Multi-terminal collaborative remote academic tutoring resource scheduling method and system

Abstract

The invention provides a multi-terminal collaborative remote academic tutoring resource scheduling method and system, and the method comprises the steps: carrying out the recognition of the cognitive load of a current student based on the interaction data and tutoring scene information in a tutoring session request initiated by a tutoring terminal in a remote academic tutoring process, and obtaining the cognitive load level of the current student, performing label matching on the cognitive load level and a tutoring strategy template of remote academic tutoring to obtain a hierarchical scheduling instruction of the main tutoring terminal; differentiated resource distribution is carried out on remote academic tutoring resources among the auxiliary tutoring terminals through resource adaptation features in interaction feedback data of the auxiliary tutoring terminals, and resource distribution schemes of the auxiliary tutoring terminals are obtained; and according to each resource distribution scheme, carrying out real-time adjustment on the distribution proportion and the terminal adaptive format in the hierarchical scheduling instruction until the remote academic tutoring session ends. Based on the scheme, cognitive load-driven differentiated resource scheduling in remote academic tutoring can be realized.

Description

Technical Field

[0001] This application relates to the field of resource scheduling technology, and more specifically, to a method and system for scheduling remote academic tutoring resources in a multi-terminal collaborative manner. Background Technology

[0002] Remote academic tutoring refers to academic guidance activities that utilize internet communication technology and smart terminal devices, where teachers transmit teaching resources in real time from a main tutoring terminal to multiple auxiliary tutoring terminals on the student's end. This enables teachers and students to interact remotely and collaborate across multiple screens. Teachers can explain knowledge points, answer homework questions, and track learning progress. Students can receive explanations, complete exercises, and receive real-time feedback through different terminals, thus breaking through the limitations of time and space and ensuring the continuity and personalization of academic support.

[0003] Existing technologies typically employ a uniform resource delivery method, failing to dynamically adjust resource allocation strategies based on students' real-time cognitive load. In remote tutoring, materials are often delivered to all students using fixed content organization and standardized formats, without differentiating between students' varying cognitive burdens. When students are under high cognitive load, continuing to deliver challenging, high-density resources leads to decreased attention, comprehension difficulties, and even learning frustration. Conversely, maintaining a regular pace of resource delivery when students are under low cognitive load, without timely additions of challenging or supplementary content, results in low learning efficiency. Therefore, it is difficult to accurately match resource allocation strategies with students' actual learning states, ultimately impacting the personalization and effectiveness of remote tutoring. Consequently, achieving cognitive load-driven differentiated resource allocation in remote tutoring to improve the adaptability of multi-terminal collaborative tutoring has become a significant challenge for the industry. Summary of the Invention

[0004] This application provides a method and system for scheduling remote academic tutoring resources in a multi-terminal collaborative manner, which can realize differentiated resource scheduling driven by cognitive load in remote academic tutoring, thereby improving the adaptation efficiency of multi-terminal collaborative tutoring.

[0005] Firstly, this application provides a method for scheduling remote academic tutoring resources through multi-terminal collaboration, including: During remote academic tutoring, the tutoring session requests initiated by the main tutoring terminal are obtained, and the interactive feedback data of each tutoring terminal is collected periodically. Based on the interactive data and tutoring scenario information in the tutoring session request, the cognitive load of the current student is identified to obtain the cognitive load level of the current student. The cognitive load level is then matched with the tutoring strategy template of remote academic tutoring to obtain the hierarchical scheduling instruction of the main tutoring terminal. By using the resource adaptation characteristics in the various interactive feedback data, the remote academic tutoring resources among the various tutoring terminals are distributed in a differentiated manner to obtain the resource distribution scheme for each tutoring terminal. The distribution ratio and terminal adaptation format in the hierarchical scheduling instructions are adjusted in real time according to each resource distribution scheme until the remote academic tutoring session ends.

[0006] In some embodiments, identifying the current student's cognitive load level based on the interaction data and tutoring scenario information in the tutoring session request specifically includes: Multimodal interaction data and tutoring scenario information are extracted from the tutoring session request. The interaction data includes the student's speech spectrum features, writing trajectory curvature, and screen gaze point heatmap. Based on the question difficulty coefficient and knowledge point association graph in the interaction data and tutoring scenario information, the cognitive load of the current student is evaluated in multiple dimensions to obtain the probability distribution vector of multiple cognitive load dimensions. All probability distribution vectors are fused across dimensions, and the fusion results are then mapped to a preset cognitive load level range to obtain the current student's cognitive load level.

[0007] In some embodiments, tag matching between the cognitive load level and the remote academic tutoring strategy template to obtain the hierarchical scheduling instructions for the main tutoring terminal specifically includes: A pre-built tutoring strategy template library is constructed, and each tutoring strategy template in the library is associated with a cognitive load level range, an academic label type, and a corresponding resource scheduling priority sequence. The current student's cognitive load level and academic label are jointly encoded to obtain the matching degree with each tutoring strategy template in the tutoring strategy template library, and the target template with the highest matching degree is selected. The predefined resource scheduling priority sequence in the target template is parsed to generate resource allocation weights for different auxiliary terminal categories, thereby obtaining hierarchical scheduling instructions for the primary auxiliary terminal.

[0008] In some embodiments, the remote academic tutoring resources among various tutoring terminals are distributed in a differentiated manner based on the resource adaptation characteristics in the various interactive feedback data, resulting in a resource distribution scheme for each tutoring terminal. Specifically, this scheme includes: Extract the resource adaptation features of each tutoring terminal from the interactive feedback data; Based on the resource adaptation characteristics, all auxiliary terminals are clustered and grouped. Terminals within the same group use the same resource encapsulation format, while different groups use different resource transmission strategies. For each group, a corresponding resource encoding bitrate, resource slice duration, and resource presentation style are generated, thereby obtaining the resource distribution scheme for each auxiliary and tutoring terminal.

