Learning time point recommendation method and device, electronic equipment and storage medium

Abstract

The invention discloses a learning time point recommendation method and device, electronic equipment and a computer readable storage medium, the method is applied to an intelligent learning system, the intelligent learning system comprises a first display interface and a second display interface, and at least one learning recommendation identifier is displayed on the first display interface, the learning recommendation identifier is used for indicating the target user to learn the target knowledge point at the target learning time point; and in response to the trigger operation for the at least one learning recommendation identifier, the learning content corresponding to the target learning time point is displayed on the second display interface, so that the learning efficiency of the user for a certain knowledge point can be improved.

Description

Technical Field

[0001] This disclosure relates to the field of data processing technology, and specifically to a method, apparatus, electronic device, and storage medium for recommending learning time points. Background Technology

[0002] Currently, users exhibit various learning behaviors during their learning process. Different types of learning behaviors, or the time intervals between learning behaviors, have different impacts on a user's mastery of a particular knowledge point. However, in actual learning, users often rely on subjective judgment or simple memorization to decide when to review a knowledge point, lacking a systematic understanding of their own learning behavior patterns and forgetting patterns. This often leads to inappropriate review timing. For example, reviewing too early wastes time, while reviewing too late results in significant forgetting of the knowledge, thus reducing the user's learning efficiency for that knowledge point. Summary of the Invention

[0003] This application provides a method, apparatus, electronic device, and storage medium for recommending learning time points, which can improve the learning efficiency of users for a certain knowledge point.

[0004] In a first aspect, embodiments of this application provide a learning time point recommendation method, applied to an intelligent learning system. The intelligent learning system includes a first display interface and a second display interface. The method includes:

[0005] At least one learning recommendation icon is displayed on the first display interface, wherein the learning recommendation icon is used to instruct the target user to learn the target knowledge point at the target learning time point; In response to a triggered action targeting at least one learning recommendation identifier, the learning content corresponding to the target learning time point is displayed on the second display interface.

[0006] In some embodiments, at least one learning recommendation icon is displayed on the first display interface, including: The memory change curve is displayed on the first display interface; A learning recommendation icon is displayed at at least one target learning time point on the memory change curve. The target learning time point is the time point in the future when the target user's memory strength of the target knowledge point meets the preset conditions after learning at least one learning content.

[0007] In some embodiments, the preset conditions include: The average memory strength of target users for the target knowledge points within the estimated time period is greater than a preset threshold; or, The increase in the intensity of memory of target knowledge points by target users at the target learning time point is higher than at other time points.

[0008] In some embodiments, after displaying the memory change curve on the first display interface, the method further includes: In response to a trigger operation targeting any time point on the memory change curve, a memory prediction curve starting from the triggered time point is displayed on the first display interface. The memory prediction curve predicts the change in the memory strength of the target knowledge point after the target user performs at least one learning behavior at the triggered time point.

[0009] In some embodiments, displaying a memory prediction curve starting from the triggered time point on a first display interface includes: At least one candidate learning content is displayed at the triggered time node on the first display interface; In response to a selection instruction for at least one candidate learning content, a memory prediction curve generated based on the selected candidate learning content is displayed on a first display interface.

[0010] In some embodiments, displaying a memory change curve on a first display interface includes: In response to a target user's memory review event for a target knowledge point, the first display screen shows the target user's memory change curve during the target learning period, which includes historical and future time periods.

[0011] In some embodiments, in response to a target user's memory review event for a target knowledge point, a memory change curve of the target user during the target learning period is displayed on a first display interface, including: In response to a knowledge point viewing action triggered by a target user, display at least one target knowledge point, the target user's historical learning events for the target knowledge point at at least one historical moment, and information on the contribution of the historical learning events to the target user's memory of the target knowledge point; In response to the target user's memory trigger action on the target knowledge point, the system displays the memory change curve of the target user during the target learning period.

[0012] In some embodiments, the method further includes: Send review reminders to target users at the target learning time to instruct them to process the reminders at that time.

[0013] Secondly, embodiments of this application provide a method for recommending learning time points, the method comprising: After determining that the target user performs at least one learning behavior at any time point on the memory change curve, the memory prediction curve for the target knowledge point is obtained. The memory prediction curve is the change in the memory intensity of the target user for the target knowledge point within the estimated time period, which is generated based on the learning behavior. The time points at which changes in memory intensity meet preset conditions are used as target learning time points, so that learning recommendations can be made to target users based on the target learning time points.

[0014] In some embodiments, the time point at which the change in memory intensity meets preset conditions is used as the target learning time point, including: By comparing the changes in memory intensity on the memory change curve with the changes in memory intensity on the memory prediction curve at each time point, the magnitude of the change in memory intensity at each time point can be determined. Based on the magnitude of changes in memory intensity at each time point, time points that meet preset conditions are selected as target learning time points.

[0015] In some embodiments, based on the magnitude of memory intensity change at each time point, time points that meet preset conditions are selected as target learning time points, including: The time point with the greatest change in memory intensity is selected as the target learning time point.

[0016] In some embodiments, the memory intensity change of the memory change curve and the memory intensity change of the memory prediction curve at each time point are compared to determine the magnitude of the memory intensity change at each time point, including: Calculate the difference between the memory intensity of the memory change curve and the memory intensity of the memory prediction curve at a specified time point within the estimated time period; The difference at a specified moment corresponding to each time node is taken as the magnitude of change in memory strength; or Calculate the difference between the memory intensity of the memory change curve and the memory intensity of the memory prediction curve at each time point within the estimated time period. The differences between the time windows corresponding to each time node are averaged to use the result as the magnitude of change in memory intensity.

[0017] In some embodiments, determining the memory prediction curve for a target knowledge point after the target user performs at least one learning action at any time point on the memory change curve includes: Based on learning behavior and time points, determine the memory contribution information of learning behavior at each estimated time point within the estimated time period; Based on memory contribution information, the memory strength of the target user for the target knowledge point is determined at each estimated time point, and a memory prediction curve for the target knowledge point is obtained based on the memory strength at each estimated time point.

[0018] In some embodiments, based on learning behavior and time points, memory contribution information of the learning behavior at each estimated time point within the estimated time period is determined, including: Obtain memory-influencing parameters corresponding to learning behaviors; Based on the memory impact parameters and time points of learning behavior, as well as the estimated time points, we determine the contribution information of learning behavior to the memory of target knowledge points at the estimated time points.

[0019] In some embodiments, the memory impact parameters include target decay intensity and target excitation intensity, memory impact parameters based on learning behavior and time points, and estimated time points, determining the contribution information of learning behavior to the memory of target knowledge points at the estimated time points, including: Determine the time difference between the predicted and actual time points; Based on the time difference and the target decay intensity, determine the decay indication parameters corresponding to the learning behavior; The product of the target stimuli intensity and the decay indicator parameter of the learning behavior is used as the information on the contribution of the learning behavior to the memory of the target knowledge point at the predicted time point.

[0020] In some embodiments, the method further includes: Obtain the target user's historical learning trajectory, which includes multiple learning behavior samples and memory labels for each learning behavior sample; Based on the target user's historical learning trajectory, each learning behavior sample and its memory label are fitted to determine the activation intensity and decay intensity corresponding to the learning behavior of each learning behavior sample of the target user.

[0021] In some embodiments, determining the memory strength of a target user for a target knowledge point at each estimated time point based on memory contribution information includes: Obtain information on the historical memory contribution of at least one historical learning behavior of the target user at the estimated time point; By accumulating historical memory contribution information and memory contribution information corresponding to the estimated time points, the memory strength of the target user for the target knowledge point at the estimated time point is obtained.

[0022] Thirdly, embodiments of this application provide a learning time point recommendation device, applied to an intelligent learning system. The intelligent learning system includes a first display interface and a second display interface. The device includes: The identifier display module is used to display at least one learning recommendation identifier on the first display interface, wherein the learning recommendation identifier is used to instruct the target user to learn the target knowledge point at the target learning time point; The content display module is used to respond to a trigger operation for at least one learning recommendation identifier to display learning content corresponding to the target learning time point on the second display interface.

[0023] Fourthly, embodiments of this application provide a learning time point recommendation device, the device comprising: The curve determination module is used to determine the memory prediction curve for the target knowledge point obtained after the target user performs at least one learning behavior at any time node on the memory change curve. The memory prediction curve is the change in the memory intensity of the target user for the target knowledge point within the estimated time period based on the learning behavior. The node determination module is used to identify time points where changes in memory intensity meet preset conditions as target learning time points, and to make learning recommendations to target users based on these target learning time points.

[0024] Fifthly, embodiments of this application also provide an electronic device, including a memory storing multiple instructions; a processor loading instructions from the memory to execute the steps of any of the learning time point recommendation methods provided in embodiments of this application.

[0025] Sixthly, embodiments of this application also provide a computer-readable storage medium storing a plurality of instructions adapted for loading by a processor to execute the steps of any of the learning time point recommendation methods provided in embodiments of this application.

[0026] In a seventh aspect, embodiments of this application also provide a computer program product, including a computer program or instructions, which, when executed by a processor, implement the steps in any of the learning time point recommendation methods provided in embodiments of this application.

[0027] The solution adopted in this application embodiment can be applied to an intelligent learning system. The intelligent learning system includes a first display interface and a second display interface. By displaying at least one learning recommendation icon on the first display interface, wherein the learning recommendation icon is used to instruct the target user to learn the target knowledge point at a target learning time point; in response to a trigger operation on the at least one learning recommendation icon, learning content corresponding to the target learning time point is displayed on the second display interface, thereby recommending learning about the target knowledge point to the target user at an appropriate time through the learning recommendation icon, so as to improve the user's learning efficiency of a certain knowledge point. Attached Figure Description

[0028] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0029] Figure 1 This is a flowchart illustrating the first embodiment of the learning time point recommendation method provided in this application. Figure 2 This is a schematic diagram of the memory change curve provided in the embodiments of this application; Figure 3 This is a schematic diagram of the memory prediction curve provided in the embodiments of this application; Figure 4 This is a schematic diagram of the learning recommendation identifier provided in the embodiments of this application; Figure 5 This is a schematic diagram of an interface display provided in an embodiment of this application; Figure 6 This is a flowchart illustrating a second embodiment of the learning time point recommendation method provided in this application. Figure 7 This is a flowchart illustrating the first embodiment of the knowledge point interaction method provided in this application. Figure 8 This is a flowchart illustrating a second embodiment of the knowledge point interaction method provided in this application. Figure 9 This is a schematic diagram of the structure of the first learning time point recommendation device provided in the embodiments of this application; Figure 10 This is a schematic diagram of the structure of the second learning time point recommendation device provided in the embodiments of this application; Figure 11 This is a schematic diagram of the structure of the first knowledge point interaction device provided in the embodiments of this application; Figure 12 This is a schematic diagram of the structure of the second type of knowledge point interaction device provided in the embodiments of this application; Figure 13 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0030] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. At the same time, in the description of the embodiments of this application, the terms "first," "second," etc., are only used to distinguish descriptions and should not be construed as indicating or implying relative importance. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0031] This application provides a method, apparatus, electronic device, and computer-readable storage medium for recommending learning time points.

[0032] Specifically, this embodiment will be described from the perspective of a learning time point recommendation device. This learning time point recommendation device can be integrated into an electronic device, that is, the learning time point recommendation method of this application embodiment can be executed by an electronic device. Optionally, the electronic device may include a terminal device. The terminal device may be a mobile phone, tablet computer, smart Bluetooth device, laptop computer, game console, or personal computer (PC), etc.

[0033] The learning time point recommendation method provided in this application can be applied to a learning time point recommendation system. This system can include a player terminal device and a server. The terminal can be a device that includes both receiving and transmitting hardware, i.e., a device with receiving and transmitting hardware capable of performing bidirectional communication over a bidirectional communication link. The player terminal device and the server can communicate bidirectionally via a network.

[0034] Optionally, the server can be a standalone server, or a server network or server cluster, including but not limited to computers, network hosts, single network servers, multiple network server sets, or cloud servers composed of multiple servers. Cloud servers consist of a large number of computers or network servers based on cloud computing.

[0035] The following detailed description is provided in conjunction with the accompanying drawings. In this embodiment, the execution subject is a terminal device as an example. It should be noted that the order of description in the following embodiments is not intended to limit the preferred order of the embodiments. Although a logical order is shown in the flowcharts, in some cases, the steps shown or described may be performed in a different order than that shown in the accompanying drawings.

[0036] The learning time recommendation method in this embodiment can improve the learning efficiency of users for a certain knowledge point.

[0037] Please refer to Figure 1 Taking a terminal as an example, this embodiment provides a learning time point recommendation method applied to an intelligent learning system. The intelligent learning system includes a first display interface and a second display interface. The specific process of the learning time point recommendation method can be described in the following steps 101-102, wherein: Step 101: Display at least one learning recommendation icon on the first display interface, wherein the learning recommendation icon is used to instruct the target user to learn the target knowledge point at the target learning time point.

[0038] The target knowledge point can be the knowledge point that the target user has already learned.

[0039] The learning recommendation identifiers mentioned above can be presented in styles such as text, icons, and characters.

[0040] In this embodiment, a learning recommendation icon is displayed to indicate the target knowledge point that the target user needs to learn now, as well as the target learning time point for the target user to learn the target knowledge point. This is to help the target user clarify the knowledge point that needs to be reviewed now through the target knowledge point, and to help the target user clarify the time when to learn the knowledge point through the target learning time point.

[0041] It should be noted that when two or more learning recommendation icons are displayed on the first display interface, these icons can instruct the target user to learn the same target knowledge point at different target learning time points, or to learn different target knowledge points at the same target learning time point, or to learn different target knowledge points at different target learning time points, etc. The specific settings can be configured according to the needs, and are not limited here.

[0042] Step 102: In response to a triggering operation for at least one learning recommendation identifier, display the learning content corresponding to the target learning time point on the second display interface.

[0043] The triggering operation can be a click operation, such as a single click or double click, or a press operation, such as a press for a preset duration. The specific settings can be configured according to the requirements and are not limited here.

[0044] In this embodiment, by triggering the learning recommendation icon, the learning content of the target knowledge point corresponding to the target learning time point indicated by the learning recommendation icon can be displayed on the second display interface.

[0045] The first and second display interfaces can be displayed simultaneously, or they can be displayed sequentially according to a specific logical order. The specific order can be set according to the requirements and is not limited here.

[0046] As can be seen from the above, when applied to an intelligent learning system, the intelligent learning system includes a first display interface and a second display interface. Furthermore, by displaying at least one learning recommendation icon on the first display interface, which instructs the target user to learn the target knowledge point at a target learning time, and responding to a trigger operation on the at least one learning recommendation icon, the system displays learning content corresponding to the target learning time on the second display interface. This allows for timely learning recommendations of the target knowledge point to the target user through the learning recommendation icon, thereby improving the user's learning efficiency for that knowledge point.

[0047] In some embodiments, at least one learning recommendation icon is displayed on the first display interface, including: The memory change curve is displayed on the first display interface; A learning recommendation icon is displayed at at least one target learning time point on the memory change curve. The target learning time point is the time point in the future when the target user's memory strength of the target knowledge point meets the preset conditions after learning at least one learning content.

[0048] The memory change curve represents the change in the target user's memory strength of the target knowledge point over a period of time. This curve can be generated based on the target user's historical learning events in a historical period. The memory strength change can be composed of memory strength at multiple different time points. This memory strength is used to indicate the target user's mastery of the target knowledge point; the greater the memory strength, the deeper the target user's memory of the target knowledge point.

[0049] It should be noted that since memory strength can indicate the target user's mastery of the target knowledge point, it can be indicated by percentages, fractions, decimals, etc. For example, when the target user's memory strength for the target knowledge point is 100%, it means that the target user has fully mastered the target knowledge point.

[0050] The preset conditions can be conditions corresponding to the user's memory strength reaching or about to reach the level of mastery of the target knowledge point, or conditions corresponding to the user's memory strength change being in line with the user's high memory benefit, etc. The specific settings can be set according to the needs and are not limited here.

[0051] In this embodiment, by displaying a corresponding learning recommendation icon at the target learning time point where the memory strength meets the corresponding preset conditions, the system recommends that users learn the relevant target knowledge points at time points with higher memory benefits. This allows users to intuitively understand the appropriate time to learn the target knowledge points and ensures that the memory strength after learning at the target learning time point indicated by the learning recommendation icon meets the preset conditions, thereby improving the user's learning efficiency of the target knowledge points.

