Method and system for information recommendation based on virtual user

Abstract

The present application relates to the technical field of wisdom education, especially to a kind of information recommendation method and system based on virtual user, method includes: based on effective processing proportion determines whether the exercise push meets user demand, and when determining that the exercise push does not meet user demand, corresponding processing mode is generated according to the determined reason;When the reason that does not meet user demand is that the stage adaptation rate is low, respectively based on memory strength increment and examination period proximity adjust push interval, and based on the push interval after correction determine whether to adjust the volume of single push;When the reason that does not meet user demand is that the associated recommendation is missing, the push associated exercises or the number of knowledge points is reduced based on content matching degree optimization;Adjustment is completed, based on effective processing proportion determines whether the exercise push meets user demand, for the exercise recommendation that meets user demand, the exercise recommendation is optimized according to the question diversity coefficient processing.The present application effectively improves the accuracy of exercise recommendation.

Description

Technical Field

[0001] This invention relates to the field of smart education technology, and in particular to an information recommendation method and system based on virtual users. Background Technology

[0002] With the deep integration of educational informatization and artificial intelligence technologies, personalized learning systems are increasingly being applied in vocational training, higher education, and other fields. Traditional error-recommendation methods typically rely on single question similarity calculations or knowledge point tag matching, which suffers from prominent problems such as insufficient recommendation diversity, broken cognitive connections, poor adaptability to learning stages, and lagging dynamic feedback, thus seriously affecting user satisfaction. Therefore, how to develop an information recommendation method and system based on virtual users to improve recommendation accuracy and user satisfaction in complex learning scenarios is a technical problem that urgently needs to be solved by those skilled in the art.

[0003] Chinese Patent No. CN114676334A discloses a personalized intelligent exercise recommendation method and system. The method includes: binding any exercise resource to a corresponding position in a pre-constructed knowledge system; obtaining a user description file related to the user's answer results, and constructing a user vector related to the user's incorrect answers and a resource vector related to all exercise resources based on the description file; filtering the resource vectors based on the user vectors to obtain a first exercise set; filtering the first exercise set based on exercise quality to obtain a second exercise set; the exercise quality is simultaneously related to the exercise source, error rate, type, and search rate; and recommending the second exercise set to the user. This invention effectively improves the practice effect of exercises. However, the above technical solution has the following problems: it does not consider the different learning difficulties and different exercise practice methods corresponding to different learning stages, making it difficult for a fixed exercise recommendation method to effectively adapt to actual learning situations, thus affecting the user's learning effect and user satisfaction. Summary of the Invention

[0004] To address this, the present invention provides an information recommendation method and system based on virtual users, which overcomes the problem in the prior art that it does not take into account the different learning difficulties and different ways of correcting mistakes at different learning stages. This results in a fixed method of recommending mistakes that is difficult to adapt to the actual learning situation, thereby affecting the user's learning effect and user satisfaction.

[0005] To achieve the above objectives, the present invention provides an information recommendation method and system based on virtual users, comprising:

[0006] Based on the effective processing ratio, it is determined whether the exercise push meets the user's needs. For exercise pushes that meet the user's needs, knowledge points are expanded and supplemented. For exercise pushes that do not meet the user's needs, the reasons for not meeting the user's needs are determined based on the cumulative response volume. When it is determined that the exercise push does not meet the user's needs, a corresponding processing method is generated according to the determined reason, including: if the reason is that the related recommendation is missing, the related exercises are optimized based on the content matching degree; or if the reason is that the stage adaptation rate is low, the push interval for marked knowledge points is adjusted based on the memory strength increment.

[0007] The reason for not meeting user needs is that when the stage adaptation rate is low, the push interval is reduced based on the memory strength increment. Also, the size of a single push is adjusted based on the corrected push interval. Furthermore, the adjusted push interval is corrected based on the exam period proximity to reduce the adjusted push interval.

[0008] When the reason for not meeting user needs is the lack of related recommendations, optimize the push of related exercises based on content matching or reduce the number of knowledge points.

[0009] Once the adjustment is complete, the effective processing ratio of users is retrieved again. Based on the effective processing ratio, it is determined whether the exercise recommendations meet user needs. If they do not meet user needs, the number of knowledge points is readjusted. For exercise recommendations that meet user needs, the exercise recommendations are optimized based on the question diversity coefficient.

