Competition planning method and device, electronic equipment and computer readable storage medium

Abstract

The invention discloses a competition planning method and device, electronic equipment and a computer readable storage medium. The competition planning method comprises the following steps: acquiring competition course learning information and competition simulation information of a target user for a target competition; generating knowledge point summary information of the target competition based on the competition course learning information and the competition simulation information; the knowledge point summary information is used for representing the mastering degree of the target user on each knowledge point of the target competition; and inputting the knowledge point summary information and the obtained competition summary information of the target competition into a target machine learning model to obtain competition planning information of the target user. By implementing the method and the device, automatic competition planning is realized, and the competition planning efficiency is improved compared with manual competition planning.

Description

Technical Field

[0001] This disclosure generally relates to the field of artificial intelligence. More specifically, this disclosure relates to a competition planning method, apparatus, electronic device, and computer-readable storage medium. Background Art

[0002] In the actual learning scenarios of users, in order to improve their own abilities, they often participate in many competitions. Specifically, users can prepare for competitions by learning competition courses, practicing competition questions, and simulating competition exams, and then participate in the competitions. Currently, there are many types of competitions on the market, covering different disciplines, different grades, different difficulties, different certification systems, different exam times, etc., while the ages, learning abilities, current bases, and future plans of users themselves are all different. Therefore, in order to improve the learning effect and the efficiency of competition preparation, users need to have a reasonable plan for the competitions they participate in, and the competition planning service has emerged.

[0003] Currently, users usually obtain their own competition planning information (such as what competitions to participate in at what time and the predicted scores for participating in the competitions, etc.) through competition planners. Based on this, competition planners need to understand various competition information in advance, and then comprehensively give users' competition planning information in combination with users' basic information (such as users' grades, ages, etc.), users' historical learning situations, users' competition simulation scores, etc. However, giving competition planning information manually results in low efficiency.

[0004] In view of this, there is an urgent need to provide a competition planning method, apparatus, electronic device, and computer-readable storage medium to improve the efficiency of competition planning. Summary of the Invention

[0005] In order to solve at least one or more of the above-mentioned technical problems, this disclosure proposes a competition planning method, apparatus, electronic device, and computer-readable storage medium in multiple aspects.

[0006] In a first aspect, this disclosure provides a competition planning method, the method includes: obtaining competition course learning information and competition simulation information of a target user for a target competition; generating knowledge point summary information of the target competition based on the competition course learning information and the competition simulation information; the knowledge point summary information is used to represent the mastery degree of each knowledge point of the target competition by the target user; inputting the knowledge point summary information and the obtained competition summary information of the target competition into a target machine learning model to obtain the competition planning information of the target user.

[0007] In some embodiments, generating the summary information of knowledge points for the target competition based on the competition course learning information and the competition simulation information includes: performing structured processing on the competition course learning information to obtain knowledge point learning information; and performing structured processing on the competition simulation information to obtain knowledge point simulation information; inputting the knowledge point learning information and the knowledge point simulation information into a first machine learning model to obtain the summary information of knowledge points.

[0008] In some embodiments, the method further includes: generating an exercise question bank for the target user based on the summary information of knowledge points and the knowledge point simulation information.

[0009] In some embodiments, the summary information of knowledge points includes: the probability of mastering knowledge points; the knowledge point simulation information includes: corresponding knowledge points and simulation questions one by one; generating an exercise question bank for the target user based on the summary information of knowledge points includes: screening the knowledge points of the target competition according to the probability of mastering knowledge points; inputting the screened knowledge points and the simulation questions corresponding to the screened knowledge points into a second machine learning model to obtain the exercise question bank for the target user for the target competition.

[0010] In some embodiments, the competition summary information of the target competition is obtained through the following steps: obtaining the competition basic information of each candidate competition from each competition website; for each candidate competition, performing structured processing on the competition basic information of the candidate competition to obtain the structured basic information of the candidate competition; inputting the structured basic information of each candidate competition into a third machine learning model to generate the competition summary information of each candidate competition, and extracting the competition summary information of the target competition from the competition summary information of each candidate competition.

[0011] In some embodiments, the competition summary information of the target competition includes: the knowledge point distribution information of the target competition in previous years' exams and the score line information of the target competition in previous years; the summary information of knowledge points includes: the probability of mastering knowledge points; the competition planning information includes: the predicted value of the award-winning probability; the target machine learning model includes a first target machine learning model; inputting the summary information of knowledge points and the competition summary information of the obtained target competition into the target machine learning model to obtain the competition planning information of the target user includes: determining the predicted score of the target user in the next exam of the target competition based on the knowledge point distribution information of the target competition in previous years' exams and the probability of mastering knowledge points; inputting the score line information of the target competition in previous years and the predicted score of the target user in the next exam of the target competition into the first target machine learning model to obtain the predicted value of the award-winning probability of the target user for the target competition.

[0012] In some embodiments, the target machine learning model also includes: a second target machine learning model and a third target machine learning model; the method of determining the predicted score of the target user in the next test of the target competition based on the knowledge point distribution information of the target competition in previous years and the probability of mastering the knowledge points includes: inputting the knowledge point distribution information of the target competition in previous years into the second target machine learning model to obtain the probability of occurrence of each knowledge point in the next test of the target competition; determining the predicted knowledge point information of the next test of the target competition based on the probability of occurrence of each knowledge point in the next test of the target competition; obtaining the relationship coefficient, and inputting the relationship coefficient and the probability of mastering the knowledge point into the third target machine learning model to obtain the predicted value of the accuracy of each knowledge point of the target user in the formal competition; the relationship coefficient is used to characterize the relationship between the simulated competition accuracy and the actual competition accuracy of the target competition; performing knowledge point matching on the predicted knowledge point information of the next test of the target competition and the predicted value of the accuracy of each knowledge point of the target user in the formal competition to obtain the predicted score of the target user in the next test of the target competition.