[0052] Furthermore, by identifying target learning time points and recommending learning methods to users, this approach avoids the current learning lag caused by only reviewing knowledge points when users have not been exposed to them for a long time or when they answer questions related to a knowledge point incorrectly. This lack of adaptive prompts based on the user's forgetting process prevents intervention at the optimal time.

[0053] In some embodiments, the preset conditions include: The average memory strength of target users for the target knowledge points within the estimated time period is greater than a preset threshold; or, The increase in the intensity of memory of target knowledge points by target users at the target learning time point is higher than at other time points.

[0054] It is understandable that when a target user learns the target knowledge point at different time points, it can change the target user's memory strength of the target knowledge point. By comparing the memory strength with a preset threshold, or by comparing the increase in memory strength with the increase in memory strength at other time points, it can be determined whether there is a time point where the memory strength meets the preset conditions, namely the aforementioned target learning time point, and thus automatically determine the best review time point.

[0055] Specifically, the increase in memory intensity can be obtained by calculating the difference between the memory intensity at any moment within the estimated time period and the memory intensity at the corresponding moment on the memory change curve. The specific settings can be configured according to requirements and are not limited here.

[0056] Specifically, the increase in memory intensity can be obtained by determining the increase in memory intensity between each moment (e.g., each day) in the estimated time period and the corresponding moment on the memory change curve, and then calculating the average of the increases in memory intensity across all moments.

[0057] In some embodiments, as the time interval between the target user's most recent contact with the target knowledge point increases, the target user's memory strength of the target knowledge point decreases. Therefore, the terminal can also introduce a specific threshold to dynamically monitor the target user's memory of the target knowledge point, thereby identifying the corresponding target learning time point for which learning recommendations need to be made to the target user.

[0058] Specifically, the target learning time point can also be the node where the target user's memory strength of the target knowledge point is less than a specific threshold. That is, if the target user's memory strength of a certain knowledge point drops below the specific threshold, the target learning time point can be used as the time node to trigger the review recommendation. This allows the review rhythm to be automatically scheduled according to the target user's forgetting trend of the target knowledge point, so as to obtain the best review time in advance before the target user completely forgets the target knowledge point, reduce unnecessary review time investment, improve the review hit rate, avoid ineffective or excessive review, and at the same time reduce the risk of the target user forgetting the target knowledge point.

[0059] In some embodiments, in order to make it easier for target users to intuitively understand the learning benefits they have gained at the target learning time point, learning benefit information can be displayed at the learning recommendation label, or the learning recommendation label may contain learning benefit information.

[0060] The learning benefit information can indicate the learning benefits of a target user at the target learning time point indicated by the learning recommendation icon. These learning benefits include, but are not limited to, changes in the target user's memory strength (such as an increase) within the estimated time period, and the duration during which the target user's memory strength is not lower than a specific threshold.

[0061] For example, if a target user learns at least one question containing that knowledge point three days later, and their memory strength in the next three months is not less than a certain threshold, then a time point three days apart from the current time can be used as the target learning time point, and the learning benefit information "The target user will not forget the knowledge point in the next three months" can be displayed for that target learning time point.

[0062] For example, if the target learning time point is obtained when the memory strength is less than a certain threshold, then the learning benefit information can also indicate the triggering reason for the target learning time point, such as the last learning time being X days ago and the expected forgetting risk being Y%. Natural language explanations can be output, such as "You last learned this knowledge point 5 days ago. The current memory strength has dropped below the set threshold. The system suggests reviewing it."

[0063] In this embodiment, by further displaying learning benefit information, the interpretability of the recommendation results is improved. Through visual learning recommendation indicators (such as displaying memory change curves) and natural language trigger descriptions, it is possible to clearly explain why a certain knowledge point is recommended for review at a certain time point, thereby enhancing system transparency, improving the understanding, trust, or acceptance of target users (such as teachers and students), and enhancing the learning initiative of target users.

[0064] Specifically, learning benefit information can be generated through prompt-based natural language generation, that is, based on rule-based or fine-tuned language models, chat-like explanatory text can be generated to improve the human-computer interaction experience of the system.

[0065] In some embodiments, after displaying the memory change curve on the first display interface, the method further includes: In response to a trigger operation targeting any time point on the memory change curve, a memory prediction curve starting from the triggered time point is displayed on the first display interface. The memory prediction curve predicts the change in the memory strength of the target knowledge point after the target user performs at least one learning behavior at the triggered time point.

[0066] In this embodiment, since the target user's learning behavior related to the target knowledge point at different time points will lead to changes in the target user's memory strength of the target knowledge point, the terminal predicts the change in the user's memory strength if the user performs learning behavior related to the target knowledge point at the triggered time point by responding to the user's triggering operation at a time point on the memory change curve. This generates a memory prediction curve starting from the triggered time point, prompting the user to understand what kind of change will occur in their memory strength of the target knowledge point if they perform learning behavior for the target knowledge point at the triggered time point.

[0067] The memory prediction curve covers a learning period that includes a time node and at least one time node after that time node.

[0068] For example, such as Figure 2 As shown, in Figure 2 The graph shows the changes in memory, with the horizontal axis representing time and the vertical axis representing memory intensity. Figure 2 The curve between point A and point B, from Figure 2 As can be seen, Figure 2 Point C in the model represents the time node of the last specific learning event of the target user on this knowledge point. Each point before (including) point C represents the date on which the target user actually engaged in learning behavior and the strength of their memory of the knowledge point after learning. Points after point C represent the changes in the strength of the target user's memory of the knowledge point if they do not continue learning after point C, as predicted by the model.

[0069] Among them, from Figure 2 It is evident that during the time period following point C, which includes future time periods, the target user's memory of the relevant knowledge points will decrease over time because the target user will not engage in any further learning activities related to the knowledge points. In other words, the target user's memory of the relevant knowledge points will become increasingly weaker.

[0070] If a user wants to see how their memory strength changes after learning about a specific knowledge point at a future time, they can... Figure 2The trigger operation is performed at any time point after point C on the memory change curve, such as... Figure 3 As shown, the user triggers an action at point D to simulate the target user performing at least one learning behavior at point D. The goal is to predict the target user's memory strength after point D, and then generate... Figure 3 The memory change curve between point D and point E is used as the memory prediction curve.

[0071] Among them, from Figure 3 It is evident that because the target user performs at least one learning action at point D, the target user's memory strength shows an upward trend. However, in the time period after point D, which includes the future time period, since the target user does not perform any more learning events related to the corresponding knowledge points, the memory strength after point D, which increases for a period of time, will decrease over time. In other words, the target user's memory strength of the corresponding knowledge points becomes lower and lower.

[0072] If, within a future timeframe on the memory prediction curve, there are time points where memory strength meets preset conditions, such as... Figure 4 As shown, Figure 4 Point F in the equation represents the target learning time point. Therefore, the terminal can simulate the target user performing at least one learning action at point F, predict the memory strength of the target user after performing the learning action related to the target knowledge point at point F, and generate [the necessary data / information]. Figure 4 The memory change curve between point F and point G can be used as a learning recommendation indicator, or a preset icon can be displayed at point F as a learning recommendation indicator. Then, when the user triggers point F, the memory change curve between point F and point G can be displayed. The specific settings can be configured according to the needs, and there are no restrictions here.

[0073] Among them, from Figure 4 It is evident that after performing at least one learning action at point F, the target user can fully master the target knowledge point or master it to a certain extent, such as... Figure 4 The memory strength of the target users reaches 100%, at least in the target users Figure 4 The target knowledge point will not be forgotten during the time period corresponding to point F to point G.

[0074] Optionally, after learning a certain knowledge point, the target user can use the system to calculate the memory change curve, and by triggering a certain time point on the memory change curve, calculate the change in memory intensity of the knowledge point when reviewing at a future time point. Based on the change in memory intensity at a specific time point in the memory change curve, the system can help understand the source of the review suggestions and encourage the target user to make corresponding learning plans.

[0075] In some embodiments, displaying a memory prediction curve starting from the triggered time point on a first display interface includes: At least one candidate learning content is displayed at the triggered time node on the first display interface; In response to a selection instruction for at least one candidate learning content, a memory prediction curve generated based on the selected candidate learning content is displayed on a first display interface.

[0076] Among them, the candidate learning content is learning content related to the target knowledge points, such as different types of test questions, detailed explanations of knowledge points, etc.

[0077] In this embodiment, by displaying candidate learning content at the triggered time node, the user can select the content they want to learn at the triggered time node, thereby triggering the above selection instruction and prompting the terminal to clearly understand the learning content that the target user expects to learn at the triggered time node. The above learning behavior is the behavior of learning the learning content.

[0078] Then, since different learning content contributes differently to the memory of the target knowledge point, the change in the memory strength of the target knowledge point after learning different learning content will vary. Therefore, the candidate learning content selected by the user at the triggered time node can be used as the learning content to be learned at the time node, so as to predict the change in the memory strength of the target knowledge point after the target user learns the selected candidate learning content at the triggered time node.

[0079] Specifically, the above selection instructions can be generated by triggering operations on the displayed candidate learning content. These triggering operations can be click operations, such as single-click or double-click operations, or press operations, such as press for a preset duration, or swipe operations, etc. The specific settings can be configured according to the requirements and are not limited here.

[0080] In some embodiments, displaying a memory change curve on a first display interface includes: In response to a target user's memory review event for a target knowledge point, the first display screen shows the target user's memory change curve during the target learning period, which includes historical and future time periods.

[0081] In this embodiment, the terminal needs to respond to the memory viewing event and display the memory change curve of the target user for the target knowledge point within the target learning time period, so that the user can operate on the memory change curve.

[0082] Among them, the memory viewing event is used to trigger the display of changes in the user's memory strength of the target knowledge point during the memorization process. This memory viewing event can be triggered by the user's behavior or by a pre-set triggering mechanism on the terminal. The specific settings can be configured according to the needs and are not limited here.

[0083] The target learning time period is the time period in the memory change curve that shows the change in the user's memory intensity of the target knowledge point. In this target learning time period: the historical time period is the historical period before the current moment, and the future time period is the future period after the current moment.

[0084] It should be noted that the memory change curve contains multiple time points and the memory intensity corresponding to each time point. This memory change curve can indicate the change in the memory intensity of target knowledge points after the target user conducts historical learning in a real-world scenario.

[0085] The historical time period can include at least a portion of the time period between the moment when the target user first learns the target knowledge point and the current moment. The future time period can include a time period consisting of a certain duration after the current moment. The division of the historical time period and the future time period can be set according to the needs and is not limited here.

[0086] Specifically, memory change curves can be obtained through interpretable models, such as linear point process models and logistic regression models. Interpretable models are used to generate and provide clear causal explanation structures.

[0087] In some embodiments, the end time of the target learning time period, that is, the end time of the future time period, can be determined by the time when the interval between the initial time and the initial time meets the preset time threshold, or by the time when the corresponding memory strength meets the preset threshold condition.

[0088] Specifically, the end point can be the time when the memory strength is less than the preset forgetting threshold. For example, if the target user has forgotten 95% of the target knowledge point, that is, when the target user's memory strength is 5%, it can be considered that the target user only remembers the key words in the target knowledge point and is about to forget the target knowledge point. This time point can be determined as the end point.

[0089] Specifically, the end point can be the time when the memory intensity is greater than the preset memory threshold. For example, the time when a target user has completely mastered a knowledge point due to a certain learning behavior. For instance, after watching a course, based on historical learning events, it can be inferred that the target user has completely mastered or is about to master a knowledge point. If the memory intensity of the knowledge point reaches 92%, and this 90% is greater than the preset memory threshold, then this time point can be determined as the end point.

[0090] The preset memory threshold is greater than the preset forgetting threshold. The preset forgetting threshold and the preset memory threshold can be set according to needs, and are not limited here.

[0091] In some embodiments, in response to a target user's memory review event for a target knowledge point, a memory change curve of the target user during the target learning period is displayed on a first display interface, including: In response to a knowledge point viewing action triggered by a target user, display at least one target knowledge point, the target user's historical learning events for the target knowledge point at at least one historical moment, and information on the contribution of the historical learning events to the target user's memory of the target knowledge point; In response to the target user's memory trigger action on the target knowledge point, the system displays the memory change curve of the target user during the target learning period.

[0092] In this embodiment, by displaying the target user's historical learning events for the target knowledge point, and the corresponding memory contribution information of the target user under the historical learning events, the user can intuitively understand the overall learning process of the target user for the target knowledge point at the current moment, and the impact of each learning event on the user's memory and mastery of the target knowledge point during the overall learning process.

[0093] The knowledge point viewing operation is used to trigger the viewing of at least one target knowledge point and other information related to the target knowledge point. This knowledge point viewing operation can be triggered by the user's behavior or by a pre-set triggering mechanism on the terminal. The specific settings can be configured according to the needs and are not limited here.

[0094] Among them, historical learning events can be events generated by a user's learning behavior on a certain learning content in a historical period. Such historical learning behavior includes, but is not limited to, the user's behavior of browsing, learning, or reviewing the learning content of the target knowledge point. For example, the target user's behavior of viewing and / or answering questions related to the target knowledge point.

[0095] Among them, the memory contribution information is used to indicate the contribution of a user's learning behavior to the memory of the target knowledge point. This memory contribution information can be represented by numerical values ​​such as percentages, fractions, and decimals, or by text, characters, etc. The specific settings can be set according to the needs and are not limited here.

[0096] Specifically, the memory contribution information can be expressed as a percentage, which indicates that the target user has a deeper memory of the target knowledge point after the corresponding learning behavior. For example, the memory contribution rate corresponding to the target user browsing the definition of the target knowledge point is 30%, while the memory contribution rate corresponding to the target user answering questions related to the target knowledge point is 40%.

[0097] It is understandable that the memory contribution information of each historical learning event can affect the user's memory strength of the target knowledge point over time and with the accumulation of historical learning events.

[0098] Specifically, the terminal can use the memory contribution information corresponding to different historical learning events of the target user to assess the changes in the target user's memory intensity for the target knowledge point during the target learning period. This will improve the accuracy of the assessment results when evaluating the target user's mastery of a certain knowledge point, and enable the target user to intuitively and accurately understand the memory change curve displayed in the graphical target user interface, indicating the target user's memory mastery of the target knowledge point.

[0099] It should be noted that when displaying the memory change curve of the target user for the target knowledge point, the memory intensity on the memory change curve will change over time. If the target user reviews frequently, the memory intensity of the target knowledge point will be improved. However, if the target user does not review for a long time, the memory intensity of the target knowledge point may decrease before the target user fully masters the target knowledge point.

[0100] In some embodiments, since a target user may have multiple historical learning events for a target knowledge point, these historical learning events can be arranged according to preset behavior arrangement rules. For example, based on the learning time of different historical learning events, multiple historical learning events can be arranged in chronological order, i.e., the later the learning time, the earlier the historical learning event is arranged. Or, based on the memory contribution information of different historical learning events, multiple historical learning events can be arranged from largest to smallest, i.e., the larger the memory contribution information, the earlier the historical learning event is arranged. The specific arrangement can be set according to the needs, and there are no further limitations.

[0101] In some embodiments, target knowledge points and their historical learning events can be displayed together on a graphical target user interface, and / or, historical learning events and their corresponding memory contribution information can be displayed together on a graphical target user interface, wherein the display of both includes, but is not limited to, being displayed through lists, bar charts, text, etc.

[0102] In some embodiments, the terminal may be configured with a knowledge point forgetting query function. By triggering the knowledge point forgetting query function, a knowledge point viewing operation can be realized. When the knowledge point forgetting query function is triggered, the terminal may display a function interface. The function interface may include at least a first display area and a second display area. The first display area is used to display at least one target knowledge point, while the second display area may display at least one historical learning event related to the target knowledge point, as well as information on the contribution of the historical learning event to the target user's memory of the target knowledge point.

[0103] The display format of the target knowledge points includes, but is not limited to: sorting the target knowledge points based on the target user's memory strength of the target knowledge points at the current moment, and displaying the sorted target knowledge points; at each target knowledge point, the target user's memory strength of the target knowledge point at the current moment is displayed in association, and different memory strengths can be displayed in different display styles, such as numerical values ​​of different colors, bar charts, etc.