[0009] In some embodiments, adjusting the distribution ratio and terminal adaptation format in the hierarchical scheduling instruction in real time according to each resource distribution scheme until the remote academic tutoring session ends specifically includes: Extract the resource reception feedback parameters of each tutoring terminal from the resource distribution scheme of each tutoring terminal; When the resource reception feedback parameters of any auxiliary tutoring terminal trigger the preset adjustment threshold, the optimal distribution ratio and terminal adaptation format are recalculated based on the resource adaptation characteristics of the corresponding auxiliary tutoring terminal. The adjusted distribution ratio and terminal adaptation format are synchronized to the hierarchical scheduling instruction, the resource scheduling parameters of the corresponding terminal are updated, and the feedback data of the next cycle is monitored until the session ends.

[0010] In some embodiments, the primary tutoring terminal is a teaching-leading terminal based on the teacher's identity.

[0011] In some embodiments, the tutoring terminal is a collaborative participation terminal based on student identity and tutoring identity.

[0012] Secondly, this application provides a multi-terminal collaborative remote academic tutoring resource scheduling system, including a resource distribution unit, wherein the resource distribution unit includes: The acquisition module is used to acquire tutoring session requests initiated by the main tutoring terminal during remote academic tutoring, and periodically collect interactive feedback data from each tutoring terminal. The processing module is used to identify the cognitive load of the current student based on the interaction data and tutoring scenario information in the tutoring session request, obtain the cognitive load level of the current student, match the cognitive load level with the tutoring strategy template of remote academic tutoring, and obtain the hierarchical scheduling instruction of the main tutoring terminal. The processing module is also used to distribute remote academic tutoring resources between various tutoring terminals in a differentiated manner based on the resource adaptation features in each interactive feedback data, so as to obtain a resource distribution scheme for each tutoring terminal. The execution module is used to adjust the distribution ratio and terminal adaptation format in the hierarchical scheduling instructions in real time according to each resource distribution scheme until the remote academic tutoring session ends.

[0013] Thirdly, this application provides a computer device, which includes a memory and a processor. The memory is used to store a computer program, and the processor is used to call and run the computer program from the memory, so that the computer device executes the above-described multi-terminal collaborative remote academic tutoring resource scheduling method.

[0014] Fourthly, this application provides a computer-readable storage medium storing instructions or code that, when executed on a computer, enable the computer to implement the aforementioned multi-terminal collaborative remote academic tutoring resource scheduling method.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: This application provides a method and system for scheduling remote academic tutoring resources in a multi-terminal collaborative manner. During remote academic tutoring, the system acquires tutoring session requests initiated by the main tutoring terminal and periodically collects interactive feedback data from each auxiliary tutoring terminal. Based on the interactive data and tutoring scenario information in the tutoring session requests, the system identifies the current student's cognitive load level. The system then performs tag matching between the cognitive load level and the remote academic tutoring strategy template to obtain a hierarchical scheduling instruction for the main tutoring terminal. Furthermore, the system distributes remote academic tutoring resources among the auxiliary tutoring terminals in a differentiated manner based on the resource adaptation features in the interactive feedback data, resulting in a resource distribution scheme for each auxiliary tutoring terminal. Finally, the system adjusts the distribution ratio and terminal adaptation format in the hierarchical scheduling instruction in real time according to each resource distribution scheme until the remote academic tutoring session ends.

[0016] Therefore, in this application, the distribution ratio and terminal adaptation format in the hierarchical scheduling instructions are adjusted in real time according to each resource distribution scheme until the remote academic tutoring session ends. First, by determining the hierarchical scheduling instructions, a precise mapping relationship between cognitive load status and resource scheduling strategy can be obtained, thereby establishing an upper-level scheduling decision mechanism centered on students' cognitive capacity. The hierarchical scheduling instructions match the real-time cognitive load level of students output by the cognitive load identification module with the preset tutoring strategy template, quantifying the originally ambiguous cognitive state into an execution instruction with clear scheduling weight. This execution instruction carries the priority sequence of resource allocation under different load states, enabling the main tutoring terminal to plan the order and allocation structure of various tutoring resources according to the student's current mental resource occupancy. This solves the problem of resource scheduling being disconnected from the student's actual learning status in the traditional unified push method, allowing the resource scheduling strategy to adaptively adjust with the dynamic changes in cognitive load, significantly improving resource efficiency. The accuracy of matching source scheduling with learning status is assessed. Then, by determining the resource distribution scheme, the resource adaptability and fine-grained control of distribution granularity in a multi-terminal heterogeneous environment can be achieved. This ensures the effective implementation of the cognitive load-driven scheduling strategy at the actual transmission level. The resource distribution scheme clusters and groups based on the resource adaptability characteristics of each tutoring terminal, independently configuring resource encoding bitrate, resource slice duration, and resource presentation style for different groups. This allows resources under the same tutoring strategy instruction to be distributed in parallel in a way that adapts to the capabilities of each terminal, deeply coupling the cognitive load-driven scheduling strategy with the actual capabilities of the terminal. This transforms abstract hierarchical scheduling instructions into executable, adaptable, and implementable specific resource distribution operations, improving the efficiency of resource transmission and the adaptability of resource presentation in multi-terminal collaborative scenarios, and realizing end-to-end differentiated scheduling from strategy to execution. In summary, based on the above scheme, cognitive load-driven differentiated resource scheduling in remote academic tutoring can be realized, thereby improving the adaptation efficiency of multi-terminal collaborative tutoring. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the 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.

[0018] Figure 1 This is an exemplary flowchart of a remote academic tutoring resource scheduling method with multi-terminal collaboration, as shown in some embodiments of this application. Figure 2 This is a terminal structure diagram of remote academic tutoring according to some embodiments of this application; Figure 3This is a communication structure diagram of a remote academic tutoring platform according to some embodiments of this application; Figure 4 This is a flowchart illustrating the process of determining a resource distribution scheme according to some embodiments of this application; Figure 5 This is a schematic diagram of the structure of a resource distribution unit according to some embodiments of this application; Figure 6 This is a schematic diagram of the structure of a computer device for implementing a remote academic tutoring resource scheduling method with multi-terminal collaboration, according to some embodiments of this application. Detailed Implementation

[0019] To better understand the technical solution of this application, the technical solution of this application will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0020] refer to Figure 1 The figure is an exemplary flowchart of a multi-terminal collaborative remote academic tutoring resource scheduling method according to some embodiments of this application. The multi-terminal collaborative remote academic tutoring resource scheduling method mainly includes the following steps: In step 101, during the remote academic tutoring process, the tutoring session request initiated by the main tutoring terminal is obtained, and the interactive feedback data of each auxiliary tutoring terminal is collected periodically.