[0013] In some embodiments, obtaining the relationship coefficient includes: obtaining the historical simulated mastery probability and the historical actual answering accuracy rate of each knowledge point of the target competition; inputting the historical simulated mastery probability and the historical actual answering accuracy rate of each knowledge point of the target competition into the fourth machine learning model to obtain the relationship coefficient.

[0014] In a second aspect, the present disclosure provides a competition planning device, comprising: a target information acquisition module, used to acquire competition course learning information and competition simulation information of a target user; a knowledge point summary information generation module, used to generate knowledge point summary information of the target competition based on the competition course learning information and competition simulation information of the target user; the knowledge point summary information is used to characterize the target user's mastery of each knowledge point of the target competition; a competition planning module, used to input the knowledge point summary information and the obtained competition summary information of the target competition into a target machine learning model to obtain the competition planning information of the target user.

[0015] In a third aspect, the present disclosure provides an electronic device comprising: a processor configured to execute program instructions; and a memory configured to store the program instructions, wherein when the program instructions are loaded and executed by the processor, the processor executes the method for competition planning described in the first aspect or any embodiment of the first aspect.

[0016] In a fourth aspect, the present disclosure provides a computer-readable storage medium storing program instructions that, when loaded and executed by a processor, cause the processor to execute the competition planning method described in the first aspect or any embodiment of the first aspect.

[0017] Through the competition planning method, device, electronic device, and computer-readable storage medium provided as above, the embodiments of the present disclosure obtain, for each target user, the competition course learning information and competition simulation information for the target competition to determine the mastery level of each knowledge point of the target competition by the target user, and input the obtained competition summary information of the target competition and the mastery level of each knowledge point of the target competition by the target user into the target machine learning model to output the competition planning information of the target user through the target machine learning model, achieving automated competition planning. Compared with manual competition planning, the competition planning efficiency is improved. Further, the competition basic information is automatically obtained from various competition websites. Compared with manually obtaining the competition basic information from various competition websites, the competition planning effect is further improved, and the problem of poor competition planning effect caused by limited resources obtained when manually obtaining the competition basic information is also avoided. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features, and advantages of the exemplary embodiments of the present disclosure will become readily understood. In the drawings, several embodiments of the present disclosure are shown in an exemplary rather than restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0019] Figure 1 Shows an exemplary flowchart of the competition planning method according to some embodiments of the present disclosure;

[0020] Figure 2 Shows a schematic diagram of generating knowledge point summary information based on competition course learning information and competition simulation information according to some embodiments of the present disclosure;

[0021] Figure 3 Shows a schematic diagram of generating an exercise question bank according to some embodiments of the present disclosure;

[0022] Figure 4 Shows a schematic diagram of generating competition summary information according to some embodiments of the present disclosure;

[0023] Figure 5 Shows a schematic diagram of competition planning according to some embodiments of the present disclosure;

[0024] Figure 6 Shows a schematic diagram of the AI competition planning system according to an embodiment of the present disclosure;

[0025] Figure 7Shows a specific schematic diagram of the competition planning query of some embodiments of the present disclosure;

[0026] Figure 8 Shows a schematic diagram of the competition planning device of some embodiments of the present disclosure;

[0027] Figure 9 Shows a schematic diagram of an electronic device of some embodiments of the present disclosure. Detailed implementation manners

[0028] Next, the technical solutions in the embodiments of the present disclosure will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present disclosure. Obviously, the described embodiments are some, but not all, of the embodiments of the present disclosure. Based on the embodiments in the present disclosure, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of the present disclosure.

[0029] It should be understood that the terms "including" and "comprising" used in the specification and claims of the present disclosure indicate the presence of the described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or their combinations.

[0030] It should also be understood that the terms used in the specification of the present disclosure are only for the purpose of describing specific embodiments, and are not intended to limit the present disclosure. As used in the specification and claims of the present disclosure, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include the plural forms. It should also be further understood that the term " / and / " used in the specification and claims of the present disclosure refers to any combination and all possible combinations of one or more of the related listed items, and includes these combinations.

[0031] As used in this specification and the claims, the term "if" can be interpreted as "when...", "once", "in response to determining", or "in response to detecting" according to the context. Similarly, the phrase "if determined" or "if [the described condition or event] is detected" can be interpreted as meaning "once determined", "in response to determining", "once [the described condition or event] is detected", or "in response to detecting [the described condition or event]" according to the context.

[0032] Next, the detailed implementation manners of the present disclosure will be described in detail with reference to the accompanying drawings.

[0033] Figure 1 Shows an exemplary flowchart of the competition planning method 100 of some embodiments of the present disclosure.

[0034] As Figure 1As shown, the competition planning method includes: Step S110: Obtain the competition course learning information and competition simulation information of the target user for the target competition; Step S120: Generate the knowledge point summary information of the target competition based on the competition course learning information and the competition simulation information; Step S130: Input the knowledge point summary information and the obtained competition summary information of the target competition into the target machine learning model to obtain the competition planning information of the target user.

[0035] In the embodiment of the present disclosure, the target user in Step S110 refers to a user who has purchased competition courses and participated in simulation competitions on the platform. Here, the competition planning method 100 is deployed on the platform, which can be, for example, an application program, a mini program, etc. In this way, the competition course learning information and competition simulation information of the target user can be obtained to facilitate competition planning for the target user.

[0036] In the embodiment of the present disclosure, the target competition in Step S110 refers to the competition corresponding to the competition courses purchased by the target user. There can be many types of the target competition, covering different disciplines, different grades, different difficulties, different certification systems, different exam times, etc. The target competition can be, for example, the Kangaroo Math Competition, the Eurasian English Olympiad (KGL) English Competition, the National English Competition for College Students, etc. The embodiment of the present disclosure does not specifically limit this.

[0037] In the embodiment of the present disclosure, the target user can purchase competition courses for learning and participate in simulation competitions according to the competition platform they want to participate in. Based on this, in Step S110, the competition course learning information and competition simulation information of the target user for the target competition can be obtained.

