyword of the recommended topics.

According to the method, the historical question set and the corresponding wrong question set of the target user are obtained; determining the knowledge of the related knowledge points and the target user according to the historical question set and the wrong question set; determining other users as reference users of the target users according to the related knowledge points and the corresponding mastery degree; according to the related knowledge points and the corresponding mastery degree, a recommended question is screened from the wrong question set of the reference user so as to be recommended to the target user; therefore, the topics can be recommended to the weak points of the knowledge of the user more effectively, and the exercise effect is improved.

Referring to fig. 2, the application provides a topic recommendation device based on electronic student status data, comprising:

the first acquisition module 1 is used for acquiring a historical question set and a corresponding wrong question set of a target user;

the grasping degree determining module 2 is used for determining grasping degrees of the related knowledge points and the target user on the related knowledge points according to the historical question set and the wrong question set;

a reference user determining module 3, configured to determine, according to the relevant knowledge points and the corresponding mastery degrees, other users as reference users of the target users;

and the screening module 4 is used for screening the recommended questions from the wrong question set of the reference user according to the related knowledge points and the corresponding mastery degree so as to recommend the recommended questions to the target user.

The topic recommendation device based on the electronic student identity data can be applied to a server, the server is in communication connection with the electronic student identity of a target user, the target user logs in the server through answering software of the electronic student identity, and the server acquires a historical topic set and a corresponding wrong topic set of the target user according to account information sent when the electronic student identity logs in. The historical question set is a set of questions made by a user, and the historical question set records question information of the questions made by the user; the wrong question set is a set of missed questions made by the user, and the wrong question set records question information of the missed questions made by the user. The topic information comprises at least one item of index information such as topic number, topic stem, keywords and the like.

The error question set of the target user may be manually sorted by the target user (e.g., the target user manually removes the questions that have been mastered from the error question set) or periodically and automatically sorted by the server (e.g., the mastering degree of the target user on the error question is determined according to the accuracy of repeated answers of the target user on the same error question, so as to remove the questions that have been mastered by the target user from the error question set). Thus, in some embodiments, the topic recommendation device based on the electronic student status data further comprises:

The arrangement module is used for periodically obtaining the correct answer rate (the correct answer times of the same wrong questions are divided by the total answer times) of each wrong question in the wrong question set by a target user according to a preset period (which can be set according to actual needs), and if the correct answer rate of the wrong questions exceeds a preset correct rate threshold, the wrong questions are removed from the wrong question set.

In some embodiments, the grasping level determining module 2 performs, when determining grasping levels of the relevant knowledge points and the target user on the relevant knowledge points from the history topic set and the wrong topic set:

The knowledge points related to each question in the question bank can be predetermined and stored in the local database, and the grasping degree determining module 2 can directly query the related knowledge points from the local database according to index information such as the number of the wrong question, the question stem, the keyword and the like. The knowledge points can also be determined according to the pre-trained neural network model and the problem stems of the wrong questions, for example, the neural network model is a Bert (Bidirectional Encoder Representations from Transformers) model.

The method for calculating the grasping degree of the target user on each relevant knowledge point according to the first times and the second times specifically comprises the following steps:

；

In other embodiments, the grasping level determining module 2 performs, when determining the grasping level of the relevant knowledge point and the target user to the relevant knowledge point from the history topic set and the wrong topic set:

The method for calculating the grasping degree of the target user on each relevant knowledge point according to the total triggering times and the triggering error times specifically comprises the following steps:

；

wherein ,for the total number of triggers of the ith relevant knowledge point, +.>The number of trigger errors for the i-th relevant knowledge point.