[0010] Furthermore, the process of determining whether the practice questions pushed to users meet their needs based on the effective processing ratio includes:

[0011] The comparison between the effective processing ratio and the preset effective processing ratio determines whether the exercise push meets the user's needs.

[0012] If the effective processing rate is greater than the preset effective processing rate, it is determined that the exercise push meets the user's needs, and supplemented with expanded knowledge points in subsequent exercise pushes.

[0013] If the effective processing percentage is less than or equal to the preset effective processing percentage, it is determined that the exercise push does not meet the user's needs, and the reason for not meeting the user's needs is determined based on the cumulative response volume.

[0014] Furthermore, the process of determining the reasons for non-compliance with user needs based on the cumulative response volume includes:

[0015] Obtain each user's incorrect and unprocessed questions, and mark these as invalid questions. Extract the knowledge points involved in these invalid questions.

[0016] The ratio of the number of questions containing this knowledge point to the total number of incorrect questions is recorded as the coverage percentage of a single knowledge point. The coverage percentage of each knowledge point is calculated, and the knowledge points are arranged in descending order of coverage percentage. The knowledge point with the largest fixed percentage of coverage percentage is recorded as the marked knowledge point, and the exercises corresponding to the marked knowledge point are recorded as marked exercises.

[0017] The average cumulative number of questions processed by users for each marked knowledge point is recorded as the cumulative response volume. The reasons for not meeting user needs are determined based on the cumulative response volume.

[0018] If the cumulative response count exceeds the preset cumulative response count, the cause is determined to be a lack of related recommendations, and related exercises will be optimized and pushed based on content matching degree.

[0019] If the cumulative response count is less than or equal to the preset cumulative response count, the reason is determined to be a low stage adaptation rate, and the push interval for marked knowledge points is adjusted based on the memory strength increment.

[0020] Furthermore, in response to the first preset condition, the process of determining the push interval for the marked knowledge points includes: adjusting the push interval by reducing it based on the memory intensity increment, and the adjustment range of the push interval is negatively correlated with the memory intensity increment;

[0021] The first preset condition is that the reason for not meeting the user's needs is the low stage adaptation rate.

[0022] Furthermore, the process of determining whether to adjust the size of a single push based on the comparison between the corrected push interval and the preset push interval includes:

[0023] The corrected push interval is compared with the preset push interval;

[0024] If the corrected push interval is less than or equal to the preset push interval, the size of a single push will be adjusted based on the effective interval difference.

[0025] The effective interval difference is the absolute value of the difference between the corrected push interval and the preset push interval.

[0026] Furthermore, during the adjustment of the push interval, the adjusted push interval is also corrected based on the proximity of the examination period to reduce the adjusted push interval, and the correction range of the push interval is negatively correlated with the proximity of the examination period.

[0027] Furthermore, the process of optimizing knowledge points based on content matching in response to the third preset condition includes:

[0028] Obtain the knowledge points covered by the recommended exercises and construct the corresponding knowledge point set K1; obtain the knowledge points marked by the user and construct the corresponding mark point set K2; calculate the content matching degree based on K1 and K2, wherein the content matching degree = |K1∩K2| / |K2|;

[0029] The number of knowledge points is reduced based on the difference in content matching degree, and the reduction in the number of knowledge points is positively correlated with the difference in content matching degree.

[0030] The third preset condition is that the reason for not meeting the user's needs is the lack of related recommendations.

[0031] Furthermore, the process of determining whether the exercise recommendation meets the user's needs in response to the fourth preset condition includes:

[0032] The percentage of effective processing for reacquiring users;

[0033] The effective processing ratio is compared with the preset effective processing ratio;

[0034] If the effective processing rate is greater than the preset effective processing rate, it is determined that the exercise push meets the user's needs, and supplemented with expanded knowledge points in subsequent exercise pushes.

[0035] If the effective processing percentage is less than or equal to the preset effective processing percentage, it is determined that the exercise push does not meet the user's needs, and the number of knowledge points is readjusted based on the content matching degree.

[0036] The fourth preset condition is to determine that the reason for not meeting the user's needs is the lack of related recommendations or the low stage adaptation rate, and accordingly complete the adjustment of the push interval for marked knowledge points based on the content matching degree optimization of push related exercises or based on the memory strength increment adjustment.

[0037] Furthermore, in response to the fifth preset condition, the question diversity coefficient corresponding to the exercise push is determined based on the decrease in the critical coverage ratio;

[0038] The decrease in the diversity coefficient of the topic and the critical coverage ratio are positively correlated.