[0038] Here, the competition course learning information refers to the learning situation of the user for each class hour in the competition course, including the learned class hours and the unlearned class hours; the simulation practice information refers to the correct rate of the simulation questions done by the target user.

[0039] Exemplarily, the knowledge point summary information is used to represent the mastery degree of each knowledge point of the target user for the target competition, and specifically may include: the mastery probability of each knowledge point, the answering correct rate of different difficulty questions under each knowledge point, etc. The embodiment of the present application does not specifically limit this.

[0040] For example, as Figure 2 shown, the knowledge point summary information can be: Knowledge point 1: Mastery probability, answering correct rate of difficulty type 1, answering correct rate of difficulty type 2,...; Knowledge point 2: Mastery probability, answering correct rate of difficulty type 1, answering correct rate of difficulty type 2,...; Knowledge point 3: Mastery probability, answering correct rate of difficulty type 1, answering correct rate of difficulty type 2,....

[0041] In the embodiments of the present disclosure, the specific method for generating the summary information of knowledge points of the target competition based on the competition course learning information and the competition simulation information may be as follows: Structurally process the competition course learning information to obtain the knowledge point learning information; and structurally process the competition simulation information to obtain the knowledge point simulation information; input the knowledge point learning information and the knowledge point simulation information into the first machine learning model to obtain the summary information of knowledge points.

[0042] Exemplarily, in order to improve the information processing ability, data is often structurally processed. Structural processing refers to organizing unstructured data in a certain way to meet specific application requirements. In the embodiments of the present disclosure, the competition course learning information and the competition simulation information are structurally processed to obtain structured data suitable for the machine learning model.

[0043] As Figure 2 shown, here, the knowledge point learning information may be information on whether each knowledge point has been learned, including: learned knowledge points and unlearned knowledge points. The knowledge point simulation information may include: the correct rate of each knowledge point, the correspondence between each simulation question and the knowledge points involved in the simulation question (i.e., the one-to-one corresponding knowledge points and simulation questions), and the answering details (the question difficulty of each simulation question and the answering result (correct or wrong)). Here, the question difficulty of each simulation question may be defined in advance on the test paper.

[0044] In the embodiments of the present disclosure, by structurally processing the competition course learning information, the association relationship between the class hours of study and the knowledge points is converted into the knowledge point learning information; by structurally processing the competition simulation information, the result of whether the simulation test questions are correct and the association relationship between the simulation questions and the knowledge points are converted into the knowledge point simulation information.

[0045] In the embodiments of the present disclosure, the first machine learning model is a debugged large language model (LLM model), for example, the GPT-4o model, the Doubao large language model, etc. By inputting the knowledge point learning information and the knowledge point simulation information into the first machine learning model together, the summary information of knowledge points is obtained. Here, the principle for the first machine learning model to calculate the mastery probability of each knowledge point is: weighted average of the knowledge point learning information and the knowledge point answering situation (i.e., the knowledge point correct rate), where the assignment of learned and unlearned in the knowledge point learning information and the weight values of the knowledge point learning information and the knowledge point correct rate can be given by the first machine learning model independently or can be defined in advance, and the embodiments of the present disclosure do not specifically limit this.

[0046] In the embodiments of the present disclosure, when the first machine learning model is actually calculating and determining the knowledge point accuracy rate, the difficulty of the simulation questions can be additionally considered. That is, for simulation questions of different difficulties, different weights (taking values between 0 and 1) can be given. Here, the weight of the question difficulty can also be given by the first machine learning model.

[0047] As a specific example of the present disclosure, for the learned and unlearned in the knowledge point learning information, the learned is assigned a value of 1, the unlearned is assigned a value of 0, the weight of the knowledge point learning information is 30%, and the weight of the knowledge point accuracy rate is 70%. Based on this, the knowledge point mastery probability of each knowledge point = knowledge point learning information * 30% + knowledge point accuracy rate * 70% = learned / unlearned * 30% + knowledge point accuracy rate * 70%.

[0048] In the embodiments of the present disclosure, the competition summary information of the target competition (see Figure 4 ) may include: competition basic information, event arrangement information, previous year's score line information, and the knowledge point distribution information examined by the target competition. Among them, the competition basic information may include: competition name information, competition stage information, competition difficulty information, and the grade of the target user; the event arrangement information includes: the registration time and the start time of the exam of the target competition; the previous year's score line information may include: the award-winning score line, the number of award-winning people, and the award-winning ranking ratio. The knowledge point distribution information examined by the target competition includes the examination probability of different knowledge points. As for the specific explanations and examples of each item, reference can be made to the specific descriptions of the embodiments of the following AI competition planning system, which will not be elaborated here.

[0049] In the embodiments of the present disclosure, the competition summary information of the target competition can be generated and stored locally on the platform in advance and directly obtained from the local when needed, or the competition basic information of the target competition can be obtained from the website of the target competition when needed, and then the competition summary information of the target competition can be generated based on the competition basic information of the target competition (for example, generated by a machine learning model). The embodiments of the present disclosure do not specifically limit the method for obtaining the competition summary information of the target competition, which can be determined according to the actual situation.

[0050] In the embodiments of the present disclosure, the target machine learning model can be a single machine learning model or a collection of multiple machine learning models. The embodiments of the present disclosure do not specifically limit the target machine learning model.

[0051] Exemplarily, the competition planning information may include: competition name information, competition stage information, competition time and registration time (i.e., the above-mentioned event arrangement information), and the predicted award-winning probability value, etc. The embodiments of the present disclosure do not specifically limit this.

[0052] As for the specific process of inputting the summary information of knowledge points and the competition summary information of the target competition obtained into the target machine learning model to obtain the competition planning information of the target user, reference can be made to the specific description of the following embodiments, which will not be elaborated here for the time being.