[0021] It should be noted that in this application, the main tutoring terminal is a teaching-leading terminal based on the teacher's identity; the auxiliary tutoring terminal is a collaborative participation terminal based on the student's identity and auxiliary teaching identity; the tutoring session request is a structured trigger signal used to initiate the remote academic tutoring process and carry multi-dimensional interactive data and tutoring scenario information from the student's end; the interactive feedback data is periodically collected data reflecting the resource reception status and user operation response of each auxiliary tutoring terminal during the tutoring process.

[0022] In practice, before remote academic tutoring begins, the teacher initiates a tutoring session request through the main tutoring terminal. This tutoring session request carries student-side interaction data related to the tutoring session and the current tutoring scenario information. During the tutoring session, interactive feedback data is continuously collected from each participating tutoring terminal according to a preset time period. This interactive feedback data reflects the real-time status of each terminal when receiving and presenting tutoring resources. By using the tutoring session request as the initial trigger condition, the interactive feedback data of each tutoring terminal can be obtained.

[0023] It should be noted that in this application, Figure 2This is a terminal structure diagram for remote academic tutoring. The diagram presents a three-tiered tree architecture. The core top layer is the overall remote academic tutoring platform, which is divided into two major branch terminals: the main tutoring terminal, which undertakes the core teaching and management functions, and the auxiliary tutoring terminal, which is responsible for supporting and collaborative work. The main tutoring terminal is further subdivided into three functional areas: video teaching area, interactive management area, and resource scheduling area. The auxiliary tutoring terminal is correspondingly divided into three functional areas: Q&A assistance area, material organization area, and learning assistance area.

[0024] It should be noted that in this application, Figure 3 This is a communication structure diagram of a remote academic tutoring platform. The diagram uses a business communication server as the core relay node. Both the main tutoring terminal and the auxiliary tutoring terminal establish bidirectional connections with the server. The main tutoring terminal transmits core information such as real-time teaching instructions, audio and video, and shared courseware to the auxiliary tutoring terminal through the server. The auxiliary tutoring terminal, in turn, uses the server to transmit learning feedback, Q&A content, and supplementary materials back to the main tutoring terminal. At the same time, the main and auxiliary terminals also support point-to-point direct communication to achieve rapid information exchange and teaching collaboration. The overall structure forms a stable communication logic with the server relay as the main mechanism and point-to-point direct connection as a supplement, ensuring real-time and efficient interaction of teaching data between the two ends.

[0025] In step 102, the cognitive load of the current student is identified based on the interaction data and tutoring scenario information in the tutoring session request, and the cognitive load level of the current student is obtained. The cognitive load level is then matched with the tutoring strategy template of remote academic tutoring to obtain the hierarchical scheduling instruction of the main tutoring terminal.

[0026] In some embodiments, identifying the current student's cognitive load level based on the interaction data and tutoring scenario information in the tutoring session request can be achieved through the following steps: Multimodal interaction data and tutoring scenario information are extracted from the tutoring session request. The interaction data includes the student's speech spectrum features, writing trajectory curvature, and screen gaze point heatmap. Based on the question difficulty coefficient and knowledge point association graph in the interaction data and tutoring scenario information, the cognitive load of the current student is evaluated in multiple dimensions to obtain the probability distribution vector of multiple cognitive load dimensions. All probability distribution vectors are fused across dimensions, and the fusion results are then mapped to a preset cognitive load level range to obtain the current student's cognitive load level.

[0027] It should be noted that, in this application, multimodal interaction data is a structured set of information reflecting students' multidimensional behavioral performance during remote academic tutoring; tutoring scenario information is an auxiliary judgment basis describing the objective difficulty and knowledge-related background of the current remote academic tutoring task; speech spectrum features are quantitative parameters characterizing the distribution pattern of students' speech energy in different frequency ranges during tutoring; writing trajectory curvature is a geometric feature parameter measuring the degree of curvature and trend of students' handwriting during tutoring; screen gaze point heat map is visual data used to present the duration and concentration of students' gaze in different areas of the screen during tutoring; question difficulty coefficient is a numerical indicator quantifying the complexity and cognitive challenge level of the current tutoring question; knowledge point association graph is a network structure model describing the connection relationship between the current tutoring knowledge point and other related knowledge nodes; probability distribution vector represents the data structure of the confidence distribution of students belonging to each possible state on a single cognitive load evaluation dimension; cognitive load level is a discrete evaluation result used to classify the current mental resource occupancy of students during remote academic tutoring.