[0104] Furthermore, when displaying at least one historical learning event of the target knowledge point in the second display area, the terminal can receive a selection operation of the target knowledge point in the first display area from the target user, so as to display the target knowledge point indicated by the selection operation and at least one historical learning event of the target knowledge point indicated by the selection operation in the second display area.

[0105] For example, such as Figure 5 As shown, in Figure 5 In the middle, the target knowledge point indicated by the selection operation is set as "Chapter XX, Section XX: Electric Charge and its Conservation Law", and in Figure 5 The window shown displays the target user's historical learning events related to "Chapter XX, Section XX: Electric Charge and Its Conservation Law," for example, Figure 5 The historical learning events include: "XX City College Entrance Examination Prediction Paper Question XX: 40%" held at 14:33 on February 16, 2025; "XX Practice Questions and their Analysis Question XX: 30%" held at 9:45 on March 3, 2025; "XX City Textbook Page XX Question XX: 40%" held at 10:22 on June 26, 2025; and "XXXXXX: XX%" which occurred on XX Month XX Day, XXXX Year at XXXX Minute.

[0106] in, Figure 5 The 40%, 30%, and XX% in the figure represent the memory contribution rates of different learning events for the target user, such as browsing, doing objective questions, and doing subjective questions. For example, browsing a certain knowledge point only helps the target user deepen their impression and cannot help them remember 100% of the knowledge point, so it is 30%; while the target user looking at the textbook explanation has a deeper impact on remembering the knowledge point, so it is 40%.

[0107] Then, the graphical target user interface can provide a confirmation button to "display the forgetting curve of knowledge points." Clicking this confirmation button will display another corresponding functional interface, which can show the target user's memory change curve during the target learning period, such as... Figure 2 As shown.

[0108] In some embodiments, the method further includes: Send review reminders to target users at the target learning time to instruct them to process the reminders at that time.

[0109] In this embodiment, since the terminal needs to display a learning recommendation icon at the target learning time point on the memory prediction curve to make learning recommendations, in order to further remind the target user, the terminal can directly send review reminder information for the target learning time point to the target user, so as to further remind the target user to process the review reminder information at the target learning time point.

[0110] For example, if the review prompts only include behavioral instructions that direct the target user to review the target learning time, the target user can then search for content related to the target knowledge point at the target learning time for review.

[0111] For example, if the review prompt information includes the target user's specific review behavior (e.g., a set of test questions), the target user can review the specific review behavior at the target learning time point.

[0112] Furthermore, the specific review actions in the review prompts can correspond to jump links. Users can trigger these links to jump to the learning page corresponding to the specific review action, so that the target user can review the specific review action on that learning page.

[0113] In some embodiments, when sending review prompts to target users at target learning time points, the review prompts include target learning content for target knowledge points, which is the learning content recommended for the target user to learn at the target learning time points.

[0114] The process of determining the target learning content may include: determining the memory contribution information (such as memory contribution rate) of all learning content of the target knowledge point to the memory of the target knowledge point, and selecting the learning content with the greatest memory impact as the target learning content based on the degree of impact of all learning content of the target knowledge point on the memory of the target knowledge point.

[0115] For example, the target learning content could be video explanations, examples, test questions, etc.

[0116] Specifically, the degree to which all learning content for the target knowledge point affects the memorization of the target knowledge point can be set manually or obtained by a preset algorithm.

[0117] For example, if the learning content for a target knowledge point includes questions A and B, and question A contributes 10% to the memorization of the target knowledge point while question B contributes 15%, then question B can be selected as the target learning content to recommend to the target users.

[0118] In some embodiments, when sending review reminders to target users at target learning time points, the review method of the review reminders can be determined based on the target user's behavioral preference characteristics for target knowledge points, thereby providing learning and review reminders in a way that suits the target user and improving the target user's long-term learning benefits.

[0119] The review methods include, but are not limited to, learning methods (such as redoing wrong questions, quick quizzes, concept review, etc.) and review time (such as immediate review, evening reminders, specific review times such as the next morning).

[0120] In some embodiments, the behavioral preference characteristics of target users for target knowledge points can be determined based on behavioral data from the target user's historical learning events.

[0121] Behavioral data includes, but is not limited to, multi-dimensional data such as answer accuracy, study time, study frequency, and review interval.

[0122] In this embodiment, behavioral data from the target user's historical learning events can be integrated to provide prompts, thereby improving the intelligence of the review recommendations for the target user. Furthermore, through multi-dimensional behavioral data, the review methods become more comprehensive and intelligently scheduled, improving the accuracy and personalization of the review process.

[0123] Specifically, data clustering algorithms can be used to cluster behavioral data of different historical learning events of target users to clarify the behavioral preference characteristics of target users for target knowledge points. For example, based on students' answer accuracy, students' learning abilities can be clustered, such as students with high knowledge mastery and students with low knowledge mastery. Another example is to locate students' preferred learning time periods and the length of their learning time based on students' learning time. Yet another example is to determine students' learning habits and preferences through classification.

[0124] In some embodiments, the review method can also be retrieved from a rule-based strategy library. That is, by constructing a set of review strategy templates based on student type, selecting the appropriate template according to the current student's type, and determining the corresponding review method based on the selected template.

[0125] In some embodiments, the review method can also be learned based on a reinforcement learning strategy, that is, the strategy network is trained with the "long-term knowledge retention rate" as the optimization objective, and the learning method and review time in the review method are adaptively generated.

[0126] In some embodiments, the review method can also be obtained based on the case retrieval method, that is, using the Case-Based Reasoning (CBR) method to retrieve the historical review method that is most similar to the memory change curve of the target user for the target knowledge point in the platform's historical period, and directly apply the historical review method as the review method for the target user for the target knowledge point.

[0127] In some embodiments, the method further includes: determining at least one teaching task associated with the target user and the task time corresponding to each teaching task; selecting a target teaching task whose task time is closest to the target learning time from the teaching tasks; generating task content corresponding to at least one target knowledge point in the task content of the target teaching task; and sending the task content of the target teaching task to the target user to instruct the target user to complete the task content of the target teaching task, thereby realizing the review of the target knowledge point.

[0128] For example, the target teaching task could be an exam, so that while the target users are taking the exam, they can also review the target knowledge points, which greatly improves the learning efficiency of the target users.

[0129] Understandably, multiple thresholds corresponding to different prompt levels can be set according to needs (such as the preset thresholds and specific thresholds mentioned above), for example, the thresholds corresponding to the warning level and the forgetting level. Different prompt levels will trigger different prompt methods. Review prompts will be sent to the target user based on the prompt method corresponding to the triggered prompt level, so that the target user can intuitively understand the urgency of the current prompt through the corresponding prompt method, thereby improving the user experience.

[0130] It is understood that the above thresholds (such as the above preset thresholds or the above specific thresholds) can be determined based on the difficulty level of the target knowledge point and / or the learning level of the target user.

[0131] In some embodiments, when a target user is currently learning at least two target knowledge points, the at least two target knowledge points can be sorted based on a preset forgetting sorting strategy, such as sorting the target knowledge points from low to high memory strength, i.e., placing the target knowledge points with higher forgetting risk at the top of the list, and sending the sorted target knowledge points to the target user so that the target user can clearly understand their forgetting status of each target knowledge point.

[0132] In some embodiments, a review cycle optimization model can be constructed to evaluate the score function corresponding to the review time interval of the target knowledge point, thereby finding the review time point of the target knowledge point through the review cycle optimization model.

[0133] Please refer to Figure 6 Taking a terminal as an example, the process of determining the target learning time point in Embodiment 1 is explained. This embodiment provides a learning time point recommendation method. The specific process of this learning time point recommendation method can be described in the following steps 601-602, wherein: Step 601: Determine the memory prediction curve for the target knowledge point obtained after the target user performs at least one learning behavior at any time point on the memory change curve. The memory prediction curve is the change in the memory intensity of the target user for the target knowledge point within the estimated time period, generated based on the learning behavior.

[0134] In this embodiment, since the target user’s learning behavior related to the target knowledge point at different time points will lead to changes in the target user’s memory strength of the target knowledge point, the terminal predicts the change in the target user’s memory strength based on the user’s learning behavior related to the target knowledge point at different time points on the memory change curve, so as to generate memory prediction curves corresponding to different time points.

[0135] The memory prediction curve can be a curve within an estimated time period starting from the corresponding time node. This memory prediction curve can indicate the change in the target user's memory strength of the target knowledge point within the estimated time period after the time node.

[0136] Step 602: Select the time points where the change in memory intensity meets the preset conditions as the target learning time points, and make learning recommendations to the target users based on the target learning time points.

[0137] The preset conditions can be conditions corresponding to the user's memory strength reaching or about to reach the level of mastery of the target knowledge point, or conditions corresponding to the user's memory strength change being in line with the user's high memory benefit, etc. The preset conditions can be the same as or different from the preset strength achievement conditions. The specific settings can be set according to the needs and are not limited here.

[0138] In this embodiment, the terminal can select time points whose memory intensity changes meet preset conditions based on the memory intensity changes at each time point, and use the selected time points as target learning time points to recommend to the target user, so as to improve the user's learning efficiency of target knowledge points.

[0139] Understandably, by introducing a self-excitation mechanism, the learning behavior at each time point is simulated to enhance memory strength, so as to predict the degree of mastery of a certain knowledge point by the target user over time. Based on the enhancement effect of memory strength or the trend of memory strength change, the review interval of the target user for the target knowledge point is adjusted to achieve an efficient review strategy.

[0140] As can be seen above, a memory prediction curve for a target knowledge point can be obtained by determining that the target user performs at least one learning behavior at any time point on the memory change curve. The memory prediction curve represents the change in the target user's memory intensity for the target knowledge point within an estimated time period, generated based on the learning behavior. The time point where the change in memory intensity meets the preset conditions is taken as the target learning time point, and learning recommendations are made to the target user at the target learning time point. By predicting the memory prediction curve generated by the target user's learning at the time point, and evaluating whether the time point can be used as the target learning time point for recommending learning to the user, the user can be recommended the target knowledge point at the appropriate time, thereby improving the user's learning efficiency for a certain knowledge point.

[0141] In some embodiments, point process models (such as the Hawkes point process model) can be used to dynamically model the change in the memory strength of target users for target knowledge points over time, in order to dynamically depict the forgetting trajectory of students for each knowledge point, that is, the change in memory strength. The point process model regards each interaction event between the target user and the knowledge point as a "trigger event" and dynamically updates the memory strength according to the event time interval.

[0142] Optionally, a time-sensitive memory modeling mechanism can be built in the point process model. That is, by modeling the time point of each user's learning behavior (such as the behavior of reviewing a knowledge point), the change in the target user's memory strength of the knowledge point can be dynamically estimated, providing a precise time basis for review triggering.

[0143] Optionally, multiple review effect modeling can be introduced into the point process model. That is, by identifying and accumulating the memory impact of each review behavior (the learning behavior after the first learning of the target knowledge point), each learning behavior can produce a "memory enhancement effect" in the model, thereby more realistically simulating the memory patterns of users when learning.

[0144] In some embodiments, in addition to using Hawkes point process modeling to model the memory strength of the target user, nonparametric point process models, time-sensitive neural networks (such as Time-LSTM), Transformer models, or reinforcement learning-enhanced memory models can be used for memory modeling.

[0145] Among them, nonparametric point process models, such as the Gaussian process-modulated Poisson process, can automatically learn complex time-dependent structures and are suitable for more flexible memory strength modeling.

[0146] Among them, time-sensitive neural network models, such as Time-LSTM or T-LSTM, can introduce time interval variables when updating memory states, effectively modeling the dynamic process of forgetting behavior. Since both Time-LSTM and T-LSTM belong to Long Short-Term Memory (LSTM) networks, it is evident that by introducing the recurrent neural network structure of LSTM, it is possible to capture long-term dependencies in learning behavior sequences, dynamically model the complex temporal relationships and nonlinear effects between different learning events, and thus more accurately estimate the target user's knowledge mastery and forgetting trajectory.

[0147] Among them, the reinforcement learning reinforcement memory model can model the learning strategy as a Markov decision process (MDP), indirectly model memory changes using policy networks, and reflect memory strength through state transitions.

[0148] Among them, the Transformer model can leverage the self-attention mechanism to enable parallel modeling of the entire learning event sequence, fully exploring the global dependencies between different learning events. It is particularly suitable for learning sequence scenarios where multiple knowledge points and multiple behavior types coexist. Thus, it does not rely on a fixed decay kernel function and can automatically capture the interaction effects and time sensitivity between different types of learning events through learning methods.

[0149] In some embodiments, in order to indicate the model's ability to respond to individual differences among different target users, the model can be deeply integrated with the learning behavior of the target users when performing memory modeling, so as to dynamically model the model by combining the behavioral characteristics of the target users' learning behavior such as learning frequency and accuracy.

[0150] In this embodiment, in order to model the model, it is necessary to comprehensively record the target user's learning process on the platform. That is, detailed event information is recorded for every learning event of the target user on the platform. This detailed event information includes, but is not limited to, learning timestamps (such as the time of each learning or review); learning content (such as the knowledge point identifier corresponding to a certain knowledge point); behavior type (such as first learning, review, quiz, etc.); and learning results (such as correct / incorrect answers, completion time, resource usage, etc.).

[0151] Then, by structuring and processing the detailed event information, a unified data standard is constructed to ensure the temporal integrity and behavioral diversity of the model input data, thereby enhancing the system's ability to model the forgetting patterns of target users on target knowledge points, and thus providing a data foundation for the calculation of memory strength and / or the determination of target learning time points.

[0152] In some embodiments, the time point where the change in memory intensity meets the preset conditions is taken as the target learning time point, including: based on the memory prediction curve corresponding to each time point, the node with memory intensity at a preset threshold is taken as the target learning time point, thereby introducing a preset threshold to compare the memory intensity in the memory prediction curve with the preset threshold to automatically determine the best review time point.

[0153] In some embodiments, the time point at which the change in memory intensity meets preset conditions is used as the target learning time point, including: By comparing the changes in memory intensity on the memory change curve with the changes in memory intensity on the memory prediction curve at each time point, the magnitude of the change in memory intensity at each time point can be determined. Based on the magnitude of changes in memory intensity at each time point, time points that meet preset conditions are selected as target learning time points.

[0154] Among them, the magnitude of memory intensity change is used to indicate the relative situation between the change in memory intensity after learning at a time node and the change in memory intensity before learning at a time node (memory change curve). Since memory intensity generally tends to increase after learning at a time node, the magnitude of memory intensity change can also be used as the increase in memory intensity when learning at a time node.

[0155] In this embodiment, the memory prediction curve based on each time point is compared with the memory change curve to clarify the magnitude of memory intensity change at each time point. Based on the magnitude of memory intensity change at each time point, time points that meet preset conditions are selected to automatically determine the best review time.

[0156] Specifically, based on the magnitude of change in memory intensity at each time point, time points that meet preset conditions are selected as target learning time points, including: selecting the time point with the largest magnitude of change in memory intensity as the target learning time point.

[0157] Specifically, based on the magnitude of change in memory intensity at each time point, time points that meet preset conditions are selected as target learning time points, including: selecting time points where the magnitude of change in memory intensity is greater than a preset magnitude threshold as target learning time points.

[0158] Specifically, the changes in memory intensity on the memory change curve are compared with the changes in memory intensity on the memory prediction curve at each time point to determine the magnitude of memory intensity change at each time point, including: Calculate the difference between the memory intensity of the memory change curve and the memory intensity of the memory prediction curve at a specified time point within the estimated time period; The difference at a specified moment corresponding to each time node is taken as the magnitude of change in memory strength; or Calculate the difference between the memory intensity of the memory change curve and the memory intensity of the memory prediction curve at each time point within the estimated time period. The differences between the time windows corresponding to each time node are averaged to use the result as the magnitude of change in memory intensity.

[0159] The time window can be a window that contains at least one time node, and different time windows contain the same number of time nodes. For example, all time windows contain two time nodes.