[0053] In the embodiments of the present disclosure, for each target user, the competition course learning information and competition simulation information of the target user for the target competition are obtained to determine the mastery degree of each knowledge point of the target user for the target competition, and the competition summary information of the target competition and the mastery degree of each knowledge point of the target user for the target competition obtained are input into the target machine learning model, so as to output the competition planning information of the target user through the target machine learning model, realizing automated competition planning. Compared with manual competition planning, the competition planning efficiency is improved.

[0054] So far, the completion Figure 1 of the description of the exemplary flowchart of the competition planning method shown and Figure 2 the description of the schematic diagram for generating the summary information of knowledge points shown.

[0055] To facilitate the user to practice targeted, after generating the knowledge point simulation information and the summary information of knowledge points, a practice question bank for the target user can also be generated. Specifically, Figure 3 shows a schematic diagram of generating a practice question bank in some embodiments of the present disclosure. As Figure 3 shown, a practice question bank for the target user is generated based on the summary information of knowledge points and the knowledge point simulation information.

[0056] Exemplarily, the summary information of knowledge points includes: the probability of mastering knowledge points. When generating a practice question bank for the target user, in order to improve the target user's competition preparation efficiency, filtering can be performed first based on the probability of mastering knowledge points, and the knowledge points that the target user has not fully mastered can be filtered out. For example, the knowledge points with a probability of mastering knowledge points less than 100% can be filtered out. For another example, the knowledge points with a probability of mastering knowledge points less than 98% can be filtered out. In the embodiments of the present disclosure, the knowledge points with the probability of mastering knowledge points less than 100% or 98% can be defined as the knowledge points that the target user has not fully mastered.

[0057] After screening out the knowledge points that the target user has not fully mastered, a practice question bank for the knowledge points that have not been fully mastered is generated. Specifically, the screened knowledge points and the corresponding simulation questions (including simulation questions of different difficulties) of the screened knowledge points are input into the second machine learning model to obtain the practice question bank for the target user for the target competition. Among them, the second machine learning can be a debugged large language model (LLM model), for example, the GPT-4o model, the Doubao large language model, etc. In the embodiment of the present disclosure, the screened knowledge points are used as the labels of the simulation questions corresponding to the knowledge points and input into the second machine learning model together with the simulation questions corresponding to the knowledge points. The second machine learning model learns the proposition ideas and proposition specifications of the simulation questions and outputs the practice question bank for the knowledge points that the target user has not fully mastered.

[0058] It should be noted that in the embodiment of the present disclosure, in order to strengthen the user's memory of each knowledge point, the knowledge points can be not screened, and each knowledge point and the corresponding simulation questions are input into the second machine learning model together to obtain the practice question bank for all knowledge points.

[0059] So far, the description of the schematic diagram of generating the practice question bank shown in Figure 3 is completed.

[0060] As an optional implementation manner of the embodiment of the present disclosure, the competition summary information of the target competition is obtained through the following steps: obtaining the competition basic information of each candidate competition from each competition website; for each candidate competition, performing structured processing on the competition basic information of the candidate competition to obtain the structured basic information of the candidate competition; inputting the structured basic information of each candidate competition into the third machine learning model to generate the competition summary information of each candidate competition, and extracting the competition summary information of the target competition from the competition summary information of each candidate competition.

[0061] Exemplarily, in the embodiment of the present disclosure, as shown in Figure 4 , the competition basic information of each candidate competition can be obtained from each competition website regularly (for example, every half month) through an automated script. Here, the competition basic information may include: competition name information, competition stage information, competition difficulty information, competition score line information, event arrangement information, and real question information, etc.

[0062] For each candidate competition, the basic competition information of the candidate competition is structured to obtain information that is easy for the model to understand, that is, the basic information of the candidate competition after structured processing. Here, the basic competition information after structured processing can be in the form of a table. For example, the title of the table is the competition name; the table headers are competition stage information, competition difficulty information, competition cut-off score information, and event arrangement information; the table body is the specific content corresponding to the table headers. Here, a table is generated for each obtained candidate competition.

[0063] In the embodiments of the present disclosure, the third machine learning model is obtained by debugging a large language model (LLM model), such as the GPT-4o model, the Doubao large language model, etc. The basic information of each candidate competition after structured processing is input into the third machine learning model to generate the competition summary information of each candidate competition. Specifically, the tables of each candidate competition are sequentially input into the third machine learning model, and the third machine learning model organizes the information and analyzes the data in the input tables to obtain the competition summary information of each candidate competition. When it is necessary to obtain the summary information of the target competition, the summary information of the target competition can be extracted from the competition summary information of each candidate competition.

[0064] In the embodiments of the present disclosure, by automatically obtaining the basic competition information from each competition website, compared with manually obtaining the basic competition information from each competition website, the competition planning effect is further improved, and the problem that the competition planning effect is poor due to limited resources obtained when manually obtaining the basic competition information is also avoided; at the same time, by using AI to organize the basic competition information to obtain the competition summary information of each candidate competition, the efficiency and accuracy of data analysis are improved.

[0065] So far, the Figure 4 description of the schematic diagram for generating competition summary information shown is completed.

[0066] As an optional embodiment of the embodiments of the present disclosure, the competition summary information of the target competition includes: the knowledge point distribution information of the target competition's past exams and the cut-off score information of the target competition over the years; the knowledge point summary information includes: the probability of mastering knowledge points; the competition planning information includes: the predicted value of the winning probability; the target machine learning model includes the first target machine learning model; the knowledge point summary information and the obtained competition summary information of the target competition are input into the target machine learning model to obtain the competition planning information of the target user, including: determining the predicted score of the target user in the next exam of the target competition based on the knowledge point distribution information of the target competition's past exams and the probability of mastering knowledge points; inputting the cut-off score information of the target competition over the years and the predicted score of the target user in the next exam of the target competition into the first target machine learning model to obtain the predicted value of the winning probability of the target user for the target competition.

[0067] Exemplarily, the first target machine learning model is a debugged large language model (i.e., LLM model), such as the GPT-4o model, the Doubao large language model, etc. In the embodiments of the present disclosure, the score line information of previous years of the target competition and the predicted score of the target user in the next exam of the target competition are input into the first target machine learning model. First, the first target machine learning model summarizes the score rules for winning awards based on the score line information of previous years of the target competition, and then obtains the predicted value of the award-winning probability of the target user for the target competition according to this rule and outputs it.