[0039] The fifth preset condition is to complete the optimization of push notifications of related exercises based on content matching degree or the adjustment of the push interval for marked knowledge points based on memory strength increment, and to determine that the exercise push meets the user's needs.

[0040] The present invention also provides a system for applying the aforementioned information recommendation method based on virtual users, comprising:

[0041] The data collection unit is used to collect relevant parameters of the exercise delivery system, including the exercise completion rate and accuracy rate.

[0042] A processing unit, which is connected to the acquisition unit, is used to preprocess the relevant parameters of the exercise push system;

[0043] The analysis unit is used to determine whether the exercise push meets the user's needs based on the effective processing ratio, and to determine the reason why it does not meet the user's needs based on the cumulative response volume. The analysis unit is also used to generate corresponding processing instructions according to the determined reasons for non-compliance, including: when the reason is that the related recommendation is missing, optimizing the push of related exercises based on the content matching degree, or when the reason is that the stage adaptation rate is low, adjusting the push interval for marked knowledge points based on the memory intensity increment.

[0044] The control unit, which is connected to the analysis unit, is used to redetermine the user's effective processing ratio based on the received processing instructions, and adjust the corresponding push parameters to the corresponding values ​​according to the instructions.

[0045] Compared with the prior art, the beneficial effects of the present invention are as follows: by determining whether the exercise push meets user needs based on the effective processing ratio, the present invention can timely and efficiently determine whether the exercise push system meets user needs, effectively realizing real-time monitoring of the exercise push system, analyzing the reasons for non-compliance based on the cumulative response volume, generating corresponding adjustment methods according to the determined reasons for not meeting user needs, and adjusting parameters including push interval, number of knowledge points, and number of extended knowledge points. The adjustment parameters are re-determined based on the received processing instructions, and the corresponding adjustment parameters are adjusted to the corresponding values ​​according to the instructions. While effectively realizing accurate analysis of production line capacity, the production efficiency of the production line is effectively improved.

[0046] Furthermore, by determining whether the exercise recommendations meet user needs based on the effective processing ratio, this invention can quickly determine whether the exercise recommendations meet user needs, thereby improving the efficiency of determining whether the exercise recommendations meet user needs. In turn, when they do not meet the needs, the reasons can be quickly analyzed, thus improving the accuracy and rationality of the exercise recommendations.

[0047] Furthermore, by comparing the average cumulative processing volume of marked exercises for each user with the preset average cumulative processing volume, the present invention determines the reasons why the exercises do not meet the user's needs. Based on the historical processing volume, it can more accurately analyze the reasons why the exercises push does not meet the user's needs, and thus quickly analyze the reasons when the needs are not met, thereby improving the accuracy and rationality of the exercises push.

[0048] Furthermore, by adjusting the push interval based on memory strength increments, this invention avoids the problem that a fixed push interval is difficult to meet the needs of different users in actual use, thus leading to an unreasonable push interval setting. This further improves the rationality of the exercise recommendation method and enhances user satisfaction.

[0049] Furthermore, this invention determines whether to adjust the size of a single push based on the comparison between the corrected push interval and the preset push interval, thereby achieving a balance between push size and push frequency. This further improves the rationality of the exercise recommendation method and enhances user satisfaction.

[0050] Furthermore, by correcting the adjusted push interval based on the proximity of the exam date, this invention further ensures the accuracy of the push interval adjustment, thereby improving the rationality of the exercise recommendation method and enhancing user satisfaction.

[0051] Furthermore, this invention optimizes the push of related exercises by comparing the content matching degree with the preset content matching degree, thereby further improving the relevance between the pushed exercises. This not only enhances the rationality of the exercise recommendation method but also further improves user satisfaction.

[0052] Furthermore, by determining whether the number of knowledge points needs to be repeatedly adjusted based on the effective processing ratio and the preset effective processing ratio, the present invention avoids over-adjustment in the actual adjustment process, thereby further improving the rationality of the exercise recommendation method and enhancing user satisfaction.

[0053] Furthermore, by determining the question diversity coefficient corresponding to the question push based on the decrease in the critical coverage ratio, this invention further improves the rationality of question diversity, enhances the rationality of the question recommendation method, and improves user satisfaction. Attached Figure Description

[0054] Figure 1 This is a flowchart of the information recommendation method based on virtual users according to the present invention;

[0055] Figure 2 This is a unit connection diagram of the information recommendation system based on virtual users according to the present invention;

[0056] Figure 3 This is a flowchart illustrating the process of determining whether the exercise push meets user needs based on a comparison between the effective processing ratio and a preset effective processing ratio.

