Intelligent pen and paper-based mistake question recommendation method and device, and storage medium

Abstract

The application discloses a wrong question recommendation method based on intelligent paper and pen, obtains the identification number of a question corresponding to a student's answer area, records the writing content of the student in real time when the student answers questions, compares the writing content with the answer of the question with the identification number after the student finishes answering questions, judges the student's answer result, and if the answer is wrong, the identification number is recorded in a wrong question set; the content of the homework of the day and the student's exercise expectation are input into a recommendation module, the recommendation module extracts corresponding questions from the wrong question set and recommends the questions to the student as pre-exercise content, and the recommendation module also recommends the required knowledge points related to the content of the homework of the day according to the past exercise situation. The application has the beneficial effect that the deconstruction module of the application can analyze and structure the questions in the wrong question set according to the error type, obtain fast-answer pre-exercise, the student can quickly review and check the corresponding knowledge points, and the system can also more quickly judge the student's mastery of the knowledge points, thereby providing a reference basis for subsequent recommendation.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of educational internet services, in particular to a wrong question recommendation method and device based on intelligent pen and paper and a storage medium. BACKGROUND

[0002] The wrong question set is formed by collecting and sorting the wrong questions in the learning process of students. The traditional wrong question set needs to be manually copied by students, which is time-consuming. At the same time, it is troublesome to flip through due to the lack of indexing mechanism. Moreover, the questions and answer analysis in the wrong question set are usually written in, and the answer will be seen involuntarily during review, so the review effect is general. Moreover, when the number of wrong question sets is large, it is not convenient to simplify, and it is very inconvenient to review. For example, ZL201911022072.5 "Question recommendation method, device and terminal equipment" can conveniently organize the wrong question set, but it does not have a recommendation function and cannot efficiently review the wrong questions.

[0003] Now some schools begin to use intelligent pen and paper system. The intelligent pen can collect the homework information written by students and display it on the designated device, which can conveniently and quickly analyze the written content and record the wrong questions to form a wrong question set. However, for science questions, it may take ten minutes to completely answer a wrong question. If the foundation is not solid enough, even if it is done again, it cannot play a good review effect. For some students with solid foundation but carelessness, they will jump to do the questions due to seeing the analysis, which cannot cultivate the habit of careful doing. For some students who are not familiar with the problem solving skills, they actually need help to read the question and find the implicit conditions, rather than simply repeating the questions. How to efficiently use the science wrong question set in the intelligent pen and paper system to consolidate the learning effect is still a technical problem to be solved. SUMMARY

[0004] In view of the above problems, the present application aims to provide a wrong question recommendation method, device and storage medium based on intelligent pen and paper, which can improve the review and practice effect.

[0005] To achieve the technical purpose, the scheme of the present application is as follows: a wrong question recommendation method, comprising:

[0006] The writing content of daily homework is obtained through the pen and paper module, and the writing content is uploaded to the analysis module of the server for analysis and comparison. After the analysis is completed, the wrong question set is obtained;

[0007] The current homework content and practice expectation value are input into the recommendation module, the corresponding questions in the wrong question set are extracted by the recommendation module and input into the deconstruction module, and the deconstruction module generates pre-practice in a specified form;

[0008] The writing content of pre-practice is obtained through the pen and paper module, and the writing content is uploaded to the server for analysis.

[0009] As preferred, the error collection and the student's answering process are input into the analysis module, and the questions are marked according to the error types;

[0010] Meanwhile, the knowledge points and the examination problems corresponding to the questions are analyzed to obtain the knowledge labels corresponding to the questions.

[0011] As preferred, the recommendation module is used to compare the current homework content with the error collection, and when the similarity of the knowledge points and the examination problems of the homework content and the questions in the error collection reaches a threshold value, the question enters the candidate list of the deconstruction module;

[0012] The deconstruction module selects the corresponding question as the pre-exercise content from the candidate list according to the student's exercise expectation and the reproduction weight;

[0013] Meanwhile, the corresponding knowledge points are given as the required knowledge points according to the error type of the question.

