Metaverse learning behavior guidance methods, devices, electronic devices and storage media

Abstract

This invention provides a method, device, electronic device, and storage medium for guiding learning behavior in a metaverse environment. By acquiring embodied operation signals within a sliding window and extracting embodied behavior features, it determines continuous operation states. When continuous abnormal operation states are identified, the intensity of guidance intervention is dynamically calculated based on the cumulative duration. Finally, a corresponding guidance strategy is matched to guide learning behavior. The method, device, electronic device, and storage medium provided by this invention achieve a proactive and seamless immersive behavior guidance mechanism through dynamic evaluation and adaptive intervention of the user's continuous operation states. This overcomes the shortcomings of traditional result-oriented or node-triggered methods, such as intervention lag and interaction breaks. It achieves smooth, accurate, and timely correction before errors occur, thereby significantly improving the efficiency of skill training and the user's immersive learning experience in a metaverse environment.

Description

Technical Field

[0001] This invention relates to the field of human-computer interaction technology, and in particular to a metaverse learning behavior guidance method, device, electronic device, and storage medium. Background Technology

[0002] In application scenarios such as skills training, equipment operation instruction, or process drills, systems typically need to analyze users' operations in a virtual environment and provide corresponding prompts and assistance to improve learners' training efficiency and ensure an immersive interactive experience. To meet these guidance needs, existing metaverse or virtual training systems usually employ rule-triggered mechanisms based on process nodes or intelligent interaction mechanisms based on behavioral action recognition. Specifically, they typically provide standard prompts at preset teaching task steps or operation nodes, or use multimodal models to perform discrete category recognition of users' instantaneous actions and operation commands. The corresponding interactive content is then adjusted or guidance strategies are issued only when the user's operation is determined to belong to a specific error category, or when a training process is completed based on comprehensive score feedback.

[0003] However, existing metaverse learning guidance technologies are mostly based on process nodes or behavior recognition results to trigger feedback, thus generally suffering from problems such as delayed prompts and insufficient interaction continuity. Summary of the Invention

[0004] This invention provides a metaverse learning behavior guidance method, device, electronic device, and storage medium to solve the problem that existing metaverse learning guidance technologies are mostly based on process nodes or behavior recognition results to trigger feedback, thus generally suffering from delayed prompts and insufficient interaction continuity.

[0005] This invention provides a method for guiding metaverse learning behavior, comprising: Obtain the user's specific action signals within the behavior slider window; Feature extraction is performed on the embodied operation signal within the behavior sliding window to obtain embodied behavior features; Based on the specific behavioral characteristics, the user's operational state under the behavioral sliding window is determined; When multiple consecutive sliding windows of the behavior correspond to the same abnormal operation state, the intensity of the guidance intervention is calculated based on the cumulative duration of the abnormal operation state. Based on the guidance strategy corresponding to the intensity of the guidance intervention, the user's learning behavior is guided.

[0006] According to the present invention, a metaverse learning behavior guidance method is provided, wherein the embodied operation signal includes target pointing data, operation rhythm data, and hand movement trajectory data; The step of extracting features from the embodied operation signals within the behavior sliding window to obtain embodied behavior features includes: Based on the target pointing data, the operation target is determined, and the dwell time of the user near the operation target before the key operation is obtained in the current behavior sliding window, as well as the reference dwell time of similar historical operations. Based on the dwell time and the reference dwell time, the operation hesitation characteristics are determined. Based on the operation rhythm data, the total number of independent operation actions and the number of repetitions of the same operation within the current behavior sliding window are obtained. Based on the number of repetitions and the total number of independent operation actions, the repeated trial feature is determined. Based on the hand movement trajectory data, the distance parameter between the sampling point of the user's hand movement trajectory within the current behavior sliding window and the center point of the desired operation area is obtained, and the path deviation feature is determined based on the distance parameter. Based on the operation rhythm data, the time interval parameter between adjacent operations within the current behavior sliding window is extracted, and rhythm anomaly characteristics are determined based on the dispersion of the time interval parameter. The embodied behavior feature is obtained based on at least one of the operation hesitation feature, the repeated attempt feature, the path deviation feature, and the rhythm abnormality feature.