[0068] In the embodiments of the present disclosure, the target machine learning model further includes: a second target machine learning model and a third target machine learning model; specifically, determining the predicted score of the target user in the next exam of the target competition based on the knowledge point distribution information and the knowledge point mastery probability of previous years of the target competition can be: inputting the knowledge point distribution information of previous years of the target competition into the second target machine learning model to obtain the occurrence probability of each knowledge point in the next exam of the target competition; determining the predicted knowledge point information of the next exam of the target competition based on the occurrence probability of each knowledge point in the next exam of the target competition; obtaining the correlation coefficient, and inputting the correlation coefficient and the knowledge point mastery probability into the third target machine learning model to obtain the predicted correct rate of each knowledge point of the target user in the official competition; matching the predicted knowledge point information of the next exam of the target competition and the predicted correct rate of each knowledge point of the target user in the official competition to obtain the predicted score of the target user in the next exam of the target competition.

[0069] Exemplarily, the second target machine learning model is a debugged large language model (LLM model), such as the GPT-4o model, the Doubao large language model, etc. The knowledge point distribution information of previous years of the target competition is input into the second target machine learning model, so that the second target machine learning model summarizes and analyzes the occurrence rules of each knowledge point in previous years of the exam, and then predicts the occurrence probability of different knowledge points in the next exam according to this rule to obtain the occurrence probability of each knowledge point in the next exam of the target competition.

[0070] Exemplarily, the predicted knowledge point information may include: the predicted knowledge points in the next exam of the target competition and the occurrence probability of the predicted knowledge points. In the embodiments of the present disclosure, determining the predicted knowledge point information of the next exam of the target competition based on the occurrence probability of each knowledge point in the next exam of the target competition, specifically, the knowledge points with the specified number (e.g., 20) before the occurrence probability can be used as the predicted knowledge points.

[0071] In the embodiments of this disclosure, the relationship coefficient is used to characterize the relationship between the simulated competition accuracy rate and the actual competition accuracy rate of the target competition. In actual competition planning, the relationship coefficient can be obtained through the following steps: Obtain the historical simulated mastery probabilities and historical actual answering accuracy rates of each knowledge point of the target competition; Input the historical simulated mastery probabilities and historical actual answering accuracy rates of each knowledge point of the target competition into the fourth machine learning model to obtain the relationship coefficient.

[0072] Exemplarily, in the embodiments of this disclosure, for the target competition, information of other users during historical simulation (i.e., the historical simulated mastery probabilities of each knowledge point of the target competition) and information during historical actual exams (i.e., the historical actual answering accuracy rates of each knowledge point of the target competition) can be obtained, and then the relationship coefficient can be obtained based on the information of other users during historical simulation and information during historical actual exams.

[0073] Specifically, the fourth machine learning model is a debugged large language model (LLM model), for example, the GPT-4o model, the Doubao large language model, etc. Input the historical simulated mastery probabilities and historical actual answering accuracy rates of each knowledge point of the target competition into the fourth machine learning model, so that the fourth machine learning model summarizes the relationship between the historical simulated mastery probabilities and historical actual answering accuracy rates of each knowledge point of the target competition, obtains the proportional relationship between the mastery probability in the simulated exam and the answering accuracy rate in the actual answering for each knowledge point, and the relationship coefficient can be obtained by weighted averaging the proportional relationships of each knowledge point.

[0074] After obtaining the relationship coefficient, input the relationship coefficient and the knowledge point mastery probability into the third target machine learning model, and the third target machine learning model obtains the predicted accuracy rate values of each knowledge point of the target user in the official competition based on the relationship coefficient and the knowledge point mastery probability. Here, the third target machine learning model is a debugged large language model (LLM model), for example, the GPT-4o model, the Doubao large language model, etc.

[0075] In an embodiment of the present disclosure, after obtaining the predicted correct rate values of each knowledge point of the target user in the official competition, a match is made based on the predicted knowledge point information of the next exam of the target competition and the predicted correct rate values of each knowledge point of the target user in the official competition, to obtain the predicted score of the target user in the next exam of the target competition. Specifically, for the target competition, the product of the occurrence probability of the predicted knowledge point X in the next exam of the target competition and the correct rate of the knowledge point X of the target user in the official competition is used as the predicted score of the knowledge point X, and the product of the occurrence probability of the predicted knowledge point Y in the next exam of the target competition and the correct rate of the knowledge point Y of the target user in the official competition is used to obtain the predicted score of the knowledge point Y, and so on, to obtain the predicted scores of each predicted knowledge point. Furthermore, the sum of the predicted scores of each predicted knowledge point is used as the predicted score of the target user in the next exam of the target competition.

[0076] The following combines Figure 5 the schematic diagram of the competition plan shown to describe the above competition planning process:

[0077] As Figure 5 shown, the above competition planning process is divided into three parts, namely: predicting the correct rate of each knowledge point of the target user in the exam environment, predicting the knowledge points examined in the competition, predicting the user's exam score and the winning probability.

[0078] Among them, for predicting the correct rate of each knowledge point of the target user in the exam environment, the historical simulated mastery probability of each knowledge point of the user (i.e., the historical simulated mastery probability of each knowledge point of the above target competition) and the historical actual answering correct rate of each knowledge point of the user (i.e., the historical actual answering correct rate of each knowledge point of the above target competition) are first input into the fourth machine learning model to obtain a relationship coefficient. The above relationship coefficient and the knowledge point mastery probability of the target user for each knowledge point are input into the third target machine learning model to obtain the predicted correct rate values of each knowledge point of the target user in the official competition.