[0014] As preferred, the deconstruction module is used to decompose and reconstruct the questions in the error collection, and when the knowledge points of the question are greater than or equal to two and the knowledge point blind area corresponding to the error is less than or equal to two, the deconstruction module extracts the corresponding knowledge points according to the error type to judge the question or fill-in-the-blank question to reconstruct the fast-answer pre-exercise;

[0015] The multi-dimensional answering data of the fast-answer pre-exercise are obtained by analyzing the writing content, and the reproduction rate and the reproduction mode of the corresponding knowledge points of the pre-exercise are adjusted according to the multi-dimensional answering data.

[0016] As preferred, the TF-IDF algorithm or the Euclidean distance is used to calculate the similarity of the knowledge points of the homework content and the questions in the error collection;

[0017] The questions in the error collection are sorted according to the similarity to list the candidate question list for the deconstruction module.

[0018] As preferred, the selected question and its analysis and answering content are input into the deconstruction module, and the deconstruction process is as follows:

[0019] S1, the corresponding basic knowledge points and the examination focuses are obtained through the knowledge labels of the corresponding science subjects of the question, and the basic knowledge points and the examination focuses are one-to-one corresponding to the stems in the question;

[0020] S2, the error types of the students are analyzed and predicted by comparing the answering content with the analysis, and each error type is assigned a weight of the corresponding reproduction rate;

[0021] S3, the corresponding content in the stem and the basic knowledge points are extracted according to the error type to produce the fast-answer pre-exercise;

[0022] The fast-answer pre-exercise is one of a judgment question, a selection question or a fill-in-the-blank question, and the fast-answer pre-exercise is presented through a screen and a corresponding countdown is set;

[0023] Each fast-answer pre-exercise contains only two or less basic knowledge points and / or examination points;

[0024] S4. According to the completion degree of the previous fast-answer pre-exercise, the weight of the recurrence rate of the corresponding error type in the fast-answer pre-exercise is adjusted.

[0025] S5. When the correct rate of the fast-answer pre-exercise is lower than a threshold value, a fast-answer pre-exercise with lower difficulty is supplemented.

[0026] A wrong question recommendation device, comprising:

[0027] A recommendation module is configured to recommend the pre-exercise content, the fast-answer pre-exercise and the required knowledge points related to the wrong question set to the student.

[0028] A paper and pen module is configured to compare the homework content with the wrong question set, and when the similarity of the knowledge points and the examination problems of the homework content and the questions in the wrong question set reaches a threshold value, the question enters the pre-exercise content candidate list.

[0029] A deconstruction module is configured to decompose and reconstruct the questions in the wrong question set, and the deconstruction module extracts the corresponding independent knowledge points according to the error type to reconstruct the fast-answer pre-exercise in the form of a judgment question or a fill-in-the-blank question.

[0030] A terminal device, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the method when executing the computer program.

[0031] A computer readable storage medium, the computer readable storage medium stores a computer program, wherein the computer program is executed by a processor to implement the method.

[0032] The beneficial effects of the present application are that the deconstruction module of the present application can analyze and structure the questions in the error question set according to the error types, obtain pre-exercises of quick answering, students can quickly review and test the corresponding knowledge points, and the system can also quickly judge the students' mastery of the knowledge points, thereby providing a reference basis for subsequent recommendation. At the same time, the pre-exercises of quick answering have a fast detection speed, can quickly help students consolidate the knowledge points in the error question set before review, and have a relatively small increase in homework burden. At the same time, students can also quickly review the knowledge points, thereby preparing for subsequent homework exercises today. The present application adopts the Euclidean distance calculation method, can quickly identify the questions in the error question set related to the content of the homework of the day, extract the questions as pre-exercise content, can better help students consolidate the content, and after the completion of the exercises, can quickly start the same type of questions in the subsequent homework content of the day, thereby effectively improving the review effect. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 a flowchart of the embodiment one of the present application;