[0007] According to the present invention, a metaverse learning behavior guidance method is provided, wherein determining the user's operation state under the behavior sliding window based on the embodied behavior characteristics includes: Obtain the preset feature threshold corresponding to the embodied behavioral features; The embodied behavioral features are compared with the corresponding preset feature thresholds to determine any abnormal deviations in the embodied behavioral features. Based on the abnormal deviation and the preset state determination rules, the user's operation state under the behavior sliding window is determined.

[0008] According to a metaverse learning behavior guidance method provided by the present invention, the step of obtaining a preset feature threshold corresponding to the embodied behavior feature includes: Obtain the current average feature index of the user for the embodied behavioral feature within the historical operation period, as well as the group average reference feature and initial feature baseline corresponding to the embodied behavioral feature; Based on the current average feature index of the embodied behavioral characteristics and the group average reference features, the initial feature baseline is adjusted to obtain the preset feature threshold corresponding to the embodied behavioral characteristics.

[0009] According to the present invention, a method for guiding metaverse learning behavior, wherein the intensity of guidance intervention is calculated based on the cumulative duration of the abnormal operating state, includes: Obtain the maximum duration threshold for the abnormal operation state; Calculate the ratio of the cumulative duration to the maximum duration threshold; The intensity of the guidance intervention is determined based on the smaller of the ratio and the preset intensity upper limit.

[0010] According to a metaverse learning behavior guidance method provided by the present invention, the step of guiding the user's learning behavior based on a guidance strategy corresponding to the intensity of the guidance intervention includes: When the intensity of the guidance intervention is within a first preset range, the target area for the expected interaction is determined, and spatial guidance information is applied to the target area. When the intensity of the guidance intervention is within the second preset range, a virtual demonstration trajectory that is continuously connected to the user's current body position is superimposed in the user's current operating space; When the intensity of the guidance intervention is within the third preset range, the user's current operation direction is obtained, and damping or reduction increments are applied to the input components in the current operation direction that deviate from the expected direction. The intensity values ​​of the first preset interval, the second preset interval, and the third preset interval increase sequentially.

[0011] The present invention also provides a metaverse learning behavior guidance device, comprising: The acquisition unit acquires the user's specific operation signals within the behavior sliding window; The feature extraction unit extracts features from the embodied operation signal within the behavior sliding window to obtain embodied behavior features; The state determination unit determines the user's operation state under the behavior sliding window based on the specific behavioral characteristics. The intervention intensity calculation unit calculates the guidance intervention intensity based on the cumulative duration of the abnormal operation state when the operation states corresponding to multiple consecutive behavior sliding windows are all the same abnormal operation state. The guidance unit guides the user's learning behavior based on a guidance strategy corresponding to the intensity of the guidance intervention.

[0012] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the metaverse learning behavior guidance method as described above.

[0013] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the metaverse learning behavior guidance method as described above.

[0014] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the metaverse learning behavior guidance method as described above.

[0015] The present invention provides a metaverse learning behavior guidance method, device, electronic device, and storage medium. This method acquires embodied operation signals within a behavior sliding window and extracts embodied behavior features to determine continuous operation states. When continuous abnormal operation states are identified, the guidance intervention intensity is dynamically calculated based on the cumulative duration. Finally, a corresponding guidance strategy is matched to guide learning behavior. This method achieves a proactive and seamless immersive behavior guidance mechanism through dynamic evaluation and adaptive intervention of the user's continuous operation states. This overcomes the shortcomings of traditional result-oriented or node-triggered methods, such as intervention lag and interaction breaks. It achieves smooth, accurate, and timely correction before errors occur, thus significantly improving skill training efficiency and the user's immersive learning experience in the metaverse environment. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in this invention 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 some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating the metaverse learning behavior guidance method provided by the present invention; Figure 2 This is one of the structural schematic diagrams of the metaverse learning behavior guidance device provided by the present invention; Figure 3 This is the second structural schematic diagram of the metaverse learning behavior guidance device provided by the present invention; Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0019] It should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of embodiments of the present invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0020] It should be noted that all actions involving the acquisition of signals, information, or data in this application are carried out in compliance with the relevant data protection laws and regulations of the locality and with permission from the owner of the relevant device.