[0079] For predicting the knowledge points examined in the competition, the knowledge point distribution information of the previous exams in the competition summary information is input into the second target machine learning model, so that the second target machine learning model summarizes and analyzes the occurrence rules of each knowledge point in the previous exams, and then predicts the occurrence probability of different knowledge points in the next exam according to this rule, to obtain the occurrence probability of each knowledge point in the next exam of the target competition, and determine the predicted knowledge point information of the next exam of the target competition based on the occurrence probability of each knowledge point in the next exam of the target competition.

[0080] For the prediction of the user's exam scores and winning probabilities, the predicted correct rate values of each knowledge point of the target user in the official competition and the predicted knowledge point information of the next exam in the target competition are matched for knowledge points to obtain the predicted score of the target user in the next exam of the target competition. Then, the predicted score of the target user in the next exam of the target competition and the score line information of previous years of the target competition are input into the first target machine learning model. The first target machine learning model first summarizes the score rules for winning awards based on the score line information of previous years of the target competition, and then predicts the probability that the target user can win an award according to this rule and outputs it.

[0081] Thus, the description of Figure 5 the schematic diagram of the competition plan shown is completed.

[0082] Based on the above Figure 1 - Figure 5 description of the competition method, the AI competition planning system provided by the embodiments of this disclosure will be described below through Figure 6 the schematic diagram of the AI competition planning system shown:

[0083] As Figure 6 shown, the AI competition planning system includes 4 modules, namely a competition information induction module, a user learning situation analysis module, a special topic library generation module, and a competition planning and prediction module. Specifically, the above AI competition planning system can be deployed in an application or a mini-program (here, the application and the mini-program are denoted as the platform), and users can achieve automated competition planning through the application or the mini-program.

[0084] In the embodiments of this disclosure, there can be many types of competitions, covering different disciplines, different grades, different difficulties, different certification systems, different exam times, etc. Here, the competitions can be, for example, the Kangaroo Math Competition, the Eurasian English Olympiad (KGL) English Competition, the National English Competition for College Students, etc., and the embodiments of this disclosure do not specifically limit this.

[0085] Among them, the competition information induction module can regularly update the competition basic information of each competition. In the embodiments of this disclosure, the competition basic information can include: competition name information, competition stage information, competition difficulty information, competition score line information, event arrangement information, and real question information, etc. The AI competition planning system obtains the competition summary information of each competition through inductive integration of the competition basic information of each competition (for specific descriptions, refer to the descriptions of the above embodiments) for subsequent competition planning of users.

[0086] The following uses a specific competition (Kangaroo Math Competition) as an example to explain the basic information of the competition. For the Kangaroo Math Competition, its competition name information is: Kangaroo Math Competition. The competition stage information is pre-divided, for example: A, B, C, D, E, F. Among them, stage A corresponds to grades 1 to 2 in primary school; stage B corresponds to grades 3 to 4 in primary school; stage C corresponds to grades 5 to 6 in primary school; stage D corresponds to grades 7 to 8 in middle school (first and second grades of junior high school); stage E corresponds to grades 9 to 10 in middle school (third grade of junior high school and first grade of senior high school); stage F corresponds to grades 11 to 12 in middle school (second and third grades of senior high school). The competition difficulty information matches the above competition stage information, and the difficulty gradually increases from stage A to stage F. In the disclosed embodiment, for example, it can be represented by L1, L2, L3, L4, L5, L6, or other representation methods can also be used, which is not limited here.

[0087] The competition score line information can include: the specific score lines of different awards over the years, the number of awardees, the ranking ratio, etc. The specific score lines of different awards over the years are, for example: the score line for the gold award in 2023: grade 1 in primary school: 99 points; grade 2 in primary school: 108 points; grade 3 in primary school: 99 points; grade 4 in primary school: 106 points; grade 5 in primary school: 124 points; grade 6 in primary school: 132 points; first grade of junior high school: 118 points; second grade of junior high school: 124 points; third grade of junior high school: 114 points; first grade of senior high school: 115 points; second grade of senior high school: 107 points; third grade of senior high school: 139 points. The number of awardees is, for example: gold award: 227 people; silver award: 75 people; bronze award: 137 people. The ranking ratio can be, for example: gold award: top 10%; silver award: top 20%; bronze award: top 35%.

[0088] The event arrangement information can include: the competition registration time and the competition exam time. For example, the competition registration time is: March 18, 2023; the competition exam time is: March 29, 2023; the exam time for stages A, B, and C is: 10:00 - 11:15 on March 29, 2023; the exam time for stages D, E, and F is: 14:00 - 15:15 on March 29, 2023.

[0089] The true question information can include the true questions of each competition stage over the years. In the disclosed embodiment, the true question information can include: the corresponding knowledge points and true question topics.

[0090] Also as Figure 6 shown, the user learning situation analysis module can include: the competition course learning situation and the simulation practice situation. Among them, the competition course learning situation refers to the user's learning situation for each class hour, including learned and unlearned; the simulation practice situation refers to the correct rate of the simulation questions. By analyzing the user's learning situation and simulation situation, data support can be provided for the special question bank generation module and the competition planning and prediction module.

[0091] Also as Figure 6 shown, the special topic question bank generation module is used to generate a special topic question bank for the user according to the analysis results obtained by the user learning situation analysis module, so that the user can practice targeted.

[0092] Also as Figure 6 shown, the competition planning and prediction module may include: competition participation plan and performance prediction. Specifically, by comprehensively analyzing the summarized information of the competition induction module and the learning situation and simulation situation of the user learning situation analysis module, the competitions that the user can participate in (i.e., the competition participation plan) and the possible results in the competitions (i.e., the performance prediction) are obtained, so that the user can use this as a reference for competition preparation.

[0093] So far, the description of the AI competition planning system shown Figure 6 is completed.

[0094] Figure 7 shows a specific schematic diagram of competition planning query for some embodiments of the present disclosure.