In the first embodiment, the reference user determination module 3 performs, when determining the other user as the reference user of the target user based on the relevant knowledge points and the corresponding grasping degree:

For example, according to the difference between the number of knowledge points related to the mistopic set of each first candidate user and the number of related knowledge points and the corresponding similarity, calculating the first matching degree between each first candidate user and the target user specifically includes:

For this reason, in the second embodiment, the reference user determination module 3 performs, when determining other users as reference users of the target users based on the relevant knowledge points and the corresponding mastery degrees:

The screening module 4 screens the recommended questions from the wrong question set of the reference user according to the relevant knowledge points and the corresponding mastery degree, so that when the recommended questions are recommended to the target user, all questions which are not contained in the wrong question set of the target user in the wrong question set of the reference user can be used as recommended questions to be recommended to the target user. However, the number of recommended topics is too large, which can reduce the training effect of the target user on weak knowledge points.

To this end, in some embodiments, the screening module 4 performs, when screening the recommended topics from the wrong topic set of the reference user according to the relevant knowledge points and the corresponding mastery degree, to recommend to the target user:

In some embodiments, the screening module 4 performs, when screening the recommended topics from the wrong topic set of the reference user according to the relevant knowledge points and the corresponding mastery degree, to recommend the recommended topics to the target user:

Wherein, step a403 may include:

；

wherein ,error-causing association degree after correction for correction object, +.>For correcting the second total number corresponding to the object, +.>For eliminating the second total number corresponding to the object, +.>For correcting the correspondence of the objectFirst total number->And eliminating the first total number corresponding to the object. However, the specific correction method is not limited thereto. By correcting the error-causing association degree of the correction object, the accuracy of the error-causing association degree can be improved. The steps S5-S7 can be repeatedly executed until the error-generating association degree of the combination without the relevant knowledge points is not greater than the error-generating association degree of the combination at the lower level.

When the topic recommendation device based on the electronic student identity data is applied to a server, recommendation information is sent to the electronic student identity of a target user to achieve recommendation of recommended topics, and the recommendation information comprises at least one item of index information such as a number, a topic stem, a keyword and the like of the recommended topics.

From the above, the topic recommendation device based on the electronic student status data obtains the historical topic collection and the corresponding wrong topic collection of the target user; determining the knowledge of the related knowledge points and the target user according to the historical question set and the wrong question set; determining other users as reference users of the target users according to the related knowledge points and the corresponding mastery degree; according to the related knowledge points and the corresponding mastery degree, a recommended question is screened from the wrong question set of the reference user so as to be recommended to the target user; therefore, the topics can be recommended to the weak points of the knowledge of the user more effectively, and the exercise effect is improved.

Referring to fig. 3, fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present application, where the electronic device includes: processor 301 and memory 302, the processor 301 and memory 302 being interconnected and in communication with each other by a communication bus 303 and/or other form of connection mechanism (not shown), the memory 302 storing a computer program executable by the processor 301, the processor 301 executing the computer program when the electronic device is running to perform the method of question recommendation based on electronic student status data in any of the alternative implementations of the above embodiments to perform the following functions: acquiring a historical question set and a corresponding wrong question set of a target user; determining the knowledge of the related knowledge points and the target user according to the historical question set and the wrong question set; determining other users as reference users of the target users according to the related knowledge points and the corresponding mastery degree; and screening recommended questions from the wrong question set of the reference user according to the related knowledge points and the corresponding mastery degree so as to recommend the recommended questions to the target user.

The embodiment of the application provides a computer readable storage medium, on which a computer program is stored, which when executed by a processor, performs the topic recommendation method based on electronic student status data in any of the alternative implementations of the above embodiment, so as to implement the following functions: acquiring a historical question set and a corresponding wrong question set of a target user; determining the knowledge of the related knowledge points and the target user according to the historical question set and the wrong question set; determining other users as reference users of the target users according to the related knowledge points and the corresponding mastery degree; and screening recommended questions from the wrong question set of the reference user according to the related knowledge points and the corresponding mastery degree so as to recommend the recommended questions to the target user. The computer readable storage medium may be implemented by any type or combination of volatile or non-volatile Memory devices, such as static random access Memory (Static Random Access Memory, SRAM), electrically erasable Programmable Read-Only Memory (EEPROM), erasable Programmable Read-Only Memory (Erasable Programmable Read Only Memory, EPROM), programmable Read-Only Memory (PROM), read-Only Memory (ROM), magnetic Memory, flash Memory, magnetic disk, or optical disk.