[0021] It should also be noted that existing interactive guidance mechanisms are essentially designed for discrete behavioral outcomes or final evaluation metrics, with feedback triggering heavily reliant on fixed process rules or final error category determination. When users gradually exhibit trends of loss of control or cognitive uncertainty during continuous interactive operations, the system cannot effectively capture the abnormal development of the operation process because these evolutionary processes have not yet formed clear error classification results or touched the set process nodes. This inevitably leads to a significant lag in system intervention, often resulting in post-event correction only after misoperations have accumulated or erroneous actions have become firmly established. Furthermore, the feedback output of this discrete outcome-triggered intervention mechanism is often abrupt and coarse-grained, unable to adapt to the gradual evolution of the operational state. In immersive virtual training environments, it easily and frequently interrupts the user's original interactive continuity, making it difficult to achieve smooth adjustment of the learning process.

[0022] To address the aforementioned problems, this invention provides a metaverse learning behavior guidance method. This method can be applied to skill training, equipment operation instruction, or process rehearsals in virtual simulation or metaverse training environments. Through continuous behavioral state determination, it achieves timely and smooth guidance for the learner's operational process. The method will be described in detail below with reference to specific embodiments. Figure 1 This is a flowchart illustrating the metaverse learning behavior guidance method provided by the present invention, as shown below. Figure 1 As shown, the method includes: Step 110: Obtain the user's specific operation signal within the behavior sliding window.

[0023] Here, a behavior sliding window refers to a data capture interval of a certain length set in the time dimension, which slides forward continuously over time to continuously record and capture the user's operation process within a specific time segment. For example, the window length can be set to tens to hundreds of frames, such as 40 to 120 frames, roughly corresponding to 8 to 24 seconds, to depict a continuous operation process, thereby avoiding the discreteness and judgment lag problems caused by instantaneous action recognition in existing technologies. For example, a behavior sliding window can be represented as... i.e., behavior sliding window Including , At that moment, Indicates the window length.

[0024] Here, embodied operation signals refer to the objective physical and behavioral sequence data generated when a user interacts with their body in a virtual reality (VR) or metaverse training environment.

[0025] In practice, various sensors or virtual peripherals can be used to collect multimodal user operation data in real time according to a fixed sampling period. As an example ranging from large to small, these embodied operation signals can cover the overall user posture and environmental interaction data in a broad sense, or they can be more specific continuous signal information such as hand posture, hand movement trajectory, gaze object, gaze duration, and operation trigger frequency.

[0026] Step 120: Extract features from the embodied operation signal within the behavior sliding window to obtain embodied behavior features.

[0027] Here, embodied behavioral characteristics refer to the feature indicators that can quantitatively reflect a user's current operating habits, operational stability, or operational intentions, extracted through data processing and analysis of continuous embodied operation signals.

[0028] In practical implementation, various behavioral features can be obtained by employing algorithms such as spatial distance calculation or time series statistics for embodied operation signals of different dimensions. For example, operational hesitation features reflecting cognitive uncertainty can be extracted by analyzing the time a user spends in the target area before performing an operation; repeated attempts reflecting a cycle of failed attempts can be extracted by statistically analyzing the repetition rate of the same operation within a short period; path deviation features reflecting the stability of operation control can be extracted by calculating the distance of the hand trajectory from the expected area; and rhythmic anomaly features reflecting the degree of operational disorder can be extracted by utilizing the dispersion of the time interval between adjacent operations.

[0029] It should be noted that by extracting features from continuous segments, we can accurately reflect the stable trend of user operations rather than the absolute right or wrong of a single action.

[0030] Step 130: Based on the specific behavioral characteristics, determine the user's operation state under the behavior sliding window.

[0031] Here, "operational state" refers to a macroscopic description of the overall behavioral tendencies or cognitive stages exhibited by the user during continuous operations within the current action sliding window. It should be noted that this operational state does not depend on specific teaching task steps or nodes, but rather directly describes the stability of the current operational process.