[0028] In practical implementation, firstly, when a remote academic tutoring session begins, two types of core information are parsed and separated from the tutoring session request initiated by the main tutoring terminal. The first type is multimodal interaction data, and the second type is tutoring scenario information. For the multimodal interaction data, three specific dimension feature parameters are further extracted from the interaction data: the first dimension is speech spectrum features, which analyze the speech signals emitted by students during tutoring, divides the speech into multiple frequency bands according to frequency, and statistically analyzes the proportion of speech energy in each frequency band, thereby reflecting the student's level of tension and emotional fluctuations; the second dimension is the curvature of the writing trajectory, which... The features include: collecting students' handwriting trajectories on electronic writing devices, calculating the curvature of the trajectory curve in each local segment, and statistically analyzing the curvature change trend over time, thereby reflecting the degree of hesitation and fluency of students' writing; the third dimension is the screen gaze point heatmap, which tracks the students' eye gaze positions on the screen, records the gaze duration of each position area, and presents the concentrated and dispersed areas of gaze points in the form of a heatmap; the set of speech spectrum features, writing trajectory curvature, and screen gaze point heatmap is used as multimodal interactive data, and the set of question difficulty coefficients and knowledge point association maps is used as tutoring scenario information; Then, the multimodal interaction data and tutoring scenario information are input into a pre-built cognitive load assessment model. This model assesses students' cognitive load from multiple dimensions, each corresponding to a different perspective of cognitive load representation. The specific assessment process is as follows: In the first dimension, the model analyzes the change in the ratio of high-frequency energy to low-frequency energy in the speech spectrum features, and, combined with the difficulty level of the questions, determines the cognitive load tendency corresponding to the student's speech tension level. It outputs the probability values ​​of the student being in a low-load, medium-load, or high-load state under this dimension. The set of the three values ​​serves as the probability distribution vector for the first dimension; in the second dimension... In two dimensions, the model analyzes the fluctuation amplitude of the curvature of the writing trajectory and the frequency of abnormal peaks. Combined with the dependency depth between the current knowledge point and the previous knowledge point in the knowledge point association graph, it judges the cognitive load tendency corresponding to the degree of hesitation in writing and outputs the probability values ​​of the three load states under this dimension as the probability distribution vector of the second dimension. In the third dimension, the model analyzes the concentration of gaze points and the frequency of jumps between regions in the screen gaze point heat map to judge the cognitive load tendency corresponding to the degree of visual attention dispersion of students. It outputs the probability values ​​of the three load states under this dimension as the probability distribution vector of the third dimension.Through the above multi-dimensional parallel evaluation, the probability distribution vectors output by each dimension are used as the preliminary results of the multi-dimensional cognitive load evaluation. Finally, the probability distribution vectors of multiple cognitive load dimensions are fused across dimensions. The fusion calculation can adopt a weighted comprehensive approach, with a pre-set weight coefficient for each cognitive load dimension. This weight coefficient reflects the importance of that dimension in the comprehensive judgment of cognitive load. Fusion is performed separately for low load, medium load, and high load states. For the low load state, the probability values ​​corresponding to low load in each dimension's probability distribution vector are multiplied by their respective dimension's weight coefficient and then summed to obtain the comprehensive confidence level for the low load state. For the medium load state, a weighted comprehensive approach is adopted. The overall confidence level for the medium load state is calculated in the same way; the overall confidence level for the high load state is also calculated in the same way. After the fusion calculation is completed, three overall confidence level values ​​are obtained, corresponding to the low load, medium load, and high load states, respectively. A cognitive load level interval division rule is preset. This rule specifies the range of overall confidence level values ​​that correspond to the low load level, the medium load level, and the high load level. The three overall confidence level values ​​are compared with the level interval division rule, and the load state with the highest overall confidence level that falls into the corresponding interval is selected as the final result. This result is used as the current student's cognitive load level.

[0029] It should be noted that in this application, the difficulty level of the questions can be obtained through three comprehensive methods: First, based on statistical analysis of historical answer data, a large number of students' correct answers and average time taken on the same question are collected; the lower the correct answer rate and the longer the time taken, the higher the difficulty level. Second, based on the pre-set value of the cognitive level of knowledge points, according to Bloom's Taxonomy of Objectives, the cognitive level tested by the questions is divided into six levels: memory, comprehension, application, analysis, evaluation, and creation, with different levels corresponding to different basic difficulty ranges. Third, based on the quantitative assessment of the question's structural characteristics, the number of conditions and the length of the reasoning steps in the question are analyzed. Structural elements such as the complexity of distractors are weighted and calculated to obtain a structural difficulty value, which serves as a standardized difficulty coefficient for each question. The knowledge point association graph is created manually by subject matter experts during the curriculum system construction phase, based on the logical sequence, knowledge inclusion, and cross-chapter associations between knowledge points. Simultaneously, by analyzing a large amount of students' historical learning path data, implicit co-occurrence associations and pre- and post-dependent relationships between knowledge points are discovered. The expert graph is then dynamically supplemented and its weights are adjusted to obtain a knowledge point association graph covering the entire subject system.

[0030] In some embodiments, the following steps can be used to perform tag matching between the cognitive load level and the remote academic tutoring strategy template to obtain the hierarchical scheduling instructions for the main tutoring terminal: A pre-built tutoring strategy template library is constructed, and each tutoring strategy template in the library is associated with a cognitive load level range, an academic label type, and a corresponding resource scheduling priority sequence. The current student's cognitive load level and academic label are jointly encoded to obtain the matching degree with each tutoring strategy template in the tutoring strategy template library, and the target template with the highest matching degree is selected. The predefined resource scheduling priority sequence in the target template is parsed to generate resource allocation weights for different auxiliary terminal categories, thereby obtaining hierarchical scheduling instructions for the primary auxiliary terminal.

[0031] It should be noted that in this application, the tutoring strategy template library is a structured database used to store and manage various preset tutoring strategy templates; the tutoring strategy template is a standard configuration unit used to define resource scheduling rules under a specific cognitive load state and academic tag combination; the cognitive load level range is a set of segmented boundaries used to divide the range of cognitive load level values; the academic tag type is a classification identifier used to identify the subject area and knowledge point category to which the current tutoring content belongs; the resource scheduling priority sequence is an ordered list used to specify the order and weight ratio of different types of tutoring terminals when receiving resources; the matching degree is a quantitative indicator that measures the degree of fit between the current tutoring state and each tutoring strategy template; the target template is the selected tutoring strategy template that serves as the basis for resource scheduling in the current tutoring scenario; the resource allocation weight is a numerical parameter that specifies the proportion of each type of tutoring terminal in resource distribution; and the hierarchical scheduling instruction is an execution command used to guide the main tutoring terminal to allocate resources to each type of tutoring terminal according to differentiated weights.