[0057] Figure 4 This is a flowchart illustrating the process of determining why a response does not meet user requirements based on a comparison between the cumulative response amount and a preset cumulative response amount. Detailed Implementation

[0058] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0059] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0060] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0061] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0062] Please see Figure 1 As shown, the present invention provides an information recommendation method based on virtual users, comprising:

[0063] Based on the effective processing ratio, it is determined whether the exercise push meets the user's needs. For exercise pushes that meet the user's needs, knowledge points are expanded and supplemented. For exercise pushes that do not meet the user's needs, the reasons for not meeting the user's needs are determined based on the cumulative response volume. When it is determined that the exercise push does not meet the user's needs, a corresponding processing method is generated according to the determined reason, including: if the reason is that the related recommendation is missing, the related exercises are optimized based on the content matching degree; or if the reason is that the stage adaptation rate is low, the push interval for marked knowledge points is adjusted based on the memory strength increment.

[0064] The reason for not meeting user needs is that when the stage adaptation rate is low, the push interval is reduced based on the memory strength increment. Also, the size of a single push is adjusted based on the corrected push interval. Furthermore, the adjusted push interval is corrected based on the exam period proximity to reduce the adjusted push interval.

[0065] When the reason for not meeting user needs is the lack of related recommendations, optimize the push of related exercises based on content matching or reduce the number of knowledge points.

[0066] Once the adjustment is complete, the effective processing ratio of users is retrieved again. Based on the effective processing ratio, it is determined whether the exercise recommendations meet user needs. If they do not meet user needs, the number of knowledge points is readjusted. For exercise recommendations that meet user needs, the exercise recommendations are optimized based on the question diversity coefficient.

[0067] Please see Figure 2 As shown, the present invention provides an information recommendation system based on virtual users, comprising:

[0068] The data collection unit is used to collect relevant parameters of the exercise delivery system, including the exercise completion rate and accuracy rate.

[0069] A processing unit, which is connected to the acquisition unit, is used to preprocess the relevant parameters of the exercise push system;

[0070] An analysis unit, which is connected to the acquisition unit and the processing unit respectively, is used to determine whether the exercise push meets the user's needs based on the effective processing ratio, and to determine the reason why it does not meet the user's needs based on the cumulative response volume. The analysis unit is also used to generate corresponding processing instructions according to the determined reasons for non-compliance, including: when the reason is that the related recommendation is missing, optimizing the push of related exercises based on the content matching degree, or when the reason is that the stage adaptation rate is low, adjusting the push interval for marked knowledge points based on the memory intensity increment.

[0071] The control unit, which is connected to the analysis unit, is used to redetermine the user's effective processing ratio based on the received processing instructions, and adjust the corresponding push parameters to the corresponding values ​​according to the instructions.

[0072] Please see Figure 3 The diagram illustrates the steps of determining whether exercise recommendations meet user needs based on the effective processing ratio in an embodiment of the present invention. The process includes: determining whether exercise recommendations meet user needs based on a comparison between the effective processing ratio and a preset effective processing ratio; if the effective processing ratio is greater than the preset effective processing ratio, then the exercise recommendations meet user needs, and supplementary knowledge points are added to subsequent exercise recommendations; if the effective processing ratio is less than or equal to the preset effective processing ratio, then the exercise recommendations do not meet user needs, and the reason for not meeting user needs is determined based on the cumulative response volume.

[0073] Specifically, in this embodiment, the effective processing percentage = exercise completion rate × exercise accuracy rate; exercise completion rate = number of exercises practiced / total number of exercises; exercise accuracy rate = number of correct questions / total number of completed exercises; the preset effective processing percentage P0 = 0.7, then the comparison result between the effective processing percentage and the preset effective processing percentage is as follows:

[0074] If the effective processing rate P is greater than the preset effective processing rate P0, it is determined that the exercise push meets the user's needs, and supplemented with extended knowledge points in subsequent exercise pushes.

[0075] If the effective processing percentage P is less than or equal to the preset effective processing percentage P0, it is determined that the exercise push does not meet the user's needs, and the reason for not meeting the user's needs is determined based on the cumulative response volume.

[0076] The preset effective processing ratio is not limited to this value, and those skilled in the art can adjust the value according to actual needs.