[0032] In practical implementation, one or more embodied behavioral features extracted can be comprehensively analyzed and evaluated through preset rule combinations or classification mechanisms. For example, each embodied behavioral feature can be compared with its corresponding feature threshold. If the feature value deviates from the reference baseline, it is determined to be a specific state type. For instance, if the extracted features show operational hesitation and an increase in the value of the rhythm abnormality feature, the operational state under that window can be determined to be a hesitant state; if the operational hesitation and repeated attempts are obvious, it is determined to be an obstructed state; if the path deviation feature is significant, it is determined to be a deviation state; if other normal exploration features are present, it is determined to be an exploration state, and so on.

[0033] Step 140: When multiple consecutive sliding windows of the behavior correspond to the same abnormal operation state, the intensity of the guidance intervention is calculated based on the cumulative duration of the abnormal operation state.

[0034] Here, "abnormal operation state" refers to an operation state that deviates from the normal expected trajectory or rhythm, or exhibits a negative cognitive tendency, such as the aforementioned hesitant state, obstructed state, or deviation state. Here, "cumulative duration" refers to the total time span between the earliest consecutive occurrence of the same abnormal operation state and the current behavior sliding window. Here, "guidance intervention intensity" refers to the degree of intervention to correct or assist the user's current operation process in order to rectify abnormal behavior; a higher value indicates a deeper level of intervention.

[0035] In practice, to ensure the reliability of the judgment and to resist interference, the state sequence of adjacent sliding windows can be continuously monitored to avoid false triggers caused by accidental jitter or brief distraction. Once multiple consecutive behavior sliding windows are detected to exhibit the same abnormal operation state, such as 3 to 10 consecutive windows, a guidance mechanism is triggered and the cumulative duration of the abnormal operation state is recorded.

[0036] Subsequently, the cumulative duration is compared with a pre-set maximum duration threshold, such as by calculating a ratio, to determine the corresponding guidance intervention intensity. Generally, the longer the abnormal behavior lasts, the higher the calculated guidance intervention intensity.

[0037] Step 150: Guide the user's learning behavior based on the guidance strategy corresponding to the intensity of the guidance intervention.

[0038] Here, guidance strategy refers to a series of specific immersive feedback and intervention schemes adopted to correct abnormal user behavior and assist them in completing correct operations.

[0039] In practice, instead of using pop-ups or text prompts to forcibly interrupt the user's workflow, the guidance information is mapped into the user's current interaction space. Based on the calculated range of guidance intervention intensity, a matching guidance strategy is automatically selected and executed, ensuring that the level of intervention is consistent with the degree of abnormality in the user's behavior.

[0040] For example, when the intensity of guidance intervention is low, spatial guidance strategies can be used, such as highlighting or enhancing the boundary contour of the target area to provide directional cues; when the intensity of guidance intervention is medium, action demonstration strategies can be used, such as overlaying a continuous virtual gesture demonstration trajectory on the user's current hand position; when the intensity of guidance intervention is high, operational constraint strategies can be used, such as applying damping to deviations from the direction or reducing increments.

[0041] The method provided in this invention acquires embodied operation signals within a behavior sliding window and extracts embodied behavior features to determine continuous operation states. When continuous abnormal operation states are identified, the intensity of guidance intervention is dynamically calculated based on the cumulative duration. Finally, a corresponding guidance strategy is matched to guide learning behavior. This method achieves a proactive and seamless immersive behavior guidance mechanism through dynamic evaluation and adaptive intervention of the user's continuous operation states. This overcomes the shortcomings of traditional result-oriented or node-triggered methods, such as intervention lag and interaction breaks. It achieves smooth, accurate, and timely correction before errors occur, thereby significantly improving the efficiency of skill training in the metaverse environment and the user's immersive learning experience.

[0042] Based on the above embodiments, the embodied operation signal includes target pointing data, operation rhythm data, and hand movement trajectory data. Here, the embodied operation signal at any time t can be expressed as: ,in, This represents the specific operation signal at time t; The hand pose at time t is represented by the three-dimensional spatial coordinates (x, y, z) of the hand and the rotation quaternion q; The gaze behavior data at time t should include at least the gazed object obtained from ray testing. Simultaneously record the duration of gaze. Frequency of eye contact switching Here, the hand pose and gaze behavior data at time t can be combined into target pointing data at time t. This represents an array of vectors representing the hand movement trajectory, i.e., the hand movement trajectory data. This represents the data related to the operation rhythm.