[0160] Specifically, calculating the difference between the memory intensity of the memory change curve and the memory intensity of the memory prediction curve corresponding to each time node within each time window of the estimated time period can include: determining the first difference between the memory intensity of any two time nodes within the corresponding time window on the memory change curve, determining the second difference between the memory intensity of the same two time nodes within the corresponding time window on the memory prediction curve, and then using the difference between the first difference and the second difference as the difference between the memory intensity of the memory change curve within the corresponding time window and the memory intensity of the memory prediction curve corresponding to each time node.

[0161] Specifically, calculating the difference between the memory intensity of the memory change curve and the memory intensity of the memory prediction curve corresponding to each time node within each time window of the estimated time period can include: determining the third difference between any time node in the corresponding time window on the memory change curve and the memory intensity of the same time node in the corresponding time window on the memory prediction curve; and then using the third difference as the difference between the memory intensity of the memory change curve in the corresponding time window and the memory intensity of the memory prediction curve corresponding to each time node.

[0162] For example, the process of comparing the changes in memory intensity on the memory change curve with the changes in memory intensity on the memory prediction curve at each time point is shown in the following formula:

[0163] Where t+T is the time range for comparison, such as t being a time node and T being a time interval (such as a week). This shows the changes in memory intensity at corresponding time points on the memory prediction curve. This represents the change in memory intensity at corresponding time points on the memory change curve.

[0164] Understandably, by comparing the memory change curve and the memory prediction curve, we can obtain the memory improvement benefits of learning about target knowledge points at different time points. Based on these benefits, we can select appropriate time points as target learning time points to recommend learning about target knowledge points to target users.

[0165] Then, the argmax function can be used to filter the benefits at different time points to determine the optimal review time, i.e., the target learning time. The method of using a benefit function (such as the argmax function) to filter the benefits at different time points is as follows:

[0166] Among them, t The time point with the greatest benefit, such as the time point with the greatest change in memory intensity.

[0167] Optionally, in addition to the revenue function, the learning cost of the target knowledge point can be calculated to obtain a comprehensive score for the target knowledge point. A review plan can be arranged based on the comprehensive score. The learning cost includes, but is not limited to, the average learning time of all target users on a knowledge point, the first-time learning pass rate of the knowledge point, the complexity of the knowledge point, and the average difficulty of related test questions under the knowledge point.

[0168] In some embodiments, determining the memory prediction curve for a target knowledge point after the target user performs at least one learning action at any time point on the memory change curve includes: Based on learning behavior and time points, determine the memory contribution information of learning behavior at each estimated time point within the estimated time period; Based on memory contribution information, the memory strength of the target user for the target knowledge point is determined at each estimated time point, and a memory prediction curve for the target knowledge point is obtained based on the memory strength at each estimated time point.

[0169] The estimated time point is the moment when the target user's mastery of the target knowledge point needs to be predicted, which can be denoted as t. Correspondingly, the memory strength at the estimated time point can be denoted as λ(t).

[0170] In this embodiment, different learning behaviors contribute differently to the memory of the target knowledge points for the target user. Based on the learning behavior at a time node, the memory contribution information of the learning behavior at each time node within the estimated time period is evaluated. Then, based on the obtained memory contribution information, the memory intensity change within the estimated time period is determined, that is, the memory prediction curve corresponding to the estimated time period is generated.

[0171] Specifically, based on learning behavior and time points, the memory contribution information of learning behavior at each estimated time point within the estimated time period is determined, including: Obtain memory-influencing parameters corresponding to learning behaviors; Based on the memory impact parameters and time points of learning behavior, as well as the estimated time points, we determine the contribution information of learning behavior to the memory of target knowledge points at the estimated time points.

[0172] Among them, the memory influence parameter is a parameter used to indicate the influence of learning behavior on the memory of the target user when learning the target knowledge point. The memory influence parameter includes, but is not limited to, the target decay intensity and / or the target excitation intensity.

[0173] In this embodiment, since the target user has a corresponding implementation time when performing the learning behavior, i.e. time node, the learning behavior itself and the implementation time of the learning behavior will affect the degree of memory at different estimated time nodes within the estimated time period. That is, different learning behaviors contribute different amounts of memory information to different third time points.

[0174] It should be noted that the greater the time difference between the actual time point and the estimated time point, the lower the impact of the learning behavior that occurs at that time point on the target user's memory of the target knowledge points at the estimated time point. For example, the memory impact of an estimated time point closer to the actual time of the learning behavior is greater than that of an estimated time point farther away from the actual time of the learning behavior.

[0175] Optionally, since different learning behaviors have different impacts on the target user's memory of the target knowledge points, for example, the impact of the target user browsing the definition of the target knowledge point may be 30%, while the impact of the target user answering questions related to the target knowledge point may be greater, such as 40%. Therefore, the impact of learning behaviors on the target user's memory of the target knowledge points at the estimated time point can be evaluated by the type of learning behavior, that is, the memory impact parameter corresponding to the learning behavior can be determined by the type of learning behavior.

[0176] Specifically, memory impact parameters include target decay intensity and target excitation intensity, memory impact parameters based on learning behavior and time points, and estimated time points. Information on the contribution of learning behavior to the memory of target knowledge points at the estimated time points is determined, including: Determine the time difference between the predicted and actual time points; Based on the time difference and the target decay intensity, determine the decay indication parameters corresponding to the learning behavior; The product of the target stimuli intensity and the decay indicator parameter of the learning behavior is used as the information on the contribution of the learning behavior to the memory of the target knowledge point at the predicted time point.

[0177] It is understandable that the greater the time difference between the time node and the estimated time node, the lower the impact of the learning behavior at the time node on the target user's memory of the target knowledge points at the estimated time node. Therefore, the decay of the learning behavior at the time node on the memory strength at the estimated time node can be calculated by the time difference between the time node and the estimated time node and the target decay intensity corresponding to the learning behavior. This decay is used to represent the decay indicator parameter.

[0178] It is understandable that, since the learning behavior itself will have different degrees of stimulation on the target user's memory, after obtaining the decay of the memory intensity of the learning behavior at the estimated time point, it is also necessary to calculate the decay indicator parameter indicating the decay intensity and the target stimulation intensity indicating the stimulation of the learning behavior to obtain the degree of contribution of the learning behavior to the memory of the target knowledge point at the estimated time point, that is, the memory contribution information.

[0179] Therefore, by introducing target decay intensity and target activation intensity, we can accurately reflect the process of user memory decay over time and the activation of user memory by each learning behavior. This achieves the characterization of the time dependence of memory decay and behavioral activation, and solves the problems of staticity and inaccuracy of traditional forgetting assessment methods.

[0180] For example, the time node is set as ts, the estimated time node is t, the target decay intensity of the learning behavior is β, and the target excitation intensity of the learning behavior is α.

[0181] Therefore, the time difference between the current time point and the estimated time point is: t - ts; The decay indicator parameter corresponding to the learning behavior: e -β（t-ts） ; Information on the contribution of learning behavior to the memorization of target knowledge points: α e -β（t-ts） .

[0182] In some embodiments, since there are historical learning events of the target user on the target knowledge point in the historical period prior to the time node, and these historical learning events also have an impact on the target user's memory of the target knowledge point at the estimated time node, determining the target user's memory strength for the target knowledge point at each estimated time node based on memory contribution information includes: Obtain information on the historical memory contribution of at least one historical learning behavior of the target user at the estimated time point; By accumulating historical memory contribution information and memory contribution information corresponding to the estimated time points, the memory strength of the target user for the target knowledge point at the estimated time point is obtained.

[0183] Among them, historical learning behavior refers to learning behavior in academic learning events.

[0184] For example, if we assume that t1, t2, and t3 each involve different learning behaviors, then the memory contribution information of these three learning behaviors are respectively: α e -β（t-t1） α e -β（t-t2） α e -β（t-t3） Then, the memory contribution information of the three learning behaviors is accumulated to obtain the memory intensity of the target user for the target knowledge point at the estimated time point.

[0185] In some embodiments, in order to more accurately assess the memory strength at the estimated time point, a base memory strength can be introduced, which can be set according to needs and is not limited here.

[0186] For example, if we set the estimated time point as t, the target decay intensity of the learning behavior as β, and the target activation intensity of the learning behavior as α, then we can use the point process model to obtain the formula for determining the memory intensity at the estimated time point: λ(t) = μ + ∑α e -β（t-ti） .

[0187] Where λ(t) is the estimated memory strength at the predicted time point, μ is the basic memory strength, ti is less than t, and ti includes the time point ts and the behavioral moment corresponding to the historical learning behavior.

[0188] The formula for determining the memory strength at the predicted time point demonstrates that it portrays the impact of behavioral events (such as testing, browsing, and reading) occurring before time t on memory strength. By accumulating the contributions of different learning behaviors through an exponential kernel function, it characterizes the activation and decay process of learning behaviors at multiple different time points on the memory strength at the predicted time point, exhibiting good interpretability and temporal modeling capabilities.

[0189] In some embodiments, in order to enable the target user's learning behavior to more accurately express the target user's memory changes, the target user's different learning behaviors can be identified in this embodiment.

[0190] Specifically, the recommended learning time points also include: Obtain the target user's historical learning trajectory, which includes multiple learning behavior samples and memory labels for each learning behavior sample; Based on the target user's historical learning trajectory, each learning behavior sample and its memory label are fitted to determine the activation intensity and decay intensity corresponding to the learning behavior of each learning behavior sample of the target user.

[0191] In this embodiment, in order to support personalized memory modeling for target users, model parameters can be adjusted in a personalized manner according to the learning behavior of target users to achieve modeling of individual differences such as forgetting speed and memory activation intensity, thereby generating customized review suggestions and improving adaptability.

[0192] Specifically, personalized parameter learning methods can be introduced, that is, designing a model parameter fitting mechanism that adapts to the individual differences of the target user, so as to realize personalized modeling of memory processes such as forgetting speed and stimulating effect.

[0193] To adapt to the memory characteristics of different target users, a personalized learning and dynamic update mechanism for model parameters can be introduced. This means that the parameters in the point process model (such as decay rate and excitation intensity) can be learned in a personalized manner and dynamically updated with new data, thereby enhancing the robustness and accuracy of the model.

[0194] In some embodiments, maximum likelihood estimation can be used to fit the historical learning trajectory. That is, the sequences corresponding to the learning behavior samples of different learning events in the past period of the target user are superimposed to form the learning trajectory. Each learning event in the sequence is labeled, i.e., a memory label, which indicates the change in the memory intensity of the corresponding learning event. Then, maximum likelihood estimation is used to fit these data with memory labels. Then, the fitted trajectory is used as input, and the fitted trajectory is processed by the parameter adjustment model. In this way, the excitation intensity and decay intensity corresponding to different learning behaviors can be output, so as to realize the personalized parameter setting for the target user to adapt to the user's personalized needs.

[0195] In some embodiments, incremental parameter updates can also be performed to ensure the real-time performance and stability of the online learning scenario. That is, as time goes by, the learning events of the target user will continue to increase, and the excitation intensity and decay intensity will be continuously updated to make the parameter values ​​more stable.

[0196] In addition, regularization terms can be designed to prevent overfitting and improve the generalization ability of parameters to cope with the low accuracy when there is little data.

[0197] In some embodiments, the functional modules that implement different functions in the learning time point recommendation method can be modularly deployed through a microservice architecture. For example, it can support the embedding of multiple educational platforms (such as learning apps, question bank systems, online classrooms, etc.) and can be connected to third-party teaching resource systems through API interfaces.

[0198] It is understandable that modular deployment can achieve scalability and versatility. For example, the point-process model has a simple structure and is easy to integrate into multiple educational platforms. Furthermore, the modular design of the system facilitates replacement and iteration, and can be expanded to be applicable to learning systems of different grade levels and subjects.

[0199] Understandably, the modular design of the system structure and platform adaptability allow for the embedding of various learning systems, adapting to different educational scenarios and achieving a highly available and scalable review support system architecture.

[0200] Please refer to Figure 7 Taking a terminal as an example, this embodiment also provides a knowledge point interaction method. The specific process of this knowledge point interaction method can be described in the following steps 701, wherein: Step 701: In response to the knowledge point viewing operation triggered by the target user, display at least one target knowledge point, the target user's historical learning events for the target knowledge point at at least one historical moment, and the contribution information of each historical learning event to the target user's memory of the target knowledge point at the current moment.

[0201] The target knowledge point can be the knowledge point that the target user has already learned.

[0202] The knowledge point viewing operation is used to trigger the viewing of at least one target knowledge point and other information related to the target knowledge point. This knowledge point viewing operation can be triggered by the user's behavior or by a pre-set triggering mechanism on the terminal. The specific settings can be configured according to the needs and are not limited here.

[0203] Among them, historical learning events can be events generated by a user's learning behavior on a certain learning content in a historical period. Such historical learning behavior includes, but is not limited to, the user's behavior of browsing, learning, or reviewing the learning content of the target knowledge point. For example, the target user's behavior of viewing and / or answering questions related to the target knowledge point.

[0204] Among them, the memory contribution information is used to indicate the contribution of a user's learning behavior to the memory of the target knowledge point. This memory contribution information can be represented by numerical values ​​such as percentages, fractions, and decimals, or by text, characters, etc. The specific settings can be set according to the needs and are not limited here.

[0205] In this embodiment, by displaying the target user's historical learning events for the target knowledge point, and the corresponding memory contribution information of the target user under the historical learning events, the user can intuitively understand the overall learning process of the target user for the target knowledge point at the current moment, and the impact of each learning event on the user's memory and mastery of the target knowledge point during the overall learning process. Thus, the target user can learn specific learning events based on the memory contribution information of each historical learning event, thereby improving the user's learning efficiency for a certain knowledge point.

[0206] Specifically, the memory contribution information can be expressed as a percentage, which indicates that the target user has a deeper memory of the target knowledge point after the corresponding learning behavior. For example, the memory contribution rate corresponding to the target user browsing the definition of the target knowledge point is 30%, while the memory contribution rate corresponding to the target user answering questions related to the target knowledge point is 40%.

[0207] It is understandable that the memory contribution information of each historical learning event can affect the user's memory strength of the target knowledge point over time and with the accumulation of historical learning events.

[0208] In some embodiments, the method further includes: In response to the target user's memory trigger operation on the target knowledge point, the system displays the target user's memory change information within the target learning time period. The memory change information is generated based on the memory contribution information of each user, showing the change in the target user's memory intensity of the target knowledge point. The target learning time period includes historical time periods and future time periods.

[0209] Among them, the memory change information shows how the target user's memory strength of the target knowledge point changes during the target learning period. This memory strength change can be composed of memory strength data from multiple different time points. This memory strength indicates the target user's mastery of the target knowledge point; the greater the memory strength, the deeper the target user's memory of the target knowledge point.

[0210] It should be noted that since memory strength can indicate the target user's mastery of the target knowledge point, it can be indicated by percentages, fractions, decimals, etc. For example, when the target user's memory strength for the target knowledge point is 100%, it means that the target user has fully mastered the target knowledge point.

[0211] The target learning time period is the time period in the memory change curve that shows the change in the user's memory intensity of the target knowledge point. In this target learning time period: the historical time period is the historical period before the current moment, and the future time period is the future period after the current moment.

[0212] The historical time period can include at least a portion of the time period between the moment when the target user first learns the target knowledge point and the current moment. The future time period can include a time period consisting of a certain duration after the current moment. The division of the historical time period and the future time period can be set according to the needs and is not limited here.

[0213] Specifically, the terminal can use the memory contribution information corresponding to different historical learning events of the target user to assess the changes in the target user's memory strength for the target knowledge point during the target learning period, thereby improving the accuracy of the assessment results when evaluating the target user's mastery of a certain knowledge point.

[0214] Among these, information on changes in memory includes, but is not limited to, presentation through curves, bar charts, and other methods.

[0215] Specifically, when memory change information can be presented through a memory change curve, it enables the target user to intuitively and accurately understand the target user's memory mastery of the target knowledge points as indicated by the memory change curve displayed in the graphical target user interface.

[0216] It should be noted that when displaying the memory change curve of the target user for the target knowledge point, the memory intensity on the memory change curve will change over time. If the target user reviews frequently, the memory intensity of the target knowledge point will be improved. However, if the target user does not review for a long time, the memory intensity of the target knowledge point may decrease before the target user fully masters the target knowledge point.

[0217] Specifically, memory change curves can be obtained through interpretable models, such as linear point process models and logistic regression models. Interpretable models are used to generate and provide clear causal explanation structures.