[0077] Please see Figure 4 As shown, the process of determining the reasons for not meeting user needs based on the cumulative response volume in an embodiment of the present invention includes: obtaining each user's incorrect questions and unprocessed exercises, and recording the incorrect questions and unprocessed exercises as invalid questions; extracting the knowledge points involved in the invalid questions; recording the ratio of the number of questions containing the knowledge point to the total number of incorrect questions as the coverage ratio of a single knowledge point; calculating the coverage ratio of each knowledge point; arranging each knowledge point in descending order of coverage ratio; recording the knowledge point types with the largest fixed proportion of coverage ratio as marked knowledge points; and recording the exercises corresponding to the marked knowledge points as marked exercises; recording the average cumulative processing volume of exercises for each marked knowledge point as the cumulative response volume; determining the reasons for not meeting user needs based on the cumulative response volume; if the cumulative response volume is greater than a preset cumulative response volume, determining the reason as missing related recommendations, and optimizing the push of related exercises based on content matching degree; if the cumulative response volume is less than or equal to the preset cumulative response volume, determining the reason as low stage adaptation rate, and adjusting the push interval for marked knowledge points based on memory strength increment;

[0078] The average cumulative processing volume is the average of the cumulative processing volume of each user for each of the marked knowledge points.

[0079] Specifically, in this embodiment, the preset cumulative response quantity N0 = 1000, and the comparison result between the cumulative response quantity N and the preset cumulative response quantity N0 is as follows:

[0080] If the cumulative response count N is greater than the preset cumulative response count N0, the reason is determined to be the lack of related recommendations, and related exercises are optimized and pushed based on the content matching degree.

[0081] If the cumulative response count N is less than or equal to the preset cumulative response count N0, the reason is determined to be low stage adaptation rate, and the push interval for marked knowledge points is adjusted based on the memory strength increment.

[0082] The fixed percentage value can be adaptively set by the user according to the actual application needs. It is understood that the higher the accuracy requirement of the user's judgment of the reasons for not meeting the user's needs, the larger the fixed percentage value will be. The present invention provides a fixed percentage value, in which the fixed percentage is 30% of the number of knowledge points.

[0083] The preset cumulative response amount and the fixed proportion are not limited to these values, and those skilled in the art can adjust these values ​​according to actual needs.

[0084] Specifically, the process of determining the push interval for marked knowledge points in response to a first preset condition in this embodiment of the invention includes: adjusting the push interval by reducing it based on the memory strength increment, and the adjustment range of the push interval is negatively correlated with the memory strength increment; the first preset condition is that the reason for not meeting the user's needs is a low stage adaptation rate.

[0085] Specifically, in this embodiment, the historical learning data of each user is obtained, and the accuracy of each knowledge point in the historical learning data is extracted. The accuracy is recorded as the historical accuracy. A time-accuracy curve is plotted, and the slope of the time-accuracy curve at the end of each fixed interval period is recorded as the user's memory intensity. The difference between the current time and the memory intensity at the end of the previous fixed interval period is recorded as the memory intensity increment.

[0086] The method for confirming the fixed interval period is as follows: the time between the initial time of the exercise recommendation system and the current time is recorded as the usage time, and the usage time is evenly divided into several fixed interval periods. The historical accuracy rate at the corresponding time at the end of each fixed interval period is detected. Within the fixed interval period, for the completed questions, the accuracy rate of questions involving each knowledge point is detected. The method for confirming the accuracy rate of each knowledge point is as follows: for any knowledge point, the average accuracy rate of the questions involving that knowledge point is set. For example, the accuracy rates of questions including knowledge point A are Y1, Y2, Y3...Yk, where Yk is the accuracy rate of the k-th question including knowledge point A.

[0087] Specifically, the process of determining whether to adjust the size of a single push based on the comparison between the corrected push interval and the preset push interval in this embodiment includes: comparing the corrected push interval with the preset push interval; if the corrected push interval is less than or equal to the preset push interval, then determining to adjust the size of a single push based on the effective interval difference; the effective interval difference is the absolute value of the difference between the corrected push interval and the preset push interval.

[0088] Specifically, in this embodiment, the preset push interval M0 = 5 days, and the comparison result between the push interval and the preset push interval is as follows:

[0089] If the corrected push interval M is less than or equal to the preset push interval M0, then the single push volume is adjusted based on the effective interval difference.

[0090] If the corrected push interval M is greater than the preset push interval M0, it is determined that there is room for correction of the push interval, and the push interval is corrected repeatedly.