[0043] Step 120 includes: Based on the target pointing data, the operation target is determined, and the user's dwell time near the operation target before the key operation is obtained in the current behavior sliding window, as well as the reference dwell time of similar historical operations. Based on the dwell time and the reference dwell time, the operation hesitation characteristics are determined.

[0044] Here, target-oriented data typically refers to data reflecting the direction of a user's visual or motor focus, such as the object of attention observed by the gaze ray or data on objects nearby in the spatial position of the hand. The operational target here is the expected interactive virtual entity or area mapped to the aforementioned target-oriented data.

[0045] In practice, the target data is parsed to pinpoint the user's intended interaction target. Next, the dwell time consumed by the user in the vicinity of the target operation before triggering a key action for a specific task is calculated. This dwell time reflects the user's observation or hesitation time within the current behavior slider. Simultaneously, a reference dwell time is retrieved online; this reference dwell time is a baseline value determined based on the average dwell time before performing similar operations in historical windows.

[0046] Then, to objectively quantify the user's cognitive uncertainty, this step directly calculates the ratio of the currently acquired dwell time to the aforementioned reference dwell time, and uses this calculated ratio as the operational hesitation feature. Here, the operational hesitation feature can be calculated using the following formula, as shown below: ; In the formula, This represents the dwell time within the sliding window for the k-th line. Compared with the above reference stay time The ratio between these two values ​​represents the operational hesitation characteristic. This ratio, as a continuous numerical characteristic variable, accurately reflects the relative degree of hesitation in the current operation compared to the historical baseline.

[0047] Based on the operation rhythm data, the total number of independent operation actions and the number of repetitions of the same operation within the current behavior sliding window are obtained. Based on the number of repetitions and the total number of independent operation actions, the repeated trial feature is determined.

[0048] Here, the operation rhythm data records the temporal sequence information of user actions, such as a series of operation trigger time nodes and the number of times the operation is repeated. The total number of independent operations refers to the total number of all independent operations segmented according to the operation trigger time sequence and operation semantics within the current single behavior sliding window. The number of repetitions refers to the frequency with which an operation targeting the same semantic goal is repeatedly executed within a short period of time.

[0049] In practice, time series analysis can be used to statistically obtain the total number of independent actions and the number of repetitions of the same action. Then, based on the mathematical relationship between these two, a repetitive attempt feature representing the degree of homogeneous and redundant user actions can be determined. For example, the ratio of the number of repetitions to the total number of independent actions can be calculated, and this ratio can be directly used as the repetitive attempt feature to objectively record the user's tendency to repeat actions within the sliding window. Here, the total number of independent actions can be obtained by segmenting and statistically analyzing the action trigger time series in the action rhythm data, combined with action semantics, such as whether it is aimed at a new target or whether there is a significant interval.

[0050] Here, the repeated trial feature can be calculated using the following formula, as shown below: ; In the formula, This represents the ratio of the number of repetitions within the sliding window of the k-th action to the total number of independent actions, i.e., the repeated attempts feature.

[0051] Based on the hand movement trajectory data, the distance parameter between the sampling point of the user's hand movement trajectory within the current behavior sliding window and the center point of the desired operation area is obtained, and the path deviation feature is determined based on the distance parameter.

[0052] Here, hand movement trajectory data is used to characterize the continuous change path of the user's hand pose in a three-dimensional virtual space. It consists of a series of three-dimensional spatial position coordinates acquired at a fixed sampling period, i.e., hand movement trajectory sampling points. Here, the desired operation area center point refers to the center position of the ideal operation area that the user should accurately reach, as set according to the current training task.