[0034] Figure 2 a flowchart of the embodiment two of the present application;

[0035] Figure 3 a flowchart of the embodiment three of the present application;

[0036] Figure 4 a flowchart of the embodiment four of the present application. DETAILED DESCRIPTION

[0037] As Figures 1-4 , the present application will be further described in detail below in combination with the drawings and specific embodiments. The specific embodiments listed below are exemplary and not limiting, and the terms "include" and "have" and their common variants in the following specific embodiments are intended to cover non-exclusive inclusion. For example, the process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but can optionally further include steps or units not listed, or can optionally further include other steps or units inherent to the process, method, product or device. Any slight modification, equivalent replacement and improvement of the following specific embodiments according to the technical essence of the present application shall be included in the protection scope of the technical scheme of the present application.

[0038] Embodiment one

[0039] A wrong question recommendation method, as Figure 1 , the specific steps are as follows:

[0040] S101, using an intelligent pen with an infrared camera and dot matrix work paper, obtaining the identification number of the question corresponding to the student's answer area, recording the writing content of the student in real time when answering, comparing the writing content with the answer of the question with the identification number after the student finishes answering, determining the student's answer result, if the answer is wrong, the identification number is recorded in the wrong question set;

[0041] S102, input the daily homework content and student practice expectation into the recommendation module, the recommendation module extracts the corresponding question from the wrong question set and recommends it to the student as the pre-exercise content, and the recommendation module also recommends the required knowledge points related to the daily homework content according to the past practice;

[0042] If the student's answer to the question in the pre-exercise is correct and the completion time is less than the threshold, the recurrence weight of the question is reduced; if the answer to the question corresponding to the required knowledge point is wrong, the recurrence rate of the required knowledge point is increased;

[0043] S103, input the wrong question set and the student's answer process into the analysis module, and mark the question according to the error type;

[0044] Meanwhile, the knowledge points and the problems corresponding to the question are analyzed to obtain the knowledge label corresponding to the question.

[0045] S104, the recommendation module is provided with a deconstruction module, and the pen module is used to decompose and reconstruct the questions in the wrong question set, when the knowledge points of the question are greater than or equal to two and the blind area of the knowledge points corresponding to the wrong question is less than or equal to two, the deconstruction module will extract the corresponding independent knowledge points according to the error type to judge the question or fill in the blank to reconstruct the fast answer pre-exercise;

[0046] When the required fast answer pre-exercise is answered correctly, the recurrence rate of the knowledge point corresponding to the fast answer pre-exercise is reduced.

[0047] S105, the recommendation module records the completion rate and accuracy of the student for the pre-exercise content and the fast answer pre-exercise, and predicts the influence factor of the required knowledge point on the homework content;

[0048] Adjust the composition ratio of the pre-exercise content and the fast answer pre-exercise according to the completion rate and the accuracy, and adjust the appearance frequency of the required knowledge point according to the influence factor.

[0049] The deconstruction module of the application can analyze and structure the questions in the error question set according to the error type, obtain fast-answer pre-exercises, students can quickly review and test the corresponding knowledge points, and the system can also quickly judge the students' mastery of the knowledge points, providing a reference basis for subsequent recommendations. At the same time, the fast-answer pre-exercise has a fast detection speed and can quickly help students consolidate and review the knowledge points in the error question set before, and the homework burden is also relatively small. At the same time, students can also quickly review the knowledge points and prepare for the subsequent homework exercises today.