In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other manners. The above-described apparatus embodiments are merely illustrative, for example, the division of the units is merely a logical function division, and there may be other manners of division in actual implementation, and for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some communication interface, device or unit indirect coupling or communication connection, which may be in electrical, mechanical or other form.

Further, the units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

Furthermore, functional modules in various embodiments of the present application may be integrated together to form a single portion, or each module may exist alone, or two or more modules may be integrated to form a single portion.

In this document, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions.

The above description is only an example of the present application and is not intended to limit the scope of the present application, and various modifications and variations will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. The title recommendation method based on the electronic student identity data is characterized by comprising the following steps:

A4. according to the related knowledge points and the corresponding mastery degree, a recommended question is screened from the wrong question set of the reference user so as to be recommended to the target user;

The step A2 comprises the following steps:

calculating the grasping degree of the target user on each relevant knowledge point according to the first times and the second times;

the step of calculating the grasping degree of the target user on each relevant knowledge point according to the first times and the second times comprises the following steps:

；

wherein ,degree of mastery of the i-th relevant knowledge point for the target user, +.>For the first number of i-th relevant knowledge points, +.>The second times of the ith relevant knowledge point, n is the total number of relevant knowledge points;

step A4 includes:

According to the related knowledge point combination and the corresponding error-causing association degree, a recommendation question is screened from the error question set of the reference user so as to be recommended to the target user;

the step of obtaining the related knowledge point combination and the corresponding error-induced relevance of the error questions related to at least two related knowledge points in the error question set or the total error question set of the target user comprises the following steps:

s2, eliminating repeated relevant knowledge point combinations from the extracted relevant knowledge point combinations;

s4, calculating the ratio of the second total number in the first total number, and taking the ratio as the error-causing association degree of the target knowledge point combination;

s7, eliminating the eliminating object and error-caused association degree thereof;

in step S6, the error-causing correlation of the correction object is corrected by the following formula:

；

wherein ,error-causing association degree after correction for correction object, +.>For correcting the second total number corresponding to the object, +.>For eliminating the second total number corresponding to the object, +.>For correcting the first total number corresponding to the object, +.>And eliminating the first total number corresponding to the object.

2. The method for question recommendation based on electronic student status data as claimed in claim 1, wherein the step A3 comprises:

3. The method for question recommendation based on electronic student status data as claimed in claim 1, wherein the step A3 comprises:

4. The method of claim 1, wherein step A4 further comprises:

5. The title recommendation device based on the electronic student identity data is characterized by comprising:

the screening module is used for screening recommended questions from the wrong question set of the reference user according to the related knowledge points and the corresponding mastery degree so as to recommend the recommended questions to the target user;

the grasping degree determining module executes when determining grasping degrees of related knowledge points and the target user on the related knowledge points according to the history question set and the wrong question set:

calculating the grasping degree of the target user on each relevant knowledge point according to the first times and the second times, wherein the grasping degree specifically comprises the following steps:

；

the screening module is used for screening recommended questions from the wrong question set of the reference user according to the related knowledge points and the corresponding mastery degree so as to be recommended to the target user, and executing the following steps:

the method for obtaining the related knowledge point combination and the corresponding error-causing association degree of the error questions related to at least two related knowledge points in the error question set or the total error question set of the target user specifically comprises the following steps:

；

6. An electronic device comprising a processor and a memory, the memory storing a computer program executable by the processor, when executing the computer program, running the steps in the method for question recommendation based on electronic student status data as claimed in any one of claims 1 to 4.

7. A computer readable storage medium, on which a computer program is stored which, when being executed by a processor, performs the steps in the method for question recommendation based on electronic student status data as claimed in any one of claims 1 to 4.