[0053] In practice, the spatial distance between each hand movement trajectory sampling point extracted within the current behavior sliding window and the center point of the desired operation area is calculated to obtain a set of distance parameters. Subsequently, based on the overall statistical results of these distance parameters, path deviation features reflecting the stability of the user's hand control and the accuracy of the operation direction are determined. For example, the distance parameters corresponding to all sampling points within the window can be summed and averaged. The calculated average distance value can be directly used as the path deviation feature to objectively characterize the overall deviation of the hand trajectory within the current window. Here, the path deviation feature can be calculated using the following formula, as shown below: ; In the formula, The path deviation feature is defined within the k-th action time window; N represents the total number of hand movement trajectory sampling points within the k-th action time window. This represents the sampling point of the i-th hand movement trajectory; This indicates the center point of the desired operation area.

[0054] Based on the operation rhythm data, the time interval parameter between adjacent operations within the current behavior sliding window is extracted, and the rhythm anomaly characteristics are determined based on the dispersion of the time interval parameter.

[0055] Here, to reflect the continuity of operations, we can delve deeper into the operation rhythm data and extract an array of time interval parameters consisting of the differences between the trigger times of two adjacent independent operations within the current behavior sliding window. Here, the degree of dispersion is used to measure the fluctuation or dispersion of this set of time interval parameters.

[0056] In practice, the rhythm anomaly characteristics are directly determined based on the degree of dispersion of the calculated time interval parameters. For example, the mathematical variance of the time interval parameter array is calculated, and this variance value is directly used as the rhythm anomaly characteristic. Here, the rhythm anomaly characteristic can be calculated using the following formula, as shown below: ; In the formula, This represents the rhythmic anomaly characteristics within the k-th behavior time window; Represents the variance function; This indicates that the time series is triggered by the actions within the behavior window. The extracted array of time interval parameters for adjacent time operations. ,but It should be noted that this variance value objectively reflects the data dispersion of the operation, which varies in speed; the larger the value, the more disordered the operation rhythm.

[0057] The embodied behavior feature is obtained based on at least one of the aforementioned operational hesitation feature, repeated attempt feature, path deviation feature, and rhythm anomaly feature. Specifically, after obtaining the specific numerical features of each of the above dimensions, the finally extracted embodied behavior feature can be composed of any single numerical feature among the aforementioned operational hesitation feature, repeated attempt feature, path deviation feature, and rhythm anomaly feature, or it can be a multi-dimensional feature vector formed by concatenating multiple feature values. This embodiment of the invention does not impose specific limitations on this. For example, the embodied behavior feature can be represented as... .

[0058] The method provided in this invention extracts operational hesitation features based on the ratio of dwell time to reference dwell time, repeated attempt features based on the ratio of repetition count to total number, path deviation features based on average distance, and rhythm anomaly features based on time interval variance from target pointing data, operation rhythm data, and hand movement trajectory data. This comprehensive method obtains embodied behavior features, strictly distinguishes the boundary between feature extraction and subsequent state determination, and accurately converts the user's continuous embodied operation signals into objective numerical features. This avoids the loss of determination information caused by prematurely introducing threshold comparisons during the feature extraction stage, thereby maximizing the preservation of continuous change details during the operation process and providing a high-fidelity data foundation for more accurate and interference-resistant operation state evaluation.

[0059] Based on any of the above embodiments, step 130 includes: Obtain the preset feature threshold corresponding to the embodied behavioral feature.

[0060] Here, the preset feature threshold refers to the critical benchmark used to determine whether the corresponding embodied behavioral characteristics are within a normal and reasonable range.

[0061] In practical implementation, for each dimension of embodied behavioral feature, a corresponding judgment threshold is obtained. For example, for the hesitant operation feature, a preset hesitant time threshold can be obtained; for the repeated attempts feature, a preset repetition ratio threshold can be obtained; for the path deviation feature, a preset deviation distance threshold can be obtained, and so on. It should be noted that the threshold can be obtained by directly calling a pre-set fixed constant, or by obtaining a baseline parameter that is dynamically updated based on historical performance.

[0062] The embodied behavioral features are compared with the corresponding preset feature thresholds to determine any abnormal deviations in the embodied behavioral features.

[0063] Here, abnormal deviation refers to the degree or fluctuation trend of the extracted current specific feature value exceeding the baseline threshold.