[0032] In practical implementation, firstly, during the deployment phase of the remote academic tutoring system, a tutoring strategy template library is pre-built. This library stores multiple tutoring strategy templates, each associated with three parts of information: The first part is the cognitive load level range, which specifies when a student's cognitive load level falls within a certain range. For example, a low cognitive load level corresponds to one range, a medium load to another, and a high load to yet another. Each range can cover a single level or multiple consecutive levels. The second part is the academic tag type, used to identify the subject attribute and knowledge point affiliation of the tutoring content. For example, functions in mathematics, mechanics in physics, etc. Each template is associated with one or more academic tag types. The third part is the resource scheduling priority sequence, which, in the form of an ordered list, specifies the order in which different types of tutoring terminals receive resources and the weight ratio of each type of terminal in the total resources when applying the template. This method serves as the benchmark for subsequent matching decisions. Then, after identifying the current student's cognitive load level, the remote academic tutoring session is obtained. The academic tags are pre-set when the tutoring session is initiated. The cognitive load level and the academic tags are jointly encoded. Specifically, the numerical representation of the cognitive load level and the category identifier of the academic tag are concatenated in a fixed order to form a unified encoding string. This encoding string contains both load status information and subject content information. The jointly encoded string is then matched against the cognitive load level range and academic tag type associated with each tutoring strategy template in the tutoring strategy template library. The matching process is as follows: determining whether the cognitive load level in the encoding string falls within the template's cognitive load range... Within the known load level range, if the content falls within the range, the first matching score is obtained; otherwise, the first matching score is zero. Next, it is determined whether the academic label in the encoded string matches the academic label type of the template. If they match, the second matching score is obtained; otherwise, the second matching score is zero. The first and second matching scores are added together to obtain the overall matching degree of the template. By performing the above matching degree calculation on all templates in the tutoring strategy template library, the matching degree value corresponding to each template can be obtained. The template with the highest matching degree value among all matching degree values ​​is selected as the target template.Finally, after obtaining the selected target template, a predefined resource scheduling priority sequence is parsed from it. The resource scheduling priority sequence is an ordered list, in which different types of auxiliary and tutoring terminals are arranged in descending order of priority. For example, the first priority in the list is the smart large screen terminal, the second priority is the tablet interactive terminal, and the third priority is the mobile communication terminal. Based on this priority sequence, the resource allocation weights of each type of auxiliary and tutoring terminal are generated according to the following rules: the terminal type with the highest priority receives the largest resource allocation weight, the terminal type with the second highest priority receives the second largest resource allocation weight, and so on. The sum of the resource allocation weights of all terminal types is 100%. For example, if the smart large screen terminal is ranked first, the tablet interactive terminal is ranked second, and the mobile communication terminal is ranked last in the priority sequence, then the system can set the resource allocation weights to 50% for the smart large screen terminal, 30% for the tablet interactive terminal, and 20% for the mobile communication terminal. The system encapsulates the parsed resource scheduling priority sequence and the generated resource allocation weights into a hierarchical scheduling instruction. This instruction explicitly specifies the priority order and weight ratio by which the primary tutoring terminal should allocate tutoring resources to various tutoring terminals during subsequent resource distribution. The hierarchical scheduling instruction serves as the basis for the primary tutoring terminal to execute resource scheduling actions.

[0033] In step 103, the remote academic tutoring resources among the various tutoring terminals are distributed in a differentiated manner based on the resource adaptation features in the various interactive feedback data, thereby obtaining the resource distribution scheme for each tutoring terminal.

[0034] In some embodiments, the remote academic tutoring resources among various tutoring terminals are distributed in a differentiated manner based on the resource adaptation characteristics in the various interactive feedback data, resulting in a resource distribution scheme for each tutoring terminal, as referenced. Figure 4 The diagram is a flowchart illustrating the process of determining a resource distribution scheme in some embodiments of this application. In this embodiment, determining the resource distribution scheme can be achieved through the following steps: In step 1031, the resource adaptation features of each tutoring terminal are extracted from the interactive feedback data; In step 1032, all auxiliary terminals are clustered and grouped according to the resource adaptation characteristics. Terminals in the same group adopt the same resource encapsulation format, and different groups adopt different resource transmission strategies. In step 1033, a corresponding resource encoding bitrate, resource slice duration, and resource presentation style are generated for each group, thereby obtaining the resource distribution scheme for each auxiliary terminal.

[0035] It should be noted that, in this application, the resource adaptation feature is a set of multi-dimensional parameters describing the capability attributes and real-time status of each tutoring terminal when receiving and presenting tutoring resources; the resource encapsulation format is a technical specification that defines how tutoring resources are packaged and encoded before transmission; the resource transmission strategy is used to define the bandwidth allocation and transmission delay control scheme adopted when tutoring resources are sent to terminals in the network; the resource encoding rate is a compression parameter that controls the amount of data in tutoring resources per unit time; the resource slice duration is a time interval parameter that divides continuous tutoring resources into independent transmission units; the resource presentation style is a configuration parameter that defines the layout and interaction mode of tutoring resources when displayed on the terminal screen; and the resource distribution scheme is a complete set of configurations that guides the system to send tutoring resources to each tutoring terminal in a differentiated manner.