[0218] For example, such as Figure 2 As shown, in Figure 2 The graph shows the changes in memory, with the horizontal axis representing time and the vertical axis representing memory intensity. Figure 2 The curve between point A and point B, from Figure 2 As can be seen, Figure 2 Point C in the model represents the time node of the last specific learning event of the target user on this knowledge point. Each point before (including) point C represents the date on which the target user actually engaged in learning behavior and the strength of their memory of the knowledge point after learning. Points after point C represent the changes in the strength of the target user's memory of the knowledge point if they do not continue learning after point C, as predicted by the model.

[0219] Among them, from Figure 2 It is evident that during the time period following point C, which includes future time periods, the target user's memory of the relevant knowledge points will decrease over time because the target user will not engage in any further learning activities related to the knowledge points. In other words, the target user's memory of the relevant knowledge points will become increasingly weaker.

[0220] In some embodiments, the end time of the target learning time period, that is, the end time of the future time period, can be determined by the time when the interval between the initial time and the initial time meets the preset time threshold, or by the time when the corresponding memory strength meets the preset threshold condition.

[0221] Specifically, the end point can be the time when the memory strength is less than the preset forgetting threshold. For example, if the target user has forgotten 95% of the target knowledge point, that is, when the target user's memory strength is 5%, it can be considered that the target user only remembers the key words in the target knowledge point and is about to forget the target knowledge point. This time point can be determined as the end point.

[0222] Specifically, the end point can be the time when the memory intensity is greater than the preset memory threshold. For example, the time when a target user has completely mastered a knowledge point due to a certain learning behavior. For instance, after watching a course, based on historical learning events, it can be inferred that the target user has completely mastered or is about to master a knowledge point. If the memory intensity of the knowledge point reaches 92%, and this 90% is greater than the preset memory threshold, then this time point can be determined as the end point.

[0223] The preset memory threshold is greater than the preset forgetting threshold. The preset forgetting threshold and the preset memory threshold can be set according to needs, and are not limited here.

[0224] In some embodiments, since a target user may have multiple historical learning events for a target knowledge point, these historical learning events can be arranged according to preset behavior arrangement rules. For example, based on the learning time of different historical learning events, multiple historical learning events can be arranged in chronological order, i.e., the later the learning time, the earlier the historical learning event is arranged. Or, based on the memory contribution information of different historical learning events, multiple historical learning events can be arranged from largest to smallest, i.e., the larger the memory contribution information, the earlier the historical learning event is arranged. The specific arrangement can be set according to the needs, and there are no further limitations.

[0225] In some embodiments, target knowledge points and their historical learning events can be displayed together on a graphical target user interface, and / or, historical learning events and their corresponding memory contribution information can be displayed together on a graphical target user interface, wherein the display of both includes, but is not limited to, being displayed through lists, bar charts, text, etc.

[0226] In some embodiments, the terminal may be configured with a knowledge point forgetting query function. By triggering the knowledge point forgetting query function, a knowledge point viewing operation can be realized. When the knowledge point forgetting query function is triggered, the terminal may display a function interface. The function interface may include at least a first display area and a second display area. The first display area is used to display at least one target knowledge point, while the second display area may display at least one historical learning event related to the target knowledge point, as well as information on the contribution of the historical learning event to the target user's memory of the target knowledge point.

[0227] The display format of the target knowledge points includes, but is not limited to: sorting the target knowledge points based on the target user's memory strength of the target knowledge points at the current moment, and displaying the sorted target knowledge points; at each target knowledge point, the target user's memory strength of the target knowledge point at the current moment is displayed in association, and different memory strengths can be displayed in different display styles, such as numerical values ​​of different colors, bar charts, etc.

[0228] Furthermore, when displaying at least one historical learning event of the target knowledge point in the second display area, the terminal can receive a selection operation of the target knowledge point in the first display area from the target user, so as to display the target knowledge point indicated by the selection operation and at least one historical learning event of the target knowledge point indicated by the selection operation in the second display area.

[0229] For example, such as Figure 5 As shown, in Figure 5 In the middle, the target knowledge point indicated by the selection operation is set as "Chapter XX, Section XX: Electric Charge and its Conservation Law", and in Figure 5 The window shown displays the target user's historical learning events related to "Chapter XX, Section XX: Electric Charge and Its Conservation Law," for example, Figure 5 The historical learning events include: "XX City College Entrance Examination Prediction Paper Question XX: 40%" held at 14:33 on February 16, 2025; "XX Practice Questions and their Analysis Question XX: 30%" held at 9:45 on March 3, 2025; "XX City Textbook Page XX Question XX: 40%" held at 10:22 on June 26, 2025; and "XXXXXX: XX%" which occurred on XX Month XX Day, XXXX Year at XXXX Minute.

[0230] in, Figure 5 The 40%, 30%, and XX% in the figure represent the memory contribution rates of different learning events for the target user, such as browsing, doing objective questions, and doing subjective questions. For example, browsing a certain knowledge point only helps the target user deepen their impression and cannot help them remember 100% of the knowledge point, so it is 30%; while the target user looking at the textbook explanation has a deeper impact on remembering the knowledge point, so it is 40%.

[0231] Then, the graphical target user interface can provide a confirmation button to "display the forgetting curve of knowledge points." Clicking this confirmation button will display another corresponding functional interface, which can show the target user's memory change curve during the target learning period, such as... Figure 2 As shown.

[0232] In some embodiments, the memory change information is a memory change curve, and the method further includes: In response to a trigger operation on any time point on the memory change curve, a memory prediction curve is generated with the triggered time point as the starting point. The memory prediction curve predicts the change in the memory strength of the target knowledge point after the target user performs at least one learning behavior at the time point.

[0233] In this embodiment, since the target user's learning behavior related to the target knowledge point at different time points will lead to changes in the target user's memory strength of the target knowledge point, the terminal predicts the change in the user's memory strength if the user performs learning behavior related to the target knowledge point at that time point by responding to the user's trigger operation on a time point on the memory change curve. This generates a memory prediction curve starting from that time point, making the user understand what kind of change will occur in their memory strength of the target knowledge point if they perform learning behavior for the target knowledge point at the triggered time point.

[0234] The memory prediction curve covers a learning period that includes the triggered time point and at least one subsequent time point.

[0235] For example, based on Figure 2 As shown in the example, if a user wants to see how their memory strength changes after learning about a specific knowledge point at a future time, they can... Figure 2 The trigger operation is performed at any time point after point C on the memory change curve, such as... Figure 3 As shown, the user triggers an action at point D to simulate the target user performing at least one learning behavior at point D. The goal is to predict the target user's memory strength after point D, and then generate... Figure 3 The memory change curve between point D and point E is used as the memory prediction curve.

[0236] Among them, from Figure 3 It is evident that because the target user performs at least one learning action at point D, the target user's memory strength shows an upward trend. However, in the time period after point D, which includes the future time period, since the target user does not perform any more learning events related to the corresponding knowledge points, the memory strength after point D, which increases for a period of time, will decrease over time. In other words, the target user's memory strength of the corresponding knowledge points becomes lower and lower.

[0237] Optionally, after learning a certain knowledge point, the target user can use the system to calculate the memory change curve, and by triggering a certain time point on the memory change curve, calculate the change in memory intensity of the knowledge point when reviewing at a future time point. Based on the change in memory intensity at a specific time point in the memory change curve, the system can help understand the source of the review suggestions and encourage the target user to make corresponding learning plans.

[0238] In some embodiments, generating a memory prediction curve starting from the triggered time point includes: Display at least one candidate learning content at the triggered time point; In response to a selection instruction for at least one candidate learning content, a memory prediction curve generated based on the selected candidate learning content is displayed.

[0239] Among them, the candidate learning content is learning content related to the target knowledge points, such as different types of test questions, detailed explanations of knowledge points, etc.

[0240] In this embodiment, by displaying candidate learning content at the triggered time node, the user can select the content they want to learn at the triggered time node, thereby triggering the above selection instruction and prompting the terminal to clearly understand the learning content that the target user expects to learn at the triggered time node. The above learning behavior is the behavior of learning the learning content.

[0241] Then, since different learning content contributes differently to the memory of the target knowledge point by the target user, the change in the memory strength of the target knowledge point after learning different learning content will vary. Therefore, the candidate learning content selected by the user at the triggered time point can be used as the learning content that needs to be learned at the triggered time point, so as to predict the change in the memory strength of the target user for the target knowledge point after learning the selected candidate learning content at the triggered time point.

[0242] Specifically, the above selection instructions can be generated by triggering operations on the displayed candidate learning content. These triggering operations can be click operations, such as single-click or double-click operations, or press operations, such as press for a preset duration, or swipe operations, etc. The specific settings can be configured according to the requirements and are not limited here.

[0243] In some embodiments, a learning recommendation indicator is displayed at the target learning time point within a future time period on the memory prediction curve.

[0244] The target learning time point is the time point in the memory intensity change that meets the preset conditions.

[0245] The learning recommendation identifiers mentioned above can be presented in styles such as text, icons, and characters.

[0246] The preset conditions can be conditions corresponding to the user's memory strength reaching or about to reach the level of mastery of the target knowledge point, or conditions corresponding to the user's memory strength change being in line with the user's high memory benefit, etc. The specific settings can be set according to the needs and are not limited here.

[0247] In this embodiment, by displaying corresponding learning recommendation icons at the target learning time points, users can intuitively identify the appropriate time to learn the target knowledge points, and ensure that the memory strength after learning at the target learning time points indicated by the learning recommendation icons meets the preset conditions, thereby improving the user's learning efficiency of the target knowledge points.

[0248] Furthermore, by identifying target learning time points and recommending learning methods to users, this approach avoids the current learning lag caused by only reviewing knowledge points when users have not been exposed to them for a long time or when they answer questions related to a knowledge point incorrectly. This lack of adaptive prompts based on the user's forgetting process prevents intervention at the optimal time.

[0249] For example, based on Figure 2 The example shown illustrates that if there are time points in the future time period on the memory prediction curve where the memory strength meets preset conditions, such as... Figure 4 As shown, Figure 4 Point F in the equation represents the target learning time point. Therefore, the terminal can simulate the target user performing at least one learning action at point F, predict the memory strength of the target user after performing the learning action related to the target knowledge point at point F, and generate [the necessary data / information]. Figure 4 The memory change curve between point F and point G can be used as a learning recommendation indicator, or a preset icon can be displayed at point F as a learning recommendation indicator. Then, when the user triggers point F, the memory change curve between point F and point G can be displayed. The specific settings can be configured according to the needs, and there are no restrictions here.

[0250] Among them, from Figure 4 It is evident that after performing at least one learning action at point F, the target user can fully master the target knowledge point or master it to a certain extent, such as... Figure 4 The memory strength of the target users reaches 100%, at least in the target users Figure 4 The target knowledge point will not be forgotten during the time period corresponding to point F to point G.

[0251] In some embodiments, the target learning time point is the node where the average memory intensity of the target knowledge point within the estimated time period after the target user performs at least one learning behavior is greater than a preset threshold, or the target learning time point is the node where the increase in memory intensity of the target knowledge point after the target user performs at least one learning behavior at the target learning time point is higher than other time points.

[0252] It is understandable that when a target user learns the target knowledge point at different time points, the intensity of the target user's memory of the target knowledge point can be changed. By comparing the memory intensity with a preset threshold, or by comparing the increase in memory intensity with the increase in memory intensity at other time points, it can be determined whether there is a time point where the memory intensity meets the preset intensity standard, i.e., the aforementioned target learning time point, and thus automatically determine the best review time point.

[0253] Specifically, the increase in memory intensity can be obtained by calculating the difference between the memory intensity at any moment within the estimated time period and the memory intensity at the corresponding moment on the memory change curve. The specific value can be set according to the needs and is not limited here.

[0254] Specifically, the increase in memory intensity can be obtained by determining the increase in memory intensity between each moment (e.g., each day) in the estimated time period and the corresponding moment on the memory change curve, and then calculating the average of the increases in memory intensity across all moments.

[0255] In some embodiments, as the time interval between the target user's most recent contact with the target knowledge point increases, the target user's memory strength of the target knowledge point decreases. Therefore, the terminal can also introduce a specific threshold to dynamically monitor the target user's memory of the target knowledge point, thereby identifying the corresponding target learning time point for which learning recommendations need to be made to the target user.

[0256] Specifically, the target learning time point can also be the node where the target user's memory strength of the target knowledge point is less than a specific threshold. That is, if the target user's memory strength of a certain knowledge point drops below the specific threshold, the target learning time point can be used as the time node to trigger the review recommendation. This allows the review rhythm to be automatically scheduled according to the target user's forgetting trend of the target knowledge point, so as to obtain the best review time in advance before the target user completely forgets the target knowledge point, reduce unnecessary review time investment, improve the review hit rate, avoid ineffective or excessive review, and at the same time reduce the risk of the target user forgetting the target knowledge point.

[0257] In some embodiments, in order to make it easier for target users to intuitively understand the learning benefits they have gained at the target learning time point, learning benefit information can be displayed at the learning recommendation label, or the learning recommendation label may contain learning benefit information.

[0258] The learning benefit information can indicate the learning benefits of a target user at the target learning time point indicated by the learning recommendation icon. These learning benefits include, but are not limited to, changes in the target user's memory strength (such as an increase) within the estimated time period, and the duration during which the target user's memory strength is not lower than a specific threshold.

[0259] For example, if a target user learns at least one question containing that knowledge point three days later, and their memory strength in the next three months is not less than a certain threshold, then a time point three days apart from the current time can be used as the target learning time point, and the learning benefit information "The target user will not forget the knowledge point in the next three months" can be displayed for that target learning time point.

[0260] For example, if the target learning time point is obtained when the memory strength is less than a certain threshold, then the learning benefit information can also indicate the triggering reason for the target learning time point, such as the last learning time being X days ago and the expected forgetting risk being Y%. Natural language explanations can be output, such as "You last learned this knowledge point 5 days ago. The current memory strength has dropped below the set threshold. The system suggests reviewing it."

[0261] In this embodiment, by further displaying learning benefit information, the interpretability of the recommendation results is improved. Through visual learning recommendation indicators (such as displaying memory change curves) and natural language trigger descriptions, it is possible to clearly explain why a certain knowledge point is recommended for review at a certain time point, thereby enhancing system transparency, improving the understanding, trust, or acceptance of target users (such as teachers and students), and enhancing the learning initiative of target users.

[0262] Specifically, learning benefit information can be generated through prompt-based natural language generation, that is, based on rule-based or fine-tuned language models, chat-like explanatory text can be generated to improve the human-computer interaction experience of the system.

[0263] In some embodiments, the method further includes: Send review reminders to target users at the target learning time to instruct them to process the reminders at that time.

[0264] In this embodiment, since the terminal needs to display a learning recommendation icon at the target learning time point on the memory prediction curve to make learning recommendations, in order to further remind the target user, the terminal can directly send review reminder information for the target learning time point to the target user, so as to further remind the target user to process the review reminder information at the target learning time point.

[0265] For example, if the review prompts only include behavioral instructions that direct the target user to review the target learning time, the target user can then search for content related to the target knowledge point at the target learning time for review.

[0266] For example, if the review prompt information includes the target user's specific review behavior (e.g., a set of test questions), the target user can review the specific review behavior at the target learning time point.

[0267] Furthermore, the specific review actions in the review prompts can correspond to jump links. Users can trigger these links to jump to the learning page corresponding to the specific review action, so that the target user can review the specific review action on that learning page.

[0268] In some embodiments, when sending review prompts to target users at target learning time points, the review prompts include target learning content for target knowledge points, which is the learning content recommended for the target user to learn at the target learning time points.

[0269] The process of determining the target learning content may include: determining the memory contribution information (such as memory contribution rate) of all learning content of the target knowledge point to the memory of the target knowledge point, and selecting the learning content with the greatest memory impact as the target learning content based on the degree of impact of all learning content of the target knowledge point on the memory of the target knowledge point.

[0270] For example, the target learning content could be video explanations, examples, test questions, etc.

[0271] Specifically, the degree to which all learning content for the target knowledge point affects the memorization of the target knowledge point can be set manually or obtained by a preset algorithm.

[0272] For example, if the learning content for a target knowledge point includes questions A and B, and question A contributes 10% to the memorization of the target knowledge point while question B contributes 15%, then question B can be selected as the target learning content to recommend to the target users.