[0091] If the effective interval difference is greater than the second preset effective interval difference △A2, the single push volume will be reduced to 90% of the initial push volume. In this embodiment, the second preset effective interval difference △A2 = 2 days.

[0092] If the effective interval difference is less than or equal to the second preset effective interval difference △A2 and greater than the first preset effective interval difference △A1, the single push volume is reduced to 85% of the initial push volume. In this embodiment, the second preset effective interval difference △A1 = 1 day.

[0093] If the effective interval difference is less than or equal to the first preset effective interval difference △A1, the single push volume will be reduced to 75% of the initial push volume;

[0094] In this embodiment of the invention, the initial push volume G0 = 30. If the reduced push volume is a decimal, it is rounded to the nearest integer.

[0095] Specifically, in the process of adjusting the push interval, this embodiment also corrects the adjusted push interval based on the proximity of the examination date to reduce the adjusted push interval, and the correction range of the push interval is negatively correlated with the proximity of the examination date.

[0096] Specifically, in this embodiment, the exam proximity = the number of days from the exam date to the current time / the total number of review days;

[0097] If the exam date proximity is greater than the preset exam date proximity C0, the push interval will be adjusted by 10%.

[0098] If the exam date proximity is less than or equal to the preset exam date proximity C0, the adjustment range of the push interval is 30%; in this embodiment, the preset exam date proximity C0 is 0.25.

[0099] Specifically, the process of optimizing knowledge points based on content matching degree in response to the third preset condition in this embodiment includes:

[0100] Obtain the knowledge points covered by the recommended exercises and construct the corresponding knowledge point set K1; obtain the knowledge points marked by the user and construct the corresponding mark point set K2; calculate the content matching degree based on K1 and K2, wherein the content matching degree = |K1∩K2| / |K2|;

[0101] The number of knowledge points is reduced based on the difference in content matching degree, and the reduction in the number of knowledge points is positively correlated with the difference in content matching degree.

[0102] The third preset condition is that the reason for not meeting the user's needs is the lack of related recommendations.

[0103] Specifically, in this embodiment, the preset matching degree G0 = 0.851; the process of reducing the number of knowledge points based on the content matching degree difference is as follows:

[0104] If the content matching degree difference is greater than the second preset content matching degree difference F2, then the reduction in the number of knowledge points is ΔV2 = 0.35, wherein, in this embodiment of the invention, the second preset content matching degree difference F2 = 0.251;

[0105] If the content matching degree difference is less than or equal to the second preset content matching degree difference F2 and greater than the first preset content matching degree difference F1, then the reduction in the number of knowledge points ΔV1 = 0.2. In this embodiment of the invention, the first preset content matching degree difference F1 = 0.151.

[0106] If the content matching degree difference is less than or equal to the first preset content matching degree difference F1, then the reduction in the number of knowledge points is ΔV0 = 0.05;

[0107] The content matching degree difference is the absolute value of the difference between the content matching degree and the preset content matching degree;

[0108] Specifically, the process of determining whether the exercise push meets the user's needs in response to the fourth preset condition in this embodiment includes: re-acquiring the user's effective processing ratio; comparing the effective processing ratio with the preset effective processing ratio; if the effective processing ratio is greater than the preset effective processing ratio, it is determined that the exercise push meets the user's needs, and supplementing and expanding knowledge points in subsequent exercise pushes; if the effective processing ratio is less than or equal to the preset effective processing ratio, it is determined that the exercise push does not meet the user's needs, and the number of knowledge points is readjusted based on the content matching degree; the fourth preset condition is to determine that the reason for not meeting the user's needs is the lack of related recommendations or the low stage adaptation rate, and correspondingly complete the adjustment of the push interval for marked knowledge points based on the content matching degree to optimize the push of related exercises or based on the memory intensity increment to adjust the push interval for marked knowledge points.

[0109] Specifically, in response to the fifth preset condition, this embodiment determines the question diversity coefficient corresponding to the exercise push based on the decrease in the critical coverage ratio; the question diversity coefficient is positively correlated with the decrease in the critical coverage ratio; the fifth preset condition is to complete the adjustment of the push interval for the marked knowledge points based on the content matching degree optimization or the adjustment based on the memory strength increment, and determine that the exercise push meets the user's needs.