[0036] In practice, firstly, during remote tutoring, the system continuously receives interactive feedback data from various tutoring terminals. This data includes real-time information on the terminal's operational status and user actions. The system then parses and extracts the resource adaptation characteristics of each tutoring terminal from this data. These characteristics include four dimensions: the first is the terminal screen resolution, reflecting the pixel width and height of the terminal's display, obtained from the device attribute field in the interactive feedback data; the second is network bandwidth utilization, reflecting the percentage of total bandwidth used in the current network connection of the tutoring terminal. The first dimension is the bandwidth ratio, which is obtained from the network status field in the interactive feedback data. The second dimension is the terminal response latency, which reflects the time difference between when the system sends resources and when the tutoring terminal completes reception and begins presentation; this parameter is calculated from the timestamp field in the interactive feedback data. The third dimension is the user role type of the terminal, which identifies the user operating the tutoring terminal in remote academic tutoring, such as student or teaching assistant; this parameter is obtained from the user identifier field in the interactive feedback data. These four dimensions are used as independent resource adaptation features for the corresponding tutoring terminal. The above method can then be used to obtain... The process involves first identifying the resource adaptation characteristics of each tutoring terminal; then, based on these characteristics, clustering all tutoring terminals. The specific clustering process is as follows: Each tutoring terminal's resource adaptation characteristic is considered a point in a multi-dimensional space, with the four dimensions corresponding to different coordinate axes. The distance between any two tutoring terminals is calculated; closer distances indicate more similar resource adaptation characteristics, while greater distances indicate greater differences. The terminals are then divided into several groups based on their distance, ensuring that the distance between any two terminals within the same group is less than a preset similarity threshold. A threshold is set, and the distance between different groups is greater than this threshold. After grouping, a resource encapsulation format is determined for each group. All auxiliary terminals within the same group use the same resource encapsulation format, which includes video encoding standards, audio encoding standards, and encapsulation container types. At the same time, a resource transmission strategy is determined for each group, and different groups use different resource transmission strategies. The resource transmission strategy includes transmission priority settings, bandwidth reservation ratios, and retransmission mechanism configurations. For example, terminal groups with high screen resolution and sufficient bandwidth can use a high bitrate transmission strategy, while terminal groups with low screen resolution and limited bandwidth can use a low bitrate transmission strategy.The system uses the clustering results, the resource encapsulation format of each group, and the resource transmission strategy of each group as the initial configuration for differentiated distribution. Finally, after completing the clustering and determining the resource encapsulation format and transmission strategy for each group, three specific resource distribution parameters are generated for each group. The first parameter is the resource encoding bitrate, which is determined based on the average screen resolution and average network bandwidth utilization of the terminals within the group. Specifically, a high bitrate is selected when the screen resolution is high and bandwidth is sufficient, and a low bitrate is selected when the screen resolution is low or bandwidth is limited, ensuring that resources are presented clearly on the terminal without causing network congestion. The second parameter is the resource slice duration, which is determined based on the average response latency of the terminals within the group. Specifically, a low bitrate is selected when the response latency is low. Shorter slice durations are selected for more precise resource scheduling; longer slice durations are chosen when response latency is high to reduce the number of interactions during transmission; the third parameter is the resource presentation style, which is configured according to the user role type of the terminal within the group. Specifically, for terminals with a student role, the resource presentation style is configured as a complete learning interface including practice questions and interactive components; for terminals with a teaching support role, the resource presentation style is configured as a teaching support interface including student progress monitoring and auxiliary tools. The resource encoding bitrate, resource slice duration, and resource presentation style of each group are integrated to form the resource distribution sub-scheme corresponding to that group. The set of resource distribution sub-schemes for all groups is taken as the resource distribution scheme for the corresponding tutoring terminal. The resource distribution scheme for each tutoring terminal can be obtained in the above way.

[0037] In step 104, the distribution ratio and terminal adaptation format in the hierarchical scheduling instruction are adjusted in real time according to each resource distribution scheme until the remote academic tutoring session ends.

[0038] In some embodiments, adjusting the distribution ratio and terminal adaptation format in the hierarchical scheduling instruction in real time according to each resource distribution scheme until the remote academic tutoring session ends can be achieved through the following steps: Extract the resource reception feedback parameters of each tutoring terminal from the resource distribution scheme of each tutoring terminal; When the resource reception feedback parameters of any auxiliary tutoring terminal trigger the preset adjustment threshold, the optimal distribution ratio and terminal adaptation format are recalculated based on the resource adaptation characteristics of the corresponding auxiliary tutoring terminal. The adjusted distribution ratio and terminal adaptation format are synchronized to the hierarchical scheduling instruction, the resource scheduling parameters of the corresponding terminal are updated, and the feedback data of the next cycle is monitored until the session ends.

[0039] It should be noted that in this application, the resource reception feedback parameter is a multi-dimensional measurement indicator that quantifies the actual performance of each auxiliary tutoring terminal in receiving and presenting tutoring resources; the adjustment threshold is a critical value condition for determining whether the current resource distribution status needs to trigger dynamic adjustment; the optimal distribution ratio is the optimal numerical configuration of the allocation share of each type of resource under a specific terminal status; the terminal adaptation format is a set of resource encoding parameters and presentation specifications that match the specific terminal capabilities and status; and the resource scheduling parameter is a set of specific configuration items used to control the allocation of resources from the main tutoring terminal to each auxiliary tutoring terminal.