[0273] In some embodiments, when sending review reminders to target users at target learning time points, the review method of the review reminders can be determined based on the target user's behavioral preference characteristics for target knowledge points, thereby providing learning and review reminders in a way that suits the target user and improving the target user's long-term learning benefits.

[0274] The review methods include, but are not limited to, learning methods (such as redoing wrong questions, quick quizzes, concept review, etc.) and review time (such as immediate review, evening reminders, specific review times such as the next morning).

[0275] In some embodiments, the behavioral preference characteristics of target users for target knowledge points can be determined based on behavioral data from the target user's historical learning events.

[0276] Behavioral data includes, but is not limited to, multi-dimensional data such as answer accuracy, study time, study frequency, and review interval.

[0277] In this embodiment, behavioral data from the target user's historical learning events can be integrated to provide prompts, thereby improving the intelligence of the review recommendations for the target user. Furthermore, through multi-dimensional behavioral data, the review methods become more comprehensive and intelligently scheduled, improving the accuracy and personalization of the review process.

[0278] Specifically, data clustering algorithms can be used to cluster behavioral data of different historical learning events of target users to clarify the behavioral preference characteristics of target users for target knowledge points. For example, based on students' answer accuracy, students' learning abilities can be clustered, such as students with high knowledge mastery and students with low knowledge mastery. Another example is to locate students' preferred learning time periods and the length of their learning time based on students' learning time. Yet another example is to determine students' learning habits and preferences through classification.

[0279] In some embodiments, the review method can also be retrieved from a rule-based strategy library. That is, by constructing a set of review strategy templates based on student type, selecting the appropriate template according to the current student's type, and determining the corresponding review method based on the selected template.

[0280] In some embodiments, the review method can also be learned based on a reinforcement learning strategy, that is, the strategy network is trained with the "long-term knowledge retention rate" as the optimization objective, and the learning method and review time in the review method are adaptively generated.

[0281] In some embodiments, the review method can also be obtained based on the case retrieval method, that is, using the Case-Based Reasoning (CBR) method to retrieve the historical review method that is most similar to the memory change curve of the target user for the target knowledge point in the platform's historical period, and directly apply the historical review method as the review method for the target user for the target knowledge point.

[0282] In some embodiments, the method further includes: determining at least one teaching task associated with the target user and the task time corresponding to each teaching task; selecting a target teaching task whose task time is closest to the target learning time from the teaching tasks; generating task content corresponding to at least one target knowledge point in the task content of the target teaching task; and sending the task content of the target teaching task to the target user to instruct the target user to complete the task content of the target teaching task, thereby realizing the review of the target knowledge point.

[0283] For example, the target teaching task could be an exam, so that while the target users are taking the exam, they can also review the target knowledge points, which greatly improves the learning efficiency of the target users.

[0284] Understandably, multiple thresholds corresponding to different prompt levels can be set according to needs (such as the preset thresholds and specific thresholds mentioned above), for example, the thresholds corresponding to the warning level and the forgetting level. Different prompt levels will trigger different prompt methods. Review prompts will be sent to the target user based on the prompt method corresponding to the triggered prompt level, so that the target user can intuitively understand the urgency of the current prompt through the corresponding prompt method, thereby improving the user experience.

[0285] It is understood that the above thresholds (such as the above preset thresholds or the above specific thresholds) can be determined based on the difficulty level of the target knowledge point and / or the learning level of the target user.

[0286] In some embodiments, when a target user is currently learning at least two target knowledge points, the at least two target knowledge points can be sorted based on a preset forgetting sorting strategy, such as sorting the target knowledge points from low to high memory strength, i.e., placing the target knowledge points with higher forgetting risk at the top of the list, and sending the sorted target knowledge points to the target user so that the target user can clearly understand their forgetting status of each target knowledge point.

[0287] In some embodiments, a review cycle optimization model can be constructed to evaluate the score function corresponding to the review time interval of the target knowledge point, thereby finding the review time point of the target knowledge point through the review cycle optimization model.

[0288] Please refer to Figure 8 Taking the terminal as an example, this embodiment illustrates the memory contribution information in Example 2. This embodiment provides a knowledge point interaction method, applied to the knowledge point interaction method mentioned in Example 2. The specific process of this knowledge point interaction method can be summarized in steps 801-802, where: Step 801: Obtain historical learning records, wherein the historical learning records include at least one historical learning event for the target knowledge point.

[0289] Step 802: Determine the memory contribution information of the historical learning event to the target knowledge point at the current moment, wherein the memory contribution information is the activation and decay effect value of the historical learning event on the memory intensity at the current moment.

[0290] Among them, memory contribution information can indicate the activation and decay of historical learning events at the current moment, that is, it indicates the impact of historical learning events on the user's memory of the target knowledge point at the current moment.

[0291] In this embodiment, historical learning records can be obtained, including at least one historical learning event for the target knowledge point; the memory contribution information of the historical learning event to the target knowledge point at the current moment can be determined, wherein the memory contribution information is the activation and decay effect value of the historical learning event on the memory intensity at the current moment. Thus, the target user can learn specific learning events based on the determined memory contribution information of each historical learning event to improve the user's learning efficiency of a certain knowledge point.

[0292] In some embodiments, after determining the contribution information of historical learning events to the memory of the target knowledge point at the current moment, the method further includes: Based on memory contribution information, the memory strength of the target user for the target knowledge point at the current moment is determined, and the change in memory strength is obtained based on the memory strength. The memory strength is a quantitative value of the degree of decay of the target user's memory strength for the target knowledge point at the current moment.

[0293] The change in memory intensity can be the change in memory intensity at each time point within a specific time period including the current moment. This specific time period can be a time period starting from the current moment, a time period ending from the current moment, or a time period including both historical and future periods, with the current moment as a point within the time period. The specific time period can be set according to the needs and is not limited here.

[0294] The decay quantification value can be understood as the gradual decay of the memory strength of the target knowledge point after the target user finishes the last historical learning event. The memory strength at different time points is used as the decay quantification value of the target user to represent the decay of the target user's memory of the target knowledge point.

[0295] In this embodiment, different historical learning events contribute differently to the target user's memory of the target knowledge point. Based on the memory contribution information of different historical learning events, the impact of the historical learning event on the target user's mastery of the target knowledge point at the current moment is evaluated to predict the memory intensity change within a specific time period including the current moment.

[0296] Here, the current moment is the moment when the target user's mastery of the target knowledge point needs to be predicted, which can be denoted as t. Correspondingly, the memory strength at the current moment can be denoted as λ(t).

[0297] In some embodiments, point process models (such as the Hawkes point process model) can be used to dynamically model the change in the memory strength of target users for target knowledge points over time, in order to dynamically depict the forgetting trajectory of students for each knowledge point, that is, the change in memory strength. The point process model regards each interaction event between the target user and the knowledge point as a "trigger event" and dynamically updates the memory strength according to the event time interval.

[0298] Optionally, a time-sensitive memory modeling mechanism can be built in the point process model. That is, by modeling the time point of each user's historical learning event (such as the review behavior of a knowledge point), the change in the target user's memory strength of the knowledge point can be dynamically estimated, providing a precise time basis for review triggering.

[0299] Optionally, multiple review effect modeling can be introduced into the point process model. That is, by identifying and accumulating the memory impact of each review behavior (historical learning events after the first learning of the target knowledge point), each historical learning event can produce a "memory enhancement effect" in the model, thereby more realistically simulating the memory patterns of users when learning.

[0300] In some embodiments, in addition to using Hawkes point process modeling to model the memory strength of the target user, nonparametric point process models, time-sensitive neural networks (such as Time-LSTM), Transformer models, or reinforcement learning-enhanced memory models can be used for memory modeling.

[0301] Among them, nonparametric point process models, such as the Gaussian process-modulated Poisson process, can automatically learn complex time-dependent structures and are suitable for more flexible memory strength modeling.

[0302] Among them, time-sensitive neural network models, such as Time-LSTM or T-LSTM, can introduce time interval variables when updating memory states, effectively modeling the dynamic process of forgetting behavior. Since both Time-LSTM and T-LSTM belong to Long Short-Term Memory (LSTM) networks, it is evident that by introducing the recurrent neural network structure of LSTM, it is possible to capture long-term dependencies in the sequence of historical learning events, dynamically model the complex temporal relationships and nonlinear effects between different historical learning events, and thus more accurately estimate the target user's knowledge mastery and forgetting trajectory.

[0303] Among them, the reinforcement learning reinforcement memory model can model the learning strategy as a Markov decision process (MDP), indirectly model memory changes using policy networks, and reflect memory strength through state transitions.

[0304] Among them, the Transformer model can leverage the self-attention mechanism to enable it to model the entire sequence of historical learning events in parallel, fully exploring the global dependencies between different historical learning events. It is particularly suitable for learning sequence scenarios where multiple knowledge points and multiple behavior types coexist. Thus, it does not rely on a fixed decay kernel function and can automatically capture the interaction effects and time sensitivity between different types of historical learning events through learning methods.

[0305] In some embodiments, in order to indicate the model's ability to respond to individual differences among different target users, the model can be deeply integrated with the target user's historical learning events when performing memory modeling, so as to dynamically model the model by combining the target user's learning frequency, accuracy and other behavioral characteristics of historical learning events.

[0306] In this embodiment, in order to model the model, it is necessary to comprehensively record the target user's learning process on the platform. That is, for each historical learning event of the target user on the platform, detailed event information (i.e., the information corresponding to each historical learning event in the above historical learning record) is recorded. This detailed event information includes, but is not limited to, learning timestamps (such as the time of each learning or review); learning content (such as the knowledge point identifier corresponding to a certain knowledge point); behavior type (such as first learning, review, quiz, etc.); learning results (such as correct / incorrect answers, completion time, resources used, etc.).

[0307] Then, by structuring and processing the detailed event information, a unified data standard is constructed to ensure the temporal integrity and behavioral diversity of the model input data, thereby enhancing the system's ability to model the forgetting patterns of target users on target knowledge points, and thus providing a data foundation for the calculation of memory strength and / or the determination of target learning time points.

[0308] In some embodiments, determining the contribution information of historical learning events to the memory of target knowledge points at the current moment includes: Obtain memory impact parameters and behavioral moments corresponding to historical learning events; Based on the memory impact parameters and behavioral moments of historical learning events, as well as the current moment, determine the memory contribution information of historical learning events to the target knowledge points at the current moment.

[0309] Among them, the memory influence parameter is a parameter used to indicate the influence of historical learning events on the memory of the target user when learning the target knowledge point. The memory influence parameter includes, but is not limited to, the target decay intensity and / or the target excitation intensity.

[0310] In this embodiment, since there is a corresponding implementation time, i.e. behavior time, when the target user performs a historical learning event, the historical learning event itself and the implementation time of the historical learning event will affect the target user's memory depth of the target knowledge point at the current time. That is, different historical learning events contribute different information to the memory at the current time.

[0311] It should be noted that the greater the time difference between the action moment and the current moment, the lower the impact of the historical learning event that occurred at that action moment on the target user's memory of the target knowledge point at the current moment. For example, the memory impact of the current moment, which is closer to the action moment of the historical learning event, is greater than the memory impact of the current moment, which is farther away from the action moment of the historical learning event.

[0312] Optionally, since different historical learning events have different impacts on the target user's memory of the target knowledge point, for example, the impact of the target user browsing the definition of the target knowledge point may be 30%, while the impact of the target user answering questions related to the target knowledge point may be greater, such as 40%. Therefore, the impact of historical learning events on the target user's memory of the target knowledge point at the current moment can be evaluated by the behavioral type of the historical learning event. That is, the memory impact parameter corresponding to the historical learning event is determined by the behavioral type of the historical learning event.

[0313] Specifically, the memory influence parameters include the target decay intensity and the target excitation intensity. Based on the memory influence parameters of historical learning events and the behavioral time, as well as the current time, the memory contribution information of historical learning events to the target knowledge point at the current time is determined, including: Determine the time difference between the moment of action of a historical learning event and the current moment; Based on the time difference and the target decay intensity, determine the decay indication parameters corresponding to the historical learning events; The product of the target excitation intensity and the decay indication parameter of the historical learning event is used as the memory contribution information of the historical learning event to the target knowledge point at the current moment.

[0314] It is understandable that the greater the time difference between the action moment and the current moment, the lower the impact of the historical learning events that occurred at the action moment on the target user's memory of the target knowledge points at the current moment. Therefore, the decay of the memory strength of the historical learning events at the action moment on the current moment can be calculated by the time difference between the action moment and the current moment and the target decay intensity corresponding to the historical learning events. This decay is represented by the decay indicator parameter.

[0315] It is understandable that, since historical learning events themselves will have different degrees of stimulation on the target user's memory, after obtaining the decay of the memory intensity of the historical learning event at the current moment, it is also necessary to calculate the decay indication parameter indicating the decay intensity and the target stimulation intensity indicating the stimulation of the historical learning event to obtain the degree of memory contribution of the historical learning event to the target knowledge point at the current moment, that is, the memory contribution information.

[0316] Therefore, by introducing target decay intensity and target activation intensity, we can accurately reflect the process of user memory decay over time, as well as the activation of user memory by each historical learning event. This enables us to characterize the time dependence of memory decay and behavioral activation, and solves the problems of staticity and inaccuracy in traditional forgetting assessment methods.

[0317] For example, let the action time be ts, the current time be t, the target decay intensity of the historical learning event be β, and the target excitation intensity of the historical learning event be α.

[0318] Therefore, the time difference between the action moment and the current moment is: t - ts; The decay indicator parameter corresponding to the historical learning event: e -β（t-ts） ; Information on the contribution of historical learning events to the memorization of target knowledge points: α e -β（t-ts） .

[0319] For example, if we define t1, t2, and t3 as different historical learning events, then the memory contribution information of these three historical learning events are respectively: α e -β（t-t1） α e -β（t-t2） α e -β（t-t3） Then, the memory contribution information of the three historical learning events is accumulated to obtain the memory strength of the target user for the target knowledge point at the current moment.

[0320] In some embodiments, in order to more accurately assess the memory strength at the current moment, a base memory strength may be introduced, which can be set as needed and is not limited here.

[0321] For example, if we set the current time as t, the target decay intensity of the historical learning event as β, and the target excitation intensity of the historical learning event as α, then the formula for determining the memory intensity at the current time can be obtained using the point process model: λ(t) = μ + ∑α e -β（t-ti） .

[0322] Where λ(t) is the memory strength at the current moment, μ is the basic memory strength, ti is less than t, and ti includes the behavioral moment corresponding to the historical learning event.

[0323] The formula for determining the memory strength at the current moment illustrates the impact of behavioral events (such as testing, browsing, and reading) occurring before time t on memory strength. By accumulating the contributions of different historical learning events through an exponential kernel function, the formula characterizes the activation and decay process of memory strength at the current moment caused by historical learning events at multiple different time points, demonstrating good interpretability and temporal modeling capabilities.

[0324] In some embodiments, in order to enable the target user's learning behavior to more accurately express the target user's memory changes, the target user's different learning behaviors can be identified in this embodiment.

[0325] Specifically, knowledge point interaction methods also include: Based on the target user's historical learning trajectory, the historical learning trajectory includes multiple learning behavior samples and memory labels for each learning behavior sample; The learning behavior samples and their memory labels are fitted to determine the activation intensity and decay intensity of the learning behavior corresponding to each learning behavior sample of the target user.

[0326] In this embodiment, in order to support personalized memory modeling for target users, model parameters can be adjusted in a personalized manner according to the learning behavior of target users to achieve modeling of individual differences such as forgetting speed and memory activation intensity, thereby generating customized review suggestions and improving adaptability.

[0327] Specifically, personalized parameter learning methods can be introduced, that is, designing a model parameter fitting mechanism that adapts to the individual differences of the target user, so as to realize personalized modeling of memory processes such as forgetting speed and stimulating effect.

[0328] To adapt to the memory characteristics of different target users, a personalized learning and dynamic update mechanism for model parameters can be introduced. This means that the parameters in the point process model (such as decay rate and excitation intensity) can be learned in a personalized manner and dynamically updated with new data, thereby enhancing the robustness and accuracy of the model.