[0064] In practice, analytical methods such as numerical comparison or trend determination are used to assess whether the current embodied behavioral characteristics deviate from the reasonable range expected by normal operation. For example, the calculated value of operational hesitation characteristics can be compared with the corresponding acquired hesitation time threshold. If the former is greater than the latter, it is determined that there is an abnormal deviation of hesitation exceeding the limit. Alternatively, the current rhythm abnormality characteristic values ​​can be compared with trends to determine whether there is an abnormal deviation of rhythm dispersion gradually increasing.

[0065] Based on the abnormal deviation and the preset state determination rules, the user's operation state under the behavior sliding window is determined.

[0066] Here, the state determination rule refers to the logical combination condition that maps the abnormal deviation of one or more underlying embodied behavioral characteristics to the macroscopic operational state of the upper layer. This rule realizes the complete decoupling of the judgment of the underlying continuous behavioral characteristics from the specific teaching task content.

[0067] In practice, by inputting the abnormal deviations of each dimension as defined above, a joint logical judgment based on multiple features is performed using state determination rules. For example, if the rule detects that the operation hesitation feature is greater than the corresponding threshold and the rhythm anomaly feature shows an increase in value, then the current operation state is determined to be a hesitation state; if the rule detects that the operation hesitation feature is greater than the corresponding threshold and the repeated attempts feature is also greater than the corresponding threshold, then the current state is determined to be an obstructed state; if only the path deviation feature is greater than the corresponding threshold, then it is determined to be a deviation state; if the rhythm anomaly feature increases but the path deviation feature is less than the preset feature threshold, then it can be determined to be a normal exploration state, etc.

[0068] The method provided in this invention obtains abnormal deviations by comparing preset feature thresholds corresponding to embodied behavioral features, and then comprehensively determines the operation state by combining preset state determination rules. This method quantifies low-level continuous multimodal behavioral features into high-level operation process states, eliminating the dependence of guidance feedback on specific task steps or final task failure results. Simultaneously, by flexibly combining feature determination rules to describe the operation state, it effectively decouples the determination logic from specific teaching content, thereby significantly improving the reusability and cross-scenario transferability of the operation state determination mechanism in different metaverse training scenarios and different learning processes.

[0069] Based on any of the above embodiments, step 130, obtaining the preset feature threshold corresponding to the embodied behavioral feature, includes: Obtain the current average feature index of the user for the embodied behavioral feature within the historical operation period, as well as the group average reference feature and initial feature baseline corresponding to the embodied behavioral feature.

[0070] Here, the historical operation cycle refers to the effective time period or batch of operations experienced by the user during past metaverse training or learning interactions. Here, the current average characteristic metric refers to a measure of the user's average operational level or habits regarding a specific behavioral characteristic, based on long-term statistical results; for example, the user's average dwell time in historical data.

[0071] Furthermore, the group average reference feature here refers to a predefined feature reference scale that represents the average behavioral level of the general population or all learners. Here, the initial feature baseline is a default baseline judgment threshold pre-set for this embodied behavioral feature.

[0072] In practice, by retrieving the operation logs and long-term historical statistical data, the historical average data of individual users, the group benchmark reference data, and the initially set basic threshold parameters corresponding to a certain specific behavioral characteristic are extracted and read.

[0073] Based on the current average feature index of the embodied behavioral characteristics and the group average reference features, the initial feature baseline is adjusted to obtain the preset feature threshold corresponding to the embodied behavioral characteristics.

[0074] In practice, the difference between the current average feature index and the group average reference feature can be calculated to obtain a deviation that reflects the difference between the user's individual operating habits and proficiency and the general population. This deviation is then used to dynamically correct or compensate the initial feature baseline, thereby calculating and updating a preset feature threshold suitable for the current user. As an optional example of decreasing the threshold, the user's current average feature index can be subtracted from the group average reference feature, and the difference multiplied by a preset adaptive adjustment weight coefficient. This difference is then added as a compensation term to the initial feature baseline, and the latest personalized feature threshold is calculated using this closed-loop adjustment algorithm.

[0075] Here, the preset feature threshold for embodied behavioral characteristics can be calculated based on the following formula, as shown below: ; In the formula, The preset feature threshold represents the i-th type