[0040] In practice, firstly, during remote tutoring, feedback information on the execution of the resource distribution plan is received from each tutoring terminal according to a preset time period. Resource reception feedback parameters for each tutoring terminal are extracted from this feedback information. These parameters include three dimensions of measurement indicators: the first dimension is the resource packet loss ratio, which reflects the proportion of data packets lost during resource transmission to the total number of data packets sent. This ratio is calculated by statistically analyzing the differences between the reception confirmation information reported by the tutoring terminal and the system's transmission records. The second dimension is the number of playback stutters, which reflects the frequency of playback buffering during a unit of time. The number of times video pauses or audio interruptions occurred during playback was obtained by analyzing the playback status logs reported by the auxiliary tutoring terminals. The third dimension is the user interaction response success rate, which reflects the proportion of successful system responses after user operations on the auxiliary tutoring terminals. This is calculated by comparing the user operation records and system response records reported by the auxiliary tutoring terminals. The resource packet loss ratio, playback stutters, and user interaction response success rate of each auxiliary tutoring terminal are integrated to form the resource reception feedback parameter set for that terminal. Then, after the resource reception feedback parameters of each auxiliary tutoring terminal, each parameter of each terminal is compared with a pre-set adjustment threshold. The adjustment threshold is set at 1 / 3 of the total number of video pauses or audio interruptions. Each parameter is set individually; for example, the threshold for resource packet loss ratio is set to 5%, the threshold for playback stuttering times is set to three times per minute, and the threshold for user interaction response success rate is set to 90%. When any resource reception feedback parameter of any auxiliary terminal reaches or exceeds the corresponding adjustment threshold, it is determined that the terminal needs to trigger dynamic adjustment. The resource adaptation features extracted by the terminal in the previous steps are obtained. The resource adaptation features include the terminal's screen resolution, network bandwidth utilization, terminal response latency, and the user role type of the terminal. Based on the terminal's resource adaptation features, its optimal distribution ratio and terminal adaptation format are recalculated. The optimal distribution ratio... The calculation method is as follows: The available bandwidth limit for resource transmission is preset based on the terminal's network bandwidth utilization rate; the priority order of resource distribution is set based on the terminal's response latency time; and weights are allocated to different types of resources based on these factors, ensuring that the allocation ratio of each type of resource achieves optimal transmission efficiency under the current network conditions. The calculation method for the terminal adaptation format is as follows: The encoding resolution parameters of the resources are preset based on the terminal's screen resolution; the encoding bitrate parameters of the resources are set based on the terminal's network bandwidth utilization rate; and the presentation style parameters of the resources are selected based on the user role type of the terminal. The recalculated optimal distribution ratio and terminal adaptation format are used as the adjusted configuration for this terminal.Finally, after recalculating the optimal distribution ratio and terminal adaptation format, the adjusted distribution ratio and terminal adaptation format are synchronized to the hierarchical scheduling instructions. The hierarchical scheduling instructions are the core command set for resource scheduling executed by the primary tutoring terminal, containing resource scheduling parameters for all tutoring terminals. The configuration item corresponding to the tutoring terminal that triggered the adjustment is located within the hierarchical scheduling instructions. The original distribution ratio is replaced with the recalculated optimal distribution ratio, and the original adaptation format is replaced with the recalculated terminal adaptation format. After the replacement is completed, the updated hierarchical scheduling instructions are sent to the primary tutoring terminal. The primary tutoring terminal continues to distribute tutoring resources to the tutoring terminal according to the new resource scheduling parameters. After this adjustment is completed, the process continues into the next monitoring cycle according to the preset time period, re-executing the above process, continuously collecting resource reception feedback parameters from each tutoring terminal, monitoring whether the adjustment threshold is triggered again, and dynamically adjusting again as needed. This monitoring and adjustment process is repeated until the current remote academic tutoring session ends. The last updated hierarchical scheduling instructions serve as the final resource scheduling basis for the entire tutoring process.

[0041] Furthermore, in another aspect of this application, in some embodiments, this application provides a multi-terminal collaborative remote academic tutoring resource scheduling system, which includes a resource distribution unit, as referenced. Figure 5 The figure is a schematic diagram of the structure of a resource distribution unit according to some embodiments of this application. The resource distribution unit includes: an acquisition module 201, a processing module 202, and an execution module 203, which are described below: The acquisition module 201 in this application is mainly used to acquire tutoring session requests initiated by the main tutoring terminal during remote academic tutoring, and periodically collect interactive feedback data from each tutoring terminal. Processing module 202, in this application, is used to identify the cognitive load of the current student based on the interaction data and tutoring scenario information in the tutoring session request, obtain the cognitive load level of the current student, match the cognitive load level with the tutoring strategy template of remote academic tutoring, and obtain the hierarchical scheduling instruction of the main tutoring terminal. It should be noted that the processing module 202 is also used to distribute remote academic tutoring resources between various tutoring terminals in a differentiated manner based on the resource adaptation features in each interactive feedback data, so as to obtain a resource distribution scheme for each tutoring terminal. The execution module 203 in this application is mainly used to adjust the distribution ratio and terminal adaptation format in the hierarchical scheduling instruction in real time according to each resource distribution scheme until the remote academic tutoring session ends.

[0042] The foregoing has detailed examples of a multi-terminal collaborative remote academic tutoring resource scheduling method and system provided in the embodiments of this application. It is understood that the corresponding apparatus, in order to achieve the above functions, includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should readily recognize that, in conjunction with the units and algorithm steps of the examples described in the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed by hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0043] In some embodiments, this application also provides a computer device, the computer device including a memory and a processor, the memory for storing a computer program, and the processor for calling and running the computer program from the memory, so that the computer device executes the above-described multi-terminal collaborative remote academic tutoring resource scheduling method.

[0044] In some embodiments, reference Figure 6 The dashed lines in the figure indicate that the unit or module is optional. This figure is a structural schematic diagram of a computer device for implementing a remote academic tutoring resource scheduling method with multi-terminal collaboration according to an embodiment of this application. The remote academic tutoring resource scheduling method with multi-terminal collaboration described in the above embodiments can be achieved through… Figure 6 The computer device shown is used to implement this, and the computer device includes at least one processor 301, a memory 302 and at least one communication unit 305. The computer device may be a terminal device, a server or a chip.

[0045] Processor 301 can be a general-purpose processor or a special-purpose processor. For example, processor 301 can be a central processing unit (CPU), which can be used to control computer devices, execute software programs, and process data from software programs. The computer device may also include a communication unit 305 for inputting (receiving) and outputting (transmitting) signals.

[0046] For example, the computer device may be a chip, and the communication unit 305 may be the input and / or output circuit of the chip, or the communication unit 305 may be the communication interface of the chip, which may be a component of a terminal device, network device or other device.

[0047] For example, the computer device may be a terminal device or a server, and the communication unit 305 may be a transceiver of the terminal device or the server, or the communication unit 305 may be a transceiver circuit of the terminal device or the server.

[0048] The computer device may include one or more memories 302 storing a program 304. The program 304 can be executed by a processor 301 to generate instructions 303, causing the processor 301 to execute the method described in the above method embodiments according to the instructions 303. Optionally, the memory 302 may also store data (such as a target audit model). Optionally, the processor 301 may also read data stored in the memory 302, which may be stored at the same storage address as the program 304, or it may be stored at a different storage address than the program 304.

[0049] The processor 301 and memory 302 can be configured separately or integrated together, for example, integrated on the system on chip (SOC) of the terminal device.