[0050] Embodiment two

[0051] An error question recommendation method, as Figure 2 , the specific steps are as follows:

[0052] S201, using an intelligent pen with an infrared camera and dot matrix exercise paper, obtaining the identification number of the question corresponding to the student's answer area, recording the writing content of the student in real time when answering, comparing the writing content with the answer of the question with the identification number after the student finishes answering, determining the student's answer result, if the answer is wrong, the identification number is entered into the error question set;

[0053] S202, input the homework content and student practice expectation into the recommendation module, the recommendation module extracts the corresponding questions from the error question set and recommends them to the student as pre-exercise content, and the recommendation module also recommends the required knowledge points related to the homework content according to the past practice;

[0054] S203, input the error question set and the student's answer process into the analysis module, mark the questions according to the error type; at the same time, analyze the knowledge points and examination problems corresponding to the questions, and obtain the knowledge tags corresponding to the questions.

[0055] S204, the paper and pen module is arranged in the recommendation module, and the paper and pen module is used for comparing the homework content with the error question set, when the similarity of the knowledge points and the examination problems of the homework content and the questions in the error question set reaches a threshold value, the question enters the pre-exercise content candidate list;

[0056] According to the student's practice expectation and the recurrence weight, select the corresponding question as the pre-exercise content from the candidate list;

[0057] At the same time, according to the error type of the question, the corresponding knowledge point is given as the required knowledge point.

[0058] S205, using the Euclidean distance calculation method to calculate the similarity of the knowledge points of the homework content and the questions in the error question set;

[0059] Using the Euclidean distance calculation method to calculate the similarity of the examination problems of the homework content and the questions in the error question set;

[0060] The Euclidean distance and the wrong question entry time are used to sort the candidate list.

[0061] The Euclidean distance calculation method is used in the application, the question in the wrong question set related to the daily homework content can be quickly identified, the question is extracted as the pre-exercise content, the student can be better helped to consolidate the content, after the exercise is completed, the same type of question in the subsequent daily homework content can be quickly answered, and the review effect can be effectively improved.

[0062] Embodiment three

[0063] S301, as Figure 3 The smart pen with a camera and the dot matrix exercise paper are used to obtain the identification number of the question corresponding to the answer area of the student, the writing content of the student when answering is recorded in real time, after the student completes the answer, the writing content is compared with the answer of the question with the identification number, the answer result of the student is determined, if the answer is wrong, the identification number and the answer entry are obtained, and the wrong question set is obtained;

[0064] S302, the exercise expectation is input to the recommendation module through the mobile phone, for example: consolidating basic knowledge points, or expanding related exercises, or deepening exercises, or challenging high-order knowledge points; meanwhile, the wrong question set is input into the recommendation module, the deconstruction module in the recommendation module screens the wrong questions meeting the expectation of the student according to the exercise expectation of the student, and then the wrong questions are deconstructed to produce the pre-exercise of the quick answer type;

[0065] S303, the student only needs to spend a few minutes to complete the pre-exercise by checking on the tablet (mobile phone or computer), the pre-exercise of the quick answer type is a judgment question or a fill-in-the-blank question, the pre-exercise can quickly check the understanding of the student on the knowledge point, can greatly improve the review frequency of the wrong question, and saves review time; when the student completes the pre-exercise through the tablet, the completion rate and the correct rate are fed back to the recommendation module, the recommendation module adjusts the prediction influence factor according to the completion rate and the correct rate of each knowledge point, and then adjusts the appearance frequency of each knowledge point, and improves the review efficiency;

[0066] S304, the recommendation module extracts the corresponding question from the wrong question set and recommends the question to the student as the pre-exercise content, and the recommendation module also recommends the required knowledge points related to the daily homework content according to the past exercise situation.

[0067] Many wrong questions are essay questions, or multiple choice questions, or a question contains multiple knowledge points. When the basic knowledge points are not firm enough, the understanding cannot be effectively and quickly achieved, and there is not enough time to redo each wrong question. Through the method of the application, the wrong question set related to the daily homework content can be quickly identified, the question is extracted as the pre-exercise content, the student can be better helped to consolidate the content, after the exercise is completed, the same type of question in the subsequent daily homework content can be quickly answered, and the review effect can be effectively improved. Figure 3As shown, the present application can generate new quick-answer pre-exercises for students to quickly practice through the recommendation module and the deconstruction module. Students can review the missing knowledge points through the exercises at a high frequency, and through the analysis of the completion rate and the accuracy rate, the knowledge points needed by the students can be predicted, and through the multiple-choice questions or fill-in-the-blank questions, the students are reminded of the different elements of the knowledge points. This method can help both weak students to review and consolidate knowledge points and good students to check for missing points. At the same time, the quick-answer response mode can effectively mobilize the enthusiasm of students, and the speed and accuracy of answering can be ranked, and students can compare with similar students with similar wrong questions and similar exercise expectations.