[0329] In some embodiments, maximum likelihood estimation can be used to fit the historical learning trajectory. That is, the sequences corresponding to the learning behavior samples of different historical learning events in the past period of the target user are superimposed to form the learning trajectory. Each historical learning event in the sequence will have a label, i.e., a memory label, which indicates the change in memory intensity of the corresponding historical learning event. Then, maximum likelihood estimation is used to fit these data with memory labels. Then, the fitted trajectory is used as input, and the fitted trajectory is processed by the parameter adjustment model. In this way, the excitation intensity and decay intensity corresponding to different learning behaviors can be output, so as to realize the personalized parameter setting of the target user to adapt to the user's personalized needs.

[0330] In some embodiments, incremental parameter updates can also be performed to ensure the real-time performance and stability of the online learning scenario. That is, as time goes by, the target user's historical learning events will continue to increase, and the excitation intensity and decay intensity will be continuously updated, making the parameter values ​​more stable.

[0331] In addition, regularization terms can be designed to prevent overfitting and improve the generalization ability of parameters to cope with the low accuracy when there is little data.

[0332] In some embodiments, after obtaining the memory intensity change based on memory intensity, the method further includes: Displays memory change curves corresponding to changes in memory; After determining that the target user performs at least one learning behavior at any time point on the memory change curve, a memory prediction curve for the target knowledge point is obtained. The memory intensity change of the memory prediction curve is the change in the memory intensity of the target user for the target knowledge point within the estimated time period based on the learning behavior. The time points at which changes in memory intensity meet preset conditions are used as target learning time points, so that learning recommendations can be made to target users based on the target learning time points.

[0333] Then, the terminal can select time points whose memory intensity changes meet preset conditions based on the memory intensity changes at each time point, and use the selected time points as target learning time points to recommend to the target user, so as to improve the user's learning efficiency of target knowledge points.

[0334] In this embodiment, since the target user’s learning behavior related to the target knowledge point at different time points will lead to changes in the target user’s memory strength of the target knowledge point, the terminal predicts the change in the target user’s memory strength based on the user’s learning behavior related to the target knowledge point at different time points on the memory change curve, so as to generate memory prediction curves corresponding to different time points.

[0335] The memory prediction curve can be a curve within an estimated time period starting from the corresponding time node. This memory prediction curve can indicate the change in the target user's memory strength of the target knowledge point within the estimated time period after the time node.

[0336] The preset conditions can be conditions corresponding to the user's memory strength reaching or about to reach the level of mastery of the target knowledge point, or conditions corresponding to the user's memory strength change being in line with the user's high memory benefit, etc. The specific settings can be set according to the needs and are not limited here.

[0337] Understandably, by introducing a self-excitation mechanism, the learning behavior at each time point is simulated to enhance memory strength, so as to predict the degree of mastery of a certain knowledge point by the target user over time. Based on the enhancement effect of memory strength or the trend of memory strength change, the review interval of the target user for the target knowledge point is adjusted to achieve an efficient review strategy.

[0338] In some embodiments, the time points where the changes in memory intensity meet preset conditions are used as target learning time points, including: based on the memory prediction curves corresponding to each time point, selecting the time points where the average memory intensity of the target knowledge point within the estimated time period is greater than a preset threshold as target learning time points, thereby introducing a preset threshold to compare the memory intensity in the memory prediction curve with the preset threshold to automatically determine the best review time point.

[0339] In some embodiments, the time point at which the change in memory intensity meets preset conditions is used as the target learning time point, including: By comparing the changes in memory intensity on the memory change curve with the changes in memory intensity on the memory prediction curve at each time point, the magnitude of the change in memory intensity at each time point can be determined. Based on the magnitude of changes in memory intensity at each time point, time points that meet preset conditions are selected as target learning time points.

[0340] Among them, the magnitude of memory intensity change is used to indicate the relative situation between the change in memory intensity after learning at a time node and the change in memory intensity before learning at a time node (memory change curve). Since memory intensity generally tends to increase after learning at a time node, the magnitude of memory intensity change can also be used as the increase in memory intensity when learning at a time node.

[0341] In this embodiment, the memory prediction curve based on each time point is compared with the memory change curve to clarify the magnitude of memory intensity change at each time point. Based on the magnitude of memory intensity change at each time point, time points that meet preset conditions are selected to automatically determine the best review time.

[0342] Specifically, based on the magnitude of change in memory intensity at each time point, time points that meet preset conditions are selected as target learning time points, including: selecting the time point with the largest magnitude of change in memory intensity as the target learning time point.

[0343] Specifically, based on the magnitude of change in memory intensity at each time point, time points that meet preset conditions are selected as target learning time points, including: selecting time points where the magnitude of change in memory intensity is greater than a preset magnitude threshold as target learning time points.

[0344] Specifically, the changes in memory intensity on the memory change curve are compared with the changes in memory intensity on the memory prediction curve at each time point to determine the magnitude of memory intensity change at each time point, including: Calculate the difference between the memory intensity of the memory change curve and the memory intensity of the memory prediction curve at a specified time point within the estimated time period; The difference at a specified moment corresponding to each time node is taken as the magnitude of change in memory strength; or Calculate the difference between the memory intensity of the memory change curve and the memory intensity of the memory prediction curve at each time point within the estimated time period. The differences between the time windows corresponding to each time node are averaged to use the result as the magnitude of change in memory intensity.

[0345] The time window can be a window that contains at least one time node, and different time windows contain the same number of time nodes. For example, all time windows contain two time nodes.

[0346] Specifically, calculating the difference between the memory intensity of the memory change curve and the memory intensity of the memory prediction curve corresponding to each time node within each time window of the estimated time period can include: determining the first difference between the memory intensity of any two time nodes within the corresponding time window on the memory change curve, determining the second difference between the memory intensity of the same two time nodes within the corresponding time window on the memory prediction curve, and then using the difference between the first difference and the second difference as the difference between the memory intensity of the memory change curve within the corresponding time window and the memory intensity of the memory prediction curve corresponding to each time node.

[0347] Specifically, calculating the difference between the memory intensity of the memory change curve and the memory intensity of the memory prediction curve corresponding to each time node within each time window of the estimated time period can include: determining the third difference between any time node in the corresponding time window on the memory change curve and the memory intensity of the same time node in the corresponding time window on the memory prediction curve; and then using the difference between the third differences as the difference between the memory intensity of the memory change curve in the corresponding time window and the memory intensity of the memory prediction curve corresponding to each time node.

[0348] For example, the process of comparing the changes in memory intensity on the memory change curve with the changes in memory intensity on the memory prediction curve at each time point is shown in the following formula:

[0349] Where t+T is the time range for comparison, such as t being a time node and T being a time interval (such as a week). This shows the changes in memory intensity at corresponding time points on the memory prediction curve. This represents the change in memory intensity at corresponding time points on the memory change curve.

[0350] Understandably, by comparing the memory change curve and the memory prediction curve, we can obtain the memory improvement benefits of learning about target knowledge points at different time points. Based on these benefits, we can select appropriate time points as target learning time points to recommend learning about target knowledge points to target users.

[0351] Then, the argmax function can be used to filter the benefits at different time points to determine the optimal review time, i.e., the target learning time. The method of using a benefit function (such as the argmax function) to filter the benefits at different time points is as follows:

[0352] Among them, t The time point with the greatest benefit, such as the time point with the greatest change in memory intensity.

[0353] Optionally, in addition to the revenue function, the learning cost of the target knowledge point can be calculated to obtain a comprehensive score for the target knowledge point. A review plan can be arranged based on the comprehensive score. The learning cost includes, but is not limited to, the average learning time of all target users on a knowledge point, the first-time learning pass rate of the knowledge point, the complexity of the knowledge point, and the average difficulty of related test questions under the knowledge point.

[0354] This embodiment also provides a learning time point recommendation device applied to an intelligent learning system. The intelligent learning system includes a first display interface and a second display interface. Specifically, the learning time point recommendation device can be integrated into a terminal device. For example, as... Figure 9 As shown, the learning time point recommendation device may include: The identifier display module 901 is used to display at least one learning recommendation identifier on the first display interface, wherein the learning recommendation identifier is used to instruct the target user to learn the target knowledge point at the target learning time point; The content display module 902 is configured to respond to a trigger operation for at least one learning recommendation identifier to display learning content corresponding to the target learning time point on a second display interface.

[0355] In some embodiments, The signage display module 901 is specifically used for: The memory change curve is displayed on the first display interface; A learning recommendation icon is displayed at at least one target learning time point on the memory change curve. The target learning time point is the time point in the future when the target user's memory strength of the target knowledge point meets the preset conditions after learning at least one learning content.

[0356] In some embodiments, the preset conditions are specifically used for: The average memory strength of target users for the target knowledge points within the estimated time period is greater than a preset threshold; or, The increase in the intensity of memory of target knowledge points by target users at the target learning time point is higher than at other time points.

[0357] In some embodiments, the learning time point recommendation device further includes a curve display module, which is specifically used for: In response to a trigger operation targeting any time point on the memory change curve, a memory prediction curve starting from the triggered time point is displayed on the first display interface. The memory prediction curve predicts the change in the memory strength of the target knowledge point after the target user performs at least one learning behavior at the triggered time point.

[0358] In some embodiments, the curve display module is specifically used for: At least one candidate learning content is displayed at the triggered time node on the first display interface; In response to a selection instruction for at least one candidate learning content, a memory prediction curve generated based on the selected candidate learning content is displayed on a first display interface.

[0359] In some embodiments, the identifier display module 901 is specifically used for: In response to a target user's memory review event for a target knowledge point, the first display screen shows the target user's memory change curve during the target learning period, which includes historical and future time periods.

[0360] In some embodiments, the identifier display module 901 is specifically used for: In response to a knowledge point viewing action triggered by a target user, display at least one target knowledge point, the target user's historical learning events for the target knowledge point at at least one historical moment, and information on the contribution of the historical learning events to the target user's memory of the target knowledge point; In response to the target user's memory trigger action on the target knowledge point, the system displays the memory change curve of the target user during the target learning period.

[0361] In some embodiments, the learning time point recommendation device further includes an information sending module, which is specifically used for: Send review reminders to target users at the target learning time to instruct them to process the reminders at that time.

[0362] As can be seen from the above, this can be applied to an intelligent learning system. The intelligent learning system includes a first display interface and a second display interface. At least one learning recommendation icon is displayed on the first display interface, indicating to the target user that they should learn a target knowledge point at a target learning time. In response to a trigger operation on the at least one learning recommendation icon, learning content corresponding to the target learning time is displayed on the second display interface. Thus, by using the learning recommendation icon, learning recommendations for the target knowledge point are provided to the target user at an appropriate time, thereby improving the user's learning efficiency for that knowledge point.

[0363] This embodiment also provides a learning time point recommendation device, which can be integrated into a terminal device. For example, such as... Figure 10 As shown, the learning time point recommendation device may include: The curve determination module 1001 is used to determine the memory prediction curve for the target knowledge point obtained after the target user performs at least one learning behavior at any time node on the memory change curve. The memory prediction curve is the change in the memory intensity of the target user for the target knowledge point within the estimated time period based on the learning behavior. The node determination module 1002 is used to select time points where the changes in memory intensity meet preset conditions as target learning time points, so as to make learning recommendations to target users based on the target learning time points.

[0364] In some embodiments, the node determination module 1002 is specifically used for: By comparing the changes in memory intensity on the memory change curve with the changes in memory intensity on the memory prediction curve at each time point, the magnitude of the change in memory intensity at each time point can be determined. Based on the magnitude of changes in memory intensity at each time point, time points that meet preset conditions are selected as target learning time points.

[0365] In some embodiments, the node determination module 1002 is specifically used for: The time point with the greatest change in memory intensity is selected as the target learning time point.

[0366] In some embodiments, the node determination module 1002 is specifically used for: Calculate the difference between the memory intensity of the memory change curve and the memory intensity of the memory prediction curve at a specified time point within the estimated time period; The difference at a specified moment corresponding to each time node is taken as the magnitude of change in memory strength; or Calculate the difference between the memory intensity of the memory change curve and the memory intensity of the memory prediction curve at each time point within the estimated time period. The differences between the time windows corresponding to each time node are averaged to use the result as the magnitude of change in memory intensity.

[0367] In some embodiments, the curve determination module 1001 is specifically used for: Based on learning behavior and time points, determine the memory contribution information of learning behavior at each estimated time point within the estimated time period; Based on memory contribution information, the memory strength of the target user for the target knowledge point is determined at each estimated time point, and a memory prediction curve for the target knowledge point is obtained based on the memory strength at each estimated time point.

[0368] In some embodiments, the curve determination module 1001 is specifically used for: Obtain memory-influencing parameters corresponding to learning behaviors; Based on the memory impact parameters and time points of learning behavior, as well as the estimated time points, we determine the contribution information of learning behavior to the memory of target knowledge points at the estimated time points.

[0369] In some embodiments, the memory influence parameters include the target attenuation intensity and the target excitation intensity, and the curve determination module 1001 is specifically used for: Determine the time difference between the predicted and actual time points; Based on the time difference and the target decay intensity, determine the decay indication parameters corresponding to the learning behavior; The product of the target stimuli intensity and the decay indicator parameter of the learning behavior is used as the information on the contribution of the learning behavior to the memory of the target knowledge point at the predicted time point.

[0370] In some embodiments, the learning time point recommendation device further includes a data fitting module, which is specifically used for: Obtain the target user's historical learning trajectory, which includes multiple learning behavior samples and memory labels for each learning behavior sample; Based on the target user's historical learning trajectory, each learning behavior sample and its memory label are fitted to determine the activation intensity and decay intensity corresponding to the learning behavior of each learning behavior sample of the target user.

[0371] In some embodiments, the curve determination module 1001 is specifically used for: Obtain information on the historical memory contribution of at least one historical learning behavior of the target user at the estimated time point; By accumulating historical memory contribution information and memory contribution information corresponding to the estimated time points, the memory strength of the target user for the target knowledge point at the estimated time point is obtained.

[0372] As can be seen from the above, a memory prediction curve for a target knowledge point can be obtained by determining that the target user performs at least one learning behavior at any time point on the memory change curve. The memory prediction curve represents the change in the target user's memory intensity for the target knowledge point within an estimated time period, generated based on the learning behavior. The time point where the change in memory intensity meets the preset conditions is taken as the target learning time point, and learning recommendations are made to the target user at the target learning time point. By predicting the memory prediction curve generated by the target user's learning at the time point, and evaluating whether the memory prediction curve can be used as the target learning time point for recommending learning to the user, the user can be recommended the target knowledge point at the appropriate time, thereby improving the user's learning efficiency for a certain knowledge point.

[0373] This embodiment also provides a knowledge point interaction device, which can be integrated into a terminal device. For example, such as Figure 11 As shown, the knowledge point interactive device may include: The information display module 1101 is used to respond to the knowledge point viewing operation triggered by the target user, and to display at least one target knowledge point, the target user's historical learning events for the target knowledge point at at least one historical moment, and the contribution information of each historical learning event to the target user's memory of the target knowledge point at the current moment.

[0374] As can be seen from the above, by responding to the knowledge point viewing operation triggered by the target user, at least one target knowledge point, the target user's historical learning events for the target knowledge point at at least one historical moment, and the memory contribution information of each historical learning event to the target user's mastery of the target knowledge point at the current moment can be displayed. Thus, the target user can learn specific learning events based on the memory contribution information of each historical learning event, thereby improving the user's learning efficiency for a certain knowledge point.

[0375] This embodiment also provides a knowledge point interaction device, which can be integrated into a terminal device. For example, such as Figure 12 As shown, the knowledge point interactive device may include: The record acquisition module 1201 is used to acquire historical learning records, wherein the historical learning records include at least one historical learning event for the target knowledge point; The information determination module 1202 is used to determine the memory contribution information of historical learning events to the target knowledge point at the current moment, wherein the memory contribution information is the activation and decay effect value of historical learning events on the memory intensity at the current moment.

[0376] As can be seen from the above, by acquiring historical learning records, which include at least one historical learning event for the target knowledge point, and determining the memory contribution information of the historical learning event to the target knowledge point at the current moment, where the memory contribution information is the activation and decay effect value of the historical learning event on the memory intensity at the current moment, the target user can learn specific learning events based on the determined memory contribution information of each historical learning event to learn the target knowledge point, thereby improving the user's learning efficiency for a certain knowledge point.