[0095] As Figure 7 shown, when the target user opens the competition planning platform and enters the mobile phone number on the first interface (i.e., the login interface), after clicking the "Apply for Query" button, the user can apply for querying their competition planning information. After the target user clicks the "Apply for Query" button, it jumps to the second interface (i.e., the competition planning result display interface), and this interface displays the competition planning query results of the target user, which may include: course learning records, exam simulation records, special practice recommendations, and subsequent competition plans. As a specific example, the course learning record is: XX class hours have been learned; the exam simulation record is: XX simulations have been completed; the special practice recommendations may include: special practice on geometric image observation, special training on number sense and calculation ability; the subsequent competition plans may include at least one recommended competition to participate in, and each recommended competition to participate in may include: competition name information, competition stage information, competition time and registration time (i.e., the above-mentioned event arrangement information), and winning probability ( Figure 7 60% and 40% in

[0096] When the target user clicks the course learning record button on the second interface, a third interface (i.e., the class hour learning record interface) pops up above this second interface, and this interface displays the names of the unlearned class hours and the corresponding key knowledge points, as well as the names of the learned class hours and the corresponding key knowledge points.

[0097] The target user clicks the exam simulation record button on the second interface, and a fourth interface (i.e., the exam simulation record interface) pops up above the second interface. This interface shows the test paper names of each simulation exam participated by the target user and the simulation scores of each test paper ( Figure 7 such as 99 points, 90 points, 89 points, and 80 points in

[0098] ). The target user clicks one of the buttons for the recommended competitions on the second interface, and a fifth interface (i.e., the recommended competition interface) pops up above the second interface. In addition to showing the competition name information, competition stage information, competition time and registration time (i.e., the above-mentioned event arrangement information) and the winning probability, this interface also shows the cut-off scores, winning rankings and the number of awardees of this competition over the years.

[0099] So far, the description of the competition planning query shown in Figure 7 is completed.

[0100] Figure 8 shows a schematic diagram of a competition planning device according to some embodiments of the present disclosure.

[0101] As Figure 8 shown, the competition planning device 80 includes: a target information acquisition module 810 for acquiring the competition course learning information and competition simulation information of the target user; a knowledge point summary information generation module 820 for generating knowledge point summary information of the target competition based on the competition course learning information and competition simulation information of the target user; the knowledge point summary information is used to represent the target user's mastery of each knowledge point of the target competition; a competition planning module 830 for inputting the knowledge point summary information and the competition summary information of the obtained target competition into the target machine learning model to obtain the competition planning information of the target user.

[0102] As an optional embodiment of the present disclosure, the above-mentioned knowledge point summary information generation module is specifically configured to: perform structured processing on the competition course learning information to obtain knowledge point learning information; and perform structured processing on the competition simulation information to obtain knowledge point simulation information; input the knowledge point learning information and the knowledge point simulation information into the first machine learning model to obtain the knowledge point summary information.

[0103] As an optional embodiment of the present disclosure, the above-mentioned competition planning device further includes: a practice question bank generation module for generating a practice question bank for the target user based on the knowledge point summary information and the knowledge point simulation information.

[0104] As an alternative embodiment of the present disclosure, the knowledge point summary information includes: the probability of mastering knowledge points; the knowledge point simulation information includes: the corresponding knowledge points and simulation questions; the above exercise question bank generation module is specifically configured to: screen the knowledge points of the target competition according to the probability of mastering knowledge points; input the screened knowledge points and the simulation questions corresponding to the screened knowledge points into the second machine learning model to obtain an exercise question bank for the target user for the target competition.

[0105] As an alternative embodiment of the present disclosure, the competition planning device further includes: a competition basic information acquisition module, configured to acquire the competition basic information of each candidate competition from each competition website; a structured processing module, configured to perform structured processing on the competition basic information of each candidate competition to obtain the structured basic information of the candidate competition; an information extraction module, configured to input the structured basic information of each candidate competition into a third machine learning model to generate the competition summary information of each candidate competition, and extract the competition summary information of the target competition from the competition summary information of each candidate competition.

[0106] As an alternative embodiment of the present disclosure, the competition summary information of the target competition includes: the knowledge point distribution information of the target competition in previous years' examinations and the score line information of the target competition in previous years; the knowledge point summary information includes: the probability of mastering knowledge points; the competition planning information includes: the predicted value of the award-winning probability; the target machine learning model includes the first target machine learning model; the above competition planning module is specifically configured to: determine the predicted score of the target user in the next examination of the target competition based on the knowledge point distribution information of the target competition in previous years' examinations and the probability of mastering knowledge points; input the score line information of the target competition in previous years and the predicted score of the target user in the next examination of the target competition into the first target machine learning model to obtain the predicted value of the award-winning probability of the target user for the target competition.

[0107] As an alternative embodiment of the present disclosure, the target machine learning model further includes: a second target machine learning model and a third target machine learning model; the competition planning module is further configured to: input the knowledge point distribution information of the target competition in previous years' examinations into the second target machine learning model to obtain the occurrence probability of each knowledge point in the next examination of the target competition; determine the predicted knowledge point information of the next examination of the target competition based on the occurrence probability of each knowledge point in the next examination of the target competition; obtain the relationship coefficient, and input the relationship coefficient and the probability of mastering knowledge points into the third target machine learning model to obtain the predicted correct rate of each knowledge point of the target user in the official competition; the relationship coefficient is used to characterize the relationship between the correct rate of the simulation competition and the correct rate of the actual competition of the target competition; perform knowledge point matching on the predicted knowledge point information of the next examination of the target competition and the predicted correct rate of each knowledge point of the target user in the official competition to obtain the predicted score of the target user in the next examination of the target competition.

[0108] As an optional embodiment of this disclosure, the above competition planning module is further configured to: obtain the historical simulated mastery probabilities and historical actual answering correct rates of each knowledge point of the target competition; input the historical simulated mastery probabilities and historical actual answering correct rates of each knowledge point of the target competition into the fourth machine learning model to obtain a correlation coefficient.

[0109] For the specific implementation manners and technical effects, reference may be made to the specific descriptions of the embodiments of the above method 100, which will not be elaborated herein.