[0110] Specifically, in this embodiment, the question diversity coefficient = α1 × the reduction rate of the critical coverage ratio; where α1 is an adjustment coefficient, and the value of α1 can be adaptively set by the user according to the actual application needs. It can be understood that the greater the impact of the reduction rate of the critical coverage ratio on the question diversity coefficient, the larger the value of α1. The present invention provides a value of α1, in which α1 = 1.24, and the minimum adjustment of the critical coverage ratio is 10% of the initial critical coverage ratio.

[0111] The initial critical coverage percentage is the minimum coverage percentage corresponding to a fixed number of values, which is recorded as the initial critical coverage percentage.

[0112] Specifically, the present invention also provides a system for applying the aforementioned information recommendation method based on virtual users, comprising:

[0113] The data collection unit is used to collect relevant parameters of the exercise delivery system, including the exercise completion rate and accuracy rate.

[0114] A processing unit, which is connected to the acquisition unit, is used to preprocess the relevant parameters of the exercise push system;

[0115] An analysis unit, which is connected to the acquisition unit and the processing unit respectively, is used to determine whether the exercise push meets the user's needs based on the effective processing ratio, and to determine the reason why it does not meet the user's needs based on the cumulative response volume. The analysis unit is also used to generate corresponding processing instructions according to the determined reason for not meeting the user's needs, including: when the reason is that the related recommendation is missing, optimizing the push of related exercises based on the content matching degree, or when the reason is that the stage adaptation rate is low, adjusting the push interval for marked knowledge points based on the memory intensity increment.

[0116] The control unit, which is connected to the analysis unit, is used to redetermine the user's effective processing ratio based on the received processing instructions, and adjust the corresponding push parameters to the corresponding values ​​according to the instructions.

[0117] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0118] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A virtual user-based information recommendation method, characterized by, include: Based on the effective processing ratio, it is determined whether the exercise push meets the user's needs. For exercise pushes that meet the user's needs, knowledge points are expanded and supplemented. For exercise pushes that do not meet the user's needs, the reasons for not meeting the user's needs are determined based on the cumulative response volume. When it is determined that the exercise push does not meet the user's needs, a corresponding processing method is generated according to the determined reason, including: if the reason is that the related recommendations are missing, the related exercises are optimized and pushed based on the content matching degree; or if the reason is that the stage adaptation rate is low, the push interval for marked knowledge points is adjusted based on the user's memory strength increment. The reason for not meeting user needs is that when the stage adaptation rate is low, the push interval is adjusted to be reduced based on the user's memory strength increment. Also, the size of a single push is determined based on the corrected push interval. Furthermore, the adjusted push interval is corrected based on the proximity of the exam period to reduce the adjusted push interval. When the reason for not meeting user needs is the lack of related recommendations, optimize the push of related exercises based on content matching or reduce the number of knowledge points. Once the adjustment is complete, the effective processing ratio of users is retrieved again. Based on the effective processing ratio, it is determined whether the exercise recommendations meet the user's needs. If they do not meet the user's needs, the number of knowledge points is readjusted. For exercise recommendations that meet the user's needs, the exercise recommendations are optimized based on the question diversity coefficient. This process involves obtaining each user's incorrect and unprocessed questions, marking these as invalid questions, and extracting the knowledge points involved in these invalid questions. The ratio of the number of questions containing this knowledge point to the total number of incorrect questions is recorded as the coverage percentage of a single knowledge point. The coverage percentage of each knowledge point is calculated, and the knowledge points are arranged in descending order of coverage percentage. The knowledge point with the largest fixed percentage of coverage percentage is recorded as the marked knowledge point, and the exercises corresponding to the marked knowledge point are recorded as marked exercises. The average cumulative number of questions processed by users for each marked knowledge point is recorded as the cumulative response volume.

2. The virtual user-based information recommendation method of claim 1, wherein, The process of determining whether exercise recommendations meet user needs based on the effective processing ratio includes: The comparison between the effective processing ratio and the preset effective processing ratio determines whether the exercise push meets the user's needs. If the effective processing rate is greater than the preset effective processing rate, it is determined that the exercise push meets the user's needs, and supplemented with expanded knowledge points in subsequent exercise pushes. If the effective processing percentage is less than or equal to the preset effective processing percentage, it is determined that the exercise push does not meet the user's needs, and the reason for not meeting the user's needs is determined based on the cumulative response volume. 3.The virtual user-based information recommendation method of claim 2, wherein, The process of determining why a response does not meet user needs based on the cumulative response volume includes: Determine the reasons for non-compliance with user needs based on the cumulative response volume; If the cumulative response count exceeds the preset cumulative response count, the cause is determined to be a lack of related recommendations, and related exercises will be optimized and pushed based on content matching degree. If the cumulative response count is less than or equal to the preset cumulative response count, the reason is determined to be a low stage adaptation rate, and the push interval for marked knowledge points is adjusted based on the user's memory strength increment.