[0377] Terminal devices include smartphones, tablets, laptops, touchscreens, game consoles, personal computers (PCs), and personal digital assistants (PDAs). Alternatively, electronic devices can serve as servers.

[0378] like Figure 13 As shown, Figure 13 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device 1300 includes a processor 1301 with one or more processing cores, a memory 1302 with one or more computer-readable storage media, and a computer program stored on the memory 1302 and executable on the processor. The processor 1301 and the memory 1302 are electrically connected. Those skilled in the art will understand that the electronic device structure shown in the figure does not constitute a limitation on the electronic device, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0379] The processor 1301 is the control center of the electronic device 1300. It connects various parts of the electronic device 1300 via various interfaces and lines. By running or loading software programs and / or units stored in the memory 1302, and by calling data stored in the memory 1302, it executes various functions and processes data of the electronic device 1300, thereby providing overall monitoring of the electronic device 1300. The processor 1301 can be a central processing unit (CPU), a graphics processing unit (GPU), a network processor (NP), etc., and can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application.

[0380] In this embodiment, the processor 1301 in the electronic device 1300 loads the instructions corresponding to the processes of one or more applications into the memory 1302 according to the following steps, and the processor 1301 runs the applications stored in the memory 1302 to realize various functions, such as: At least one learning recommendation icon is displayed on the first display interface, wherein the learning recommendation icon is used to instruct the target user to learn the target knowledge point at the target learning time point; In response to a triggering action for at least one learning recommendation identifier, learning content corresponding to the target learning time point is displayed on a second display interface.

[0381] For example: After determining that the target user performs at least one learning behavior at any time point on the memory change curve, the memory prediction curve for the target knowledge point is obtained. The memory prediction curve is the change in the memory intensity of the target user for the target knowledge point within the estimated time period, which is generated based on the learning behavior. The time points at which changes in memory intensity meet preset conditions are used as target learning time points, so that learning recommendations can be made to target users based on the target learning time points.

[0382] For example: In response to a knowledge point viewing action triggered by a target user, display at least one target knowledge point, the target user's historical learning events for the target knowledge point at at least one historical moment, and the contribution information of each historical learning event to the target user's memory of the target knowledge point at the current moment.

[0383] For example: Obtain historical learning records, which include at least one historical learning event related to the target knowledge point; Determine the memory contribution information of historical learning events to the target knowledge point at the current moment, where the memory contribution information is the activation and decay effect value of historical learning events on the memory intensity at the current moment.

[0384] Therefore, the electronic device 1300 provided in this embodiment can bring the following technical effects: improve the user's learning efficiency for a certain knowledge point.

[0385] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0386] Optional, such as Figure 13 As shown, the electronic device 1300 also includes: a touch display screen 1303, a radio frequency circuit 1304, an audio circuit 1305, an input unit 1306, and a power supply 1307. The processor 1301 is electrically connected to the touch display screen 1303, the radio frequency circuit 1304, the audio circuit 1305, the input unit 1306, and the power supply 1307. Those skilled in the art will understand that... Figure 13 The electronic device structure shown does not constitute a limitation on the electronic device and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0387] The touch display screen 1303 can be used to display a graphical user interface (GUI) and receive operation commands generated by the user interacting with the GUI. The touch display screen 1303 may include a display panel and a touch panel. The display panel can be used to display information input by the user or information provided to the user, as well as various graphical user interfaces of the electronic device. These graphical user interfaces can be composed of graphics, text, icons, video, and any combination thereof. Optionally, the display panel can be configured using a liquid crystal display (LCD), organic light-emitting diode (OLED), or other similar technologies. The touch panel can be used to collect touch operations performed by the user on or near it (such as operations performed by the user using a finger, stylus, or any suitable object or accessory on or near the touch panel), generate corresponding operation commands, and execute the corresponding program according to the operation commands. Optionally, the touch panel may include two parts: a touch detection device and a touch controller. The touch detection device detects the user's touch location and the signal generated by the touch operation, transmitting the signal to the touch controller. The touch controller receives touch information from the touch detection device, converts it into touch point coordinates, and sends it to the processor 1301. It can also receive and execute commands from the processor 1301. The touch panel can cover the display panel. When the touch panel detects a touch operation on or near it, it transmits the information to the processor 1301 to determine the type of touch event. Subsequently, the processor 1301 provides corresponding visual output on the display panel based on the type of touch event. In this embodiment, the touch panel and the display panel can be integrated into the touch display screen 1303 to achieve input and output functions. However, in some embodiments, the touch panel and the touch display screen 1303 can be implemented as two independent components to achieve input and output functions. That is, the touch display screen 1303 can also be used as part of the input unit 1306 to achieve input functions.

[0388] The radio frequency circuit 1304 can be used to transmit and receive radio frequency signals to establish wireless communication with network devices or other electronic devices, and to transmit and receive signals with network devices or other electronic devices.

[0389] Audio circuit 1305 can be used to provide an audio interface between a user and an electronic device via a speaker and a microphone. Audio circuit 1305 can convert received audio data into electrical signals and transmit them to the speaker, where the speaker converts them into sound signals for output. Conversely, the microphone converts collected sound signals into electrical signals, which are then received by audio circuit 1305, converted back into audio data, and then processed by processor 1301 before being transmitted via radio frequency circuit 1304 to, for example, another electronic device, or output to memory 1302 for further processing. Audio circuit 1305 may also include an earphone jack to provide communication between peripheral headphones and electronic devices.

[0390] The input unit 1306 can be used to receive input numbers, characters, or user characteristic information (such as fingerprints, iris, facial information, etc.), and to generate keyboard, mouse, joystick, optical, or trackball signal inputs related to user settings and function control.

[0391] Power supply 1307 is used to supply power to various components of electronic device 1300. Optionally, power supply 1307 can be logically connected to processor 1301 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. Power supply 1307 may also include one or more DC or AC power supplies, recharging systems, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0392] although Figure 13 As not shown in the diagram, the electronic device 1300 may also include a camera, sensor, wireless fidelity module, Bluetooth module, etc., which will not be described in detail here.

[0393] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0394] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0395] Therefore, embodiments of this application provide a computer-readable storage medium storing multiple computer programs that can be loaded by a processor to execute any of the learning time point recommendation methods provided in embodiments of this application. The computer program can execute the steps of the learning time point recommendation method as follows: At least one learning recommendation icon is displayed on the first display interface, wherein the learning recommendation icon is used to instruct the target user to learn the target knowledge point at the target learning time point; In response to a triggering action for at least one learning recommendation identifier, learning content corresponding to the target learning time point is displayed on a second display interface.

[0396] This computer program can also perform the following steps for recommending learning time points: After determining that the target user performs at least one learning behavior at any time point on the memory change curve, the memory prediction curve for the target knowledge point is obtained. The memory prediction curve is the change in the memory intensity of the target user for the target knowledge point within the estimated time period, which is generated based on the learning behavior. The time points at which changes in memory intensity meet preset conditions are used as target learning time points, so that learning recommendations can be made to target users based on the target learning time points.

[0397] This computer program can also execute the following steps for interactive knowledge point methods: In response to a knowledge point viewing action triggered by a target user, display at least one target knowledge point, the target user's historical learning events for the target knowledge point at at least one historical moment, and the contribution information of each historical learning event to the target user's memory of the target knowledge point at the current moment.

[0398] This computer program can also execute the following steps for interactive knowledge point methods: Obtain historical learning records, which include at least one historical learning event related to the target knowledge point; Determine the memory contribution information of historical learning events to the target knowledge point at the current moment, where the memory contribution information is the activation and decay effect value of historical learning events on the memory intensity at the current moment.

[0399] As can be seen, the computer program can be loaded by the processor to execute any of the learning time point recommendation methods provided in the embodiments of this application, thereby bringing about the following technical effects: improving the user's learning efficiency for a certain knowledge point.

[0400] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0401] The computer-readable storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0402] Since the computer program stored in the computer-readable storage medium can execute any of the learning time point recommendation methods provided in the embodiments of this application, it can achieve the beneficial effects that any of the learning time point recommendation methods provided in the embodiments of this application can achieve, as detailed in the preceding embodiments, and will not be repeated here.

[0403] According to one aspect of this application, a computer program product or computer program is also provided, comprising computer instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the electronic device to perform the methods provided in the various optional implementations of the above embodiments.

[0404] In the above embodiments of the learning time point recommendation device, computer-readable storage medium, electronic device, and computer program product, the descriptions of each embodiment have different focuses. Parts not described in detail in a particular embodiment can be referred to in the relevant descriptions of other embodiments. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes and beneficial effects of the learning time point recommendation device, computer-readable storage medium, computer program product, electronic device, and their corresponding units described above can be referred to the description of the learning time point recommendation method in the above embodiments, and will not be repeated here.

[0405] The foregoing has provided a detailed description of a learning time point recommendation method, apparatus, electronic device, computer-readable storage medium, and computer program product provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for recommending learning time points, characterized in that, This is applied to an intelligent learning system, which includes a first display interface and a second display interface, including: At least one learning recommendation icon is displayed on the first display interface, wherein the learning recommendation icon is used to instruct the target user to learn the target knowledge point at the target learning time point; In response to a triggering operation for at least one learning recommendation identifier, learning content corresponding to the target learning time point is displayed on a second display interface; The provision that at least one learning recommendation icon is displayed on the first display interface includes: The memory change curve is displayed on the first display interface; A learning recommendation icon is displayed at at least one target learning time point on the memory change curve. The target learning time point is a time point in the future where the memory strength of the target knowledge point after the target user has learned at least one learning content meets the preset conditions.

2. The method according to claim 1, characterized in that, The preset conditions include: The average memory strength of the target users for the target knowledge points within the estimated time period is greater than a preset threshold; or, The target user's memory intensity of the target knowledge point at the target learning time point increased more than at other time points.

3. The method according to claim 1, characterized in that, After displaying the memory change curve on the first display interface, the method further includes: In response to a trigger operation on any time point of the memory change curve, a memory prediction curve starting from the triggered time point is displayed on the first display interface. The memory prediction curve is a prediction of the change in memory intensity of the target knowledge point after the target user performs at least one learning behavior at the triggered time point.

4. The method according to claim 3, characterized in that, The step of displaying the memory prediction curve on the first display interface, starting from the triggered time point, includes: At least one candidate learning content is displayed at the triggered time node on the first display interface; In response to a selection instruction for at least one candidate learning content, a memory prediction curve generated based on the selected candidate learning content is displayed on the first display interface.

5. The method according to claim 1, characterized in that, Displaying the memory change curve on the first display interface includes: In response to the target user's memory viewing event for the target knowledge point, the first display interface displays the target user's memory change curve within the target learning time period, which includes historical time periods and future time periods.

6. The method according to claim 5, characterized in that, The step of responding to the target user's memory review event for the target knowledge point by displaying the target user's memory change curve during the target learning time period on the first display interface includes: In response to the knowledge point viewing operation triggered by the target user, at least one target knowledge point, the target user's historical learning events for the target knowledge point at at least one historical moment, and the contribution information of the historical learning events to the target user's memory of the target knowledge point; In response to the target user's memory trigger operation on the target knowledge point, the memory change curve of the target user during the target learning time period is displayed.

7. The method according to claim 1, characterized in that, The method further includes: Send a review prompt message to the target user at the target learning time point to instruct the target user to process the review prompt message at the target learning time point.

8. A method for recommending learning time points, characterized in that, include: After determining that the target user performs at least one learning behavior at any time point on the memory change curve, a memory prediction curve for the target knowledge point is obtained, wherein the memory prediction curve is generated based on the learning behavior to show the change in the target user's memory intensity for the target knowledge point within an estimated time period. The time points at which the changes in memory intensity meet preset conditions are taken as target learning time points, so as to make learning recommendations to the target users based on the target learning time points.

9. The method according to claim 8, characterized in that, The step of using the time point where the change in memory intensity meets the preset conditions as the target learning time point includes: The memory intensity change of the memory change curve and the memory intensity change of the memory prediction curve corresponding to each time point are compared to determine the magnitude of memory intensity change at each time point. Based on the magnitude of changes in memory intensity at each time point, time points that meet preset conditions are selected as target learning time points.

10. The method according to claim 9, characterized in that, The selection of target learning time points based on the magnitude of memory intensity changes at each time point, meeting preset conditions, includes: The time point with the greatest change in memory intensity is selected as the target learning time point.

11. The method according to claim 9, characterized in that, The step of comparing the memory intensity changes in the memory change curve with the memory intensity changes in the memory prediction curve at each time point to determine the magnitude of memory intensity change at each time point includes: Calculate the difference between the memory intensity of the memory change curve and the memory intensity of the memory prediction curve at a specified time within the estimated time period; The difference at the specified time point corresponding to each time node is taken as the magnitude of the change in memory strength; or Calculate the difference between the memory intensity of the memory change curve and the memory intensity of the memory prediction curve corresponding to each time node within each time window of the estimated time period; The differences between the time windows corresponding to each time node are averaged to use the result as the magnitude of change in memory intensity.

12. The method according to any one of claims 8 to 11, characterized in that, The memory prediction curve for the target knowledge point is obtained after determining that the target user has performed at least one learning behavior at any time point on the memory change curve, including: Based on the learning behavior and the time nodes, determine the memory contribution information of the learning behavior at each estimated time node within the estimated time period; Based on the memory contribution information, the memory intensity of the target user for the target knowledge point at each estimated time point is determined, so as to obtain the memory prediction curve for the target knowledge point based on the memory intensity at each estimated time point.

13. The method according to claim 12, characterized in that, The step of determining the memory contribution information of the learning behavior at each estimated time point within the estimated time period based on the learning behavior and the time points includes: Obtain the memory impact parameters corresponding to the learning behavior; Based on the memory impact parameters of the learning behavior, the time node, and the estimated time node, the memory contribution information of the learning behavior to the target knowledge point at the estimated time node is determined.

14. The method according to claim 13, characterized in that, The memory impact parameters include target decay intensity and target excitation intensity. The determination of the memory impact parameters based on the learning behavior, the time node, and the estimated time node, and the determination of the learning behavior's contribution to the memory of the target knowledge point at the estimated time node, includes: Determine the time difference between the stated time point and the estimated time point; Based on the time difference and the target attenuation intensity, determine the attenuation indication parameter corresponding to the learning behavior; The product of the target excitation intensity of the learning behavior and the decay indication parameter is used as the memory contribution information of the learning behavior to the target knowledge point at the estimated time node.

15. The method according to claim 14, characterized in that, The method further includes: Obtain the target user's historical learning trajectory, which includes multiple learning behavior samples and memory labels for each learning behavior sample; Based on the target user's historical learning trajectory, each learning behavior sample and its memory label are fitted to determine the excitation intensity and decay intensity corresponding to the learning behavior to which each learning behavior sample belongs.

16. The method according to claim 12, characterized in that, The step of determining the memory strength of the target user for the target knowledge point at each estimated time point based on the memory contribution information includes: Obtain the historical memory contribution information of at least one historical learning behavior of the target user at the estimated time point; By accumulating the historical memory contribution information and the memory contribution information corresponding to the estimated time node, the memory intensity of the target user for the target knowledge point at the estimated time node is obtained.

17. A learning time point recommendation device, characterized in that, Applied to an intelligent learning system, the intelligent learning system includes a first display interface and a second display interface, and the device includes: The identifier display module is used to display at least one learning recommendation identifier on the first display interface, wherein the learning recommendation identifier is used to instruct the target user to learn the target knowledge point at the target learning time point; The content display module is used to respond to a trigger operation for at least one learning recommendation identifier to display learning content corresponding to the target learning time point on the second display interface; The step of displaying at least one learning recommendation icon on the first display interface includes: displaying a memory change curve on the first display interface; and displaying a learning recommendation icon at at least one target learning time point on the memory change curve, wherein the target learning time point is a time point in the future when the target user's memory strength of the target knowledge point meets preset conditions after learning at least one learning content.

18. An electronic device, characterized in that, The system includes a processor and a memory, the memory storing multiple instructions; the processor loads instructions from the memory to perform the steps of the learning time point recommendation method as described in any one of claims 1 to 16.

19. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a plurality of instructions adapted for loading by a processor to perform the steps of the learning time point recommendation method as described in any one of claims 1 to 16.

Images