[0068] Embodiment Four

[0069] The existing error question set is not used much. On the one hand, some questions have small errors, and students may not want to repeat the practice; on the other hand, some errors are on the problem-solving skills, and if students still do old questions, they can easily find hidden conditions and do not have the effect of repeated practice; some students have a weak foundation, and for comprehensive questions, it takes a long time and requires a high foundation to answer after consolidating the basic knowledge points. Therefore, directly repeating old questions not only wastes time but also cannot achieve good practice effect.

[0070] At the same time, different knowledge points are needed for each homework, and if the corresponding knowledge points can be reviewed and consolidated through the questions in the error question set before doing the questions, the effect will be doubled.

[0071] Through the TF-IDF algorithm, suitable questions are selected, the questions and their analysis and answering content are input into the deconstruction module, and then the corresponding quick-answer pre-exercises are generated, and the specific steps are as follows:

[0072] S401, the corresponding basic knowledge points and examination points are obtained through the knowledge tags of the subject corresponding to the question, and the basic knowledge points and examination points are one-to-one corresponding to the stem in the question;

[0073] S402, through the comparison between the answering content and the analysis, the error type of the student is analyzed and predicted, and each error type is assigned a corresponding recurrence rate weight;

[0074] S403, according to the error type, the corresponding content in the stem and or basic knowledge points are extracted to produce quick-answer pre-exercises;

[0075] The question of the quick-answer pre-exercise is one of a judgment question, a multiple-choice question or a fill-in-the-blank question, and the quick-answer pre-exercise is presented through a screen and a corresponding countdown is set;

[0076] Each quick-answer pre-exercise only contains two or less basic knowledge points and or examination points;

[0077] For example: for example in the physical "magnetic field on the action of moving charge", according to the knowledge point involved in the current homework, several judgment questions can be generated. The questions are as follows (1) charged particles in the magnetic field will be affected by the magnetic field force. (2) according to the formula T = 2πr / v, the motion period T of charged particles in uniform magnetic field is inversely proportional to v. Students only need a few minutes to complete the judgment on the tablet (only 1-2 knowledge points), and can see the answer immediately, and consolidate the basic knowledge points.

[0078] S404, according to the completion degree of the previous fast answering pre-exercise, the weight of the recurrence rate of the corresponding error type in this fast answering pre-exercise is adjusted; if the judgment is wrong, the probability of the judgment, selection question related to this knowledge point appearing again next time is increased, and if the judgment is correct, the frequency of appearance is reduced;

[0079] S405, when the correct rate of this fast answering pre-exercise is lower than the threshold value, a fast answering pre-exercise with lower difficulty is supplemented, and the fast answering pre-exercise supplemented only contains one basic knowledge point and or examination point. After reducing the difficulty, the students with weak foundation can better practice because of the single knowledge point.

[0080] A wrong question recommendation device, comprising: a recommendation module, configured to recommend the wrong question set, the pre-exercise content related to this homework, the fast answering pre-exercise and the required knowledge point to the student;

[0081] A pen and paper module is configured to compare the homework content with the wrong question set, when the similarity of the knowledge points and the examination problems of the homework content and the questions in the wrong question set reaches the threshold value, the question enters the pre-exercise content candidate list; the pen and paper module includes a smart pen and a dot matrix paper, the smart pen is provided with a camera, and the smart pen can record the writing content on the dot matrix paper;

[0082] A deconstruction module is configured to decompose and reconstruct the questions in the wrong question set, the deconstruction module will extract the corresponding independent knowledge points according to the error type, and reconstruct the fast answering pre-exercise with judgment questions or fill-in-the-blank questions.