[0110] So far, the description of Figure 8 the device shown is completed.

[0111] Correspondingly, the embodiments of this disclosure also provide Figure 8 the hardware structure diagram of the device shown, specifically as Figure 9 shown. The electronic device 90 may be the device for implementing the above method 100. As Figure 9 shown, the electronic device 90 includes: a processor 910 and a memory 920. Among them, the memory 920 is configured to store program instructions; the processor 910 is configured to load and execute the program instructions stored in the memory 920 to implement the corresponding method embodiments of the competition planning as shown above.

[0112] As an embodiment, the memory 920 may be any electronic, magnetic, optical or other physical storage device that can contain or store information, such as program instructions, data, and so on. For example, the memory 920 may be: a volatile memory, a non-volatile memory or a similar storage medium. Specifically, the memory 920 may be a RAM (Random Access Memory), a flash memory, a storage drive (such as a hard disk drive), a solid-state drive, any type of storage disk (such as an optical disc, a DVD, etc.), or a similar storage medium, or a combination thereof.

[0113] So far, the description of Figure 9 the electronic device shown is completed.

[0114] Although multiple embodiments of this disclosure have been shown and described herein, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art can think of many changes, alterations, and alternative ways without departing from the spirit and scope of this disclosure. It should be understood that various alternative solutions to the embodiments of this disclosure described herein may be adopted in the practice of this disclosure. The appended claims are intended to define the scope of protection of this disclosure and thus cover equivalents or alternative solutions within the scope of these claims.

Claims

1. A competition planning method, characterized in that: The method comprises: Obtain target users’ competition course learning information and competition simulation information for target competitions; Generate knowledge point summary information of the target competition based on the competition course learning information and the competition simulation information; the knowledge point summary information is used to represent the mastery degree of each knowledge point of the target competition by the target user; The knowledge point summary information and the obtained competition summary information of the target competition are input into the target machine learning model to obtain the competition planning information of the target user.

2. The method according to claim 1, characterized in that The generating of the knowledge point summary information of the target competition based on the competition course learning information and the competition simulation information includes: Performing structured processing on the competition course learning information to obtain knowledge point learning information; and Performing structured processing on the competition simulation information to obtain knowledge point simulation information; The knowledge point learning information and the knowledge point simulation information are input into a first machine learning model to obtain the knowledge point summary information.

3. The method according to claim 2, characterized in that The method further comprises: A practice question bank for the target user is generated based on the knowledge point summary information and the knowledge point simulation information.

4. The method according to claim 3, characterized in that The knowledge point summary information includes: knowledge point mastery probability; the knowledge point simulation information includes: one-to-one corresponding knowledge points and simulation questions; the generating of the exercise question bank for the target user based on the knowledge point summary information includes: Screening the knowledge points of the target competition according to the mastery probability of the knowledge points; The screened knowledge points and the simulation questions corresponding to the screened knowledge points are input into the second machine learning model to obtain a practice question bank for the target user for the target competition.

5. The method according to claim 1, characterized in that The competition summary information of the target competition includes: knowledge point distribution information of the target competition in previous years and score line information of the target competition in previous years; the knowledge point summary information includes: knowledge point mastery probability; the competition planning information includes: winning probability prediction value; the target machine learning model includes a first target machine learning model; The step of inputting the knowledge point summary information and the obtained competition summary information of the target competition into the target machine learning model to obtain the competition planning information of the target user includes: Determine the predicted score of the target user in the next test of the target competition based on the knowledge point distribution information of the target competition in previous years and the probability of mastering the knowledge points; The score line information of the target competition in previous years and the predicted score of the target user in the next exam of the target competition are input into the first target machine learning model to obtain the predicted value of the target user's winning probability for the target competition.

6. The method according to claim 5, characterized in that The target machine learning model further includes: a second target machine learning model and a third target machine learning model; the predicted score of the target user in the next test of the target competition based on the knowledge point distribution information of the target competition in previous years and the probability of mastering the knowledge points includes: Input the knowledge point distribution information of the target competition in previous years' examinations into the second target machine learning model to obtain the probability of occurrence of each knowledge point in the next examination of the target competition; Determining predicted knowledge point information for the next test of the target competition based on the probability of occurrence of each knowledge point in the next test of the target competition; Obtaining a relationship coefficient, and inputting the relationship coefficient and the probability of mastering the knowledge point into the third target machine learning model to obtain a predicted value of the accuracy of each knowledge point of the target user in the formal competition; the relationship coefficient is used to characterize the relationship between the accuracy of the simulated competition and the accuracy of the actual competition of the target competition; Knowledge point matching is performed on the predicted knowledge point information of the next test of the target competition and the predicted accuracy value of each knowledge point of the target user in the official competition to obtain the predicted score of the target user in the next test of the target competition.

7. The method according to claim 6, characterized in that Obtain relational number systems, including: Obtain the historical simulated mastery probability and historical actual answer accuracy rate of each knowledge point of the target competition; The historical simulated mastery probability of each knowledge point of the target competition and the historical actual answer accuracy rate are input into the fourth machine learning model to obtain the relationship coefficient.

8. A competition planning device, characterized in that: The device comprises: The target information acquisition module is used to obtain the target user's competition course learning information and competition simulation information; A knowledge point summary information generating module, used to generate knowledge point summary information of the target competition based on the competition course learning information and competition simulation information of the target user; the knowledge point summary information is used to characterize the mastery degree of each knowledge point of the target competition by the target user; The competition planning module is used to input the knowledge point summary information and the obtained competition summary information of the target competition into the target machine learning model to obtain the competition planning information of the target user.

9. An electronic device, characterized in that: include: a processor configured to execute program instructions; as well as A memory configured to store the program instructions, which, when loaded and executed by the processor, enables the processor to execute the competition planning method according to any one of claims 1-7.

10. A computer-readable storage medium having program instructions stored therein, characterized in that: When the program instructions are loaded and executed by a processor, the processor executes the competition planning method according to any one of claims 1-7.

Images