4. The information recommendation method based on virtual users according to claim 3, characterized in that, In response to the first preset condition, the process of determining the push interval for marked knowledge points includes: adjusting the push interval by reducing it based on the user's memory strength increment, and the adjustment range of the push interval is negatively correlated with the user's memory strength increment; The first preset condition is that the reason for not meeting the user's needs is the low stage adaptation rate.

5. The information recommendation method based on virtual users according to claim 4, characterized in that, The process of determining whether to adjust the size of a single push based on the comparison between the corrected push interval and the preset push interval includes: The corrected push interval is compared with the preset push interval; If the corrected push interval is less than or equal to the preset push interval, the size of a single push will be adjusted based on the effective interval difference. The effective interval difference is the absolute value of the difference between the corrected push interval and the preset push interval.

6. The information recommendation method based on virtual users according to claim 5, characterized in that, During the adjustment of the push interval, the adjusted push interval is also corrected based on the proximity of the examination period to reduce the adjusted push interval, and the correction range of the push interval is negatively correlated with the proximity of the examination period.

7. The information recommendation method based on virtual users according to claim 3, characterized in that, The process of optimizing knowledge points based on content matching in response to the third preset condition includes: The recommended exercises cover the knowledge points and construct a corresponding knowledge point set K1. The user-marked knowledge points are also obtained and a corresponding mark point set K2 is constructed. Based on K1 and K2, the content matching degree is calculated. ; The number of knowledge points is reduced based on the difference in content matching degree, and the reduction in the number of knowledge points is positively correlated with the difference in content matching degree. The third preset condition is that the reason for not meeting the user's needs is the lack of related recommendations.

8. The information recommendation method based on virtual users according to claim 7, characterized in that, The process of determining whether the exercise recommendation meets the user's needs in response to the fourth preset condition includes: The percentage of effective processing for reacquiring users; The effective processing ratio is compared with the preset effective processing ratio; If the effective processing rate is greater than the preset effective processing rate, it is determined that the exercise push meets the user's needs, and supplemented with expanded knowledge points in subsequent exercise pushes. If the effective processing percentage is less than or equal to the preset effective processing percentage, it is determined that the exercise push does not meet the user's needs, and the number of knowledge points is readjusted based on the content matching degree. The fourth preset condition is to determine that the reason for not meeting the user's needs is the lack of related recommendations or the low stage adaptation rate, and accordingly complete the adjustment of the push interval for marked knowledge points based on the content matching degree to optimize the push of related exercises or based on the user's memory strength increment adjustment.

9. The information recommendation method based on virtual users according to claim 8, characterized in that, In response to the fifth preset condition, the question diversity coefficient corresponding to the exercise push is determined based on the decrease in the critical coverage ratio; The decrease in the diversity coefficient of the topic and the critical coverage ratio are positively correlated. The fifth preset condition is to complete the optimization of push notifications of related exercises based on content matching degree or the adjustment of the push interval for marked knowledge points based on the user's memory strength increment, and to determine that the exercise push meets the user's needs.

10. A system applied to the information recommendation method based on virtual users as described in any one of claims 1 to 9, characterized in that, include: The data collection unit is used to collect relevant parameters of the exercise delivery system, including the exercise completion rate and accuracy rate. A processing unit, which is connected to the acquisition unit, is used to preprocess the relevant parameters of the exercise push system; An analysis unit, which is connected to the acquisition unit and the processing unit respectively, is used to determine whether the exercise push meets the user's needs based on the effective processing ratio, and to determine the reason why it does not meet the user's needs based on the cumulative response volume. The analysis unit is also used to generate corresponding processing instructions according to the determined reason for not meeting the user's needs, including: when the reason is that the related recommendation is missing, optimizing the push of related exercises based on the content matching degree, or when the reason is that the stage adaptation rate is low, adjusting the push interval for marked knowledge points based on the user's memory strength increment. The control unit, which is connected to the analysis unit, is used to redetermine the user's effective processing ratio based on the received processing instructions, and adjust the corresponding push parameters to the corresponding values ​​according to the instructions.

Images