[0050] It should be understood that each step of the above method embodiment can be completed by hardware logic circuits or software instructions in the processor 301. The processor 301 can be a CPU, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, such as discrete gate, transistor logic devices, or discrete hardware components.

[0051] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0052] For example, in some embodiments, this application also provides a computer-readable storage medium storing instructions or code that, when executed on a computer, cause the computer to implement the above-described multi-terminal collaborative remote academic tutoring resource scheduling method.

[0053] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0054] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for scheduling remote academic tutoring resources through multi-terminal collaboration, characterized in that, Includes the following steps: During remote academic tutoring, the tutoring session requests initiated by the main tutoring terminal are obtained, and the interactive feedback data of each tutoring terminal is collected periodically. Based on the interactive data and tutoring scenario information in the tutoring session request, the cognitive load of the current student is identified to obtain the cognitive load level of the current student. The cognitive load level is then matched with the tutoring strategy template of remote academic tutoring to obtain the hierarchical scheduling instruction of the main tutoring terminal. By using the resource adaptation characteristics in the various interactive feedback data, the remote academic tutoring resources among the various tutoring terminals are distributed in a differentiated manner to obtain the resource distribution scheme for each tutoring terminal. The distribution ratio and terminal adaptation format in the hierarchical scheduling instructions are adjusted in real time according to each resource distribution scheme until the remote academic tutoring session ends.

2. The method as described in claim 1, characterized in that, Based on the interaction data and tutoring scenario information in the tutoring session request, the cognitive load of the current student is identified, and the cognitive load level of the current student is obtained, specifically including: Multimodal interaction data and tutoring scenario information are extracted from the tutoring session request. The interaction data includes the student's speech spectrum features, writing trajectory curvature, and screen gaze point heatmap. Based on the question difficulty coefficient and knowledge point association graph in the interaction data and tutoring scenario information, the cognitive load of the current student is evaluated in multiple dimensions to obtain the probability distribution vector of multiple cognitive load dimensions. All probability distribution vectors are fused across dimensions, and the fusion results are then mapped to a preset cognitive load level range to obtain the current student's cognitive load level.

3. The method as described in claim 1, characterized in that, The cognitive load level and the remote academic tutoring strategy template are tagged and matched to obtain the hierarchical scheduling instructions for the main tutoring terminal, which specifically include: A pre-built tutoring strategy template library is constructed, and each tutoring strategy template in the library is associated with a cognitive load level range, an academic label type, and a corresponding resource scheduling priority sequence. The current student's cognitive load level and academic label are jointly encoded to obtain the matching degree with each tutoring strategy template in the tutoring strategy template library, and the target template with the highest matching degree is selected. The predefined resource scheduling priority sequence in the target template is parsed to generate resource allocation weights for different auxiliary terminal categories, thereby obtaining hierarchical scheduling instructions for the primary auxiliary terminal.

4. The method as described in claim 1, characterized in that, By leveraging the resource adaptation characteristics in the feedback data from various interactions, differentiated resource distribution is implemented across different tutoring terminals for remote academic tutoring. The specific resource distribution schemes for each tutoring terminal include: Extract the resource adaptation features of each tutoring terminal from the interactive feedback data; Based on the resource adaptation characteristics, all auxiliary terminals are clustered and grouped. Terminals within the same group use the same resource encapsulation format, while different groups use different resource transmission strategies. For each group, a corresponding resource encoding bitrate, resource slice duration, and resource presentation style are generated, thereby obtaining the resource distribution scheme for each auxiliary and tutoring terminal.

5. The method as described in claim 1, characterized in that, The distribution ratio and terminal adaptation format in the hierarchical scheduling instructions are adjusted in real time according to each resource distribution scheme until the remote academic tutoring session ends. Specifically, this includes: Extract the resource reception feedback parameters of each tutoring terminal from the resource distribution scheme of each tutoring terminal; When the resource reception feedback parameters of any auxiliary tutoring terminal trigger the preset adjustment threshold, the optimal distribution ratio and terminal adaptation format are recalculated based on the resource adaptation characteristics of the corresponding auxiliary tutoring terminal. The adjusted distribution ratio and terminal adaptation format are synchronized to the hierarchical scheduling instruction, the resource scheduling parameters of the corresponding terminal are updated, and the feedback data of the next cycle is monitored until the session ends.

6. The method as described in claim 1, characterized in that, The main tutoring terminal is a teaching-leading terminal based on the teacher's identity.

7. The method as described in claim 1, characterized in that, The tutoring terminal is a collaborative participation terminal based on student identity and tutoring identity.

8. A multi-terminal collaborative remote academic tutoring resource scheduling system, the multi-terminal collaborative remote academic tutoring resource scheduling system comprising a resource distribution unit, characterized in that, The resource distribution unit includes: The acquisition module is used to acquire tutoring session requests initiated by the main tutoring terminal during remote academic tutoring, and periodically collect interactive feedback data from each tutoring terminal. The processing module is used to identify the cognitive load of the current student based on the interaction data and tutoring scenario information in the tutoring session request, obtain the cognitive load level of the current student, match the cognitive load level with the tutoring strategy template of remote academic tutoring, and obtain the hierarchical scheduling instruction of the main tutoring terminal. The processing module is also used to distribute remote academic tutoring resources between various tutoring terminals in a differentiated manner based on the resource adaptation features in each interactive feedback data, so as to obtain a resource distribution scheme for each tutoring terminal. The execution module is used to adjust the distribution ratio and terminal adaptation format in the hierarchical scheduling instructions in real time according to each resource distribution scheme until the remote academic tutoring session ends.

9. A computer device, characterized in that, The computer device includes a memory and a processor. The memory is used to store computer programs, and the processor is used to call and run the computer programs from the memory, so that the computer device performs the remote academic tutoring resource scheduling method for multi-terminal collaboration as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions or code that, when executed on a computer, cause the computer to implement the remote academic tutoring resource scheduling method for multi-terminal collaboration as described in any one of claims 1 to 7.

Images