[0083] A terminal device, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to realize the method.

[0084] A computer readable storage medium, the computer readable storage medium stores a computer program, characterized in that the computer program is executed by the processor to realize the method.

[0085] In the embodiments of the present application, the size of the serial number of each process does not mean the inevitable sequence of execution order, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0086] The functional units in the embodiments of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be implemented in the form of hardware or software functional units.

[0087] When the functional units are implemented in the form of software and sold or used as an independent product, they can be stored in a storage medium accessible by a computer device. Therefore, the technical solutions of the present application, essentially or in part, or all or part of the technical solutions, can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes several steps for realizing the methods of the embodiments of the present application.

Claims

1. A method for recommending incorrect answers based on smart paper and pen, characterized in that, include: The paper-and-pen module obtains the writing content of daily homework, uploads the writing content to the analysis module on the server for analysis and comparison, and obtains a set of wrong questions after the analysis is completed; Input the current homework content and expected practice value into the recommendation module. The recommendation module extracts the corresponding questions from the wrong question set and inputs them into the deconstruction module. The deconstruction module generates pre-exercises in the specified format. The paper-and-pen module retrieves the writing content from the pre-practice exercises and uploads it to the server for analysis. Input the collection of incorrect questions and the students' answer process into the analysis module, and mark the questions according to the error type; At the same time, we analyze the knowledge points and questions that correspond to the questions to obtain the knowledge tags corresponding to the questions; The recommendation module is used to compare the current homework content with the wrong question set. When the similarity between the homework content and the knowledge points and questions in the wrong question set reaches a threshold, the question is added to the candidate list of the deconstruction module. The deconstruction module selects corresponding questions from the candidate list as pre-practice content based on students' practice expectations and reproducibility weights. At the same time, based on the error type of the question, the corresponding knowledge points are provided as the required knowledge points; The deconstruction module is used to decompose and reconstruct the questions in the wrong question set. When a question has two or more knowledge points and the knowledge point blind spot corresponding to the wrong question is two or less, the deconstruction module extracts the corresponding knowledge points according to the error type and reconstructs the quick-answer pre-practice in the form of true / false or fill-in-the-blank questions. Analyze the written content to obtain multi-dimensional answer data for the quick-response pre-exercise, and adjust the recurrence rate and recurrence method of the corresponding knowledge points in the pre-exercise based on the multi-dimensional answer data; The similarity of knowledge points between the homework content and the questions in the error collection is calculated using the TF-IDF algorithm or Euclidean distance. Sort the incorrect and correct questions according to their similarity and provide a list of alternative questions to the deconstruction module; Enter the selected question, its explanation, and the answer into the deconstruction module. The deconstruction process is as follows: S1. Obtain the corresponding basic knowledge points and key points of examination through the science knowledge tags corresponding to the question, and match the basic knowledge points and key points of examination with the question stem one by one; S2. By comparing the answers with the explanations, analyze and predict the types of errors students make, and assign a weight corresponding to the recurrence rate to each type of error. S3. Based on the error type, extract the corresponding content and / or basic knowledge points from the question stem to produce quick-response pre-exercise exercises; The questions in the quick-response pre-exercise are one of the following: true / false, multiple choice, or fill-in-the-blank questions. The quick-response pre-exercise is presented on the screen and a corresponding countdown is set. Each quick-response pre-test contains only two or fewer basic knowledge points and / or test points; S4. Adjust the weight of the recurrence rate of the corresponding error type in this quick-answer pre-exercise based on the completion level of previous quick-answer pre-exercises; S5. If the accuracy rate of this quick-response pre-practice is lower than the threshold, then supplement with a quick-response pre-practice of lower difficulty.

2. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in claim 1.

3. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in claim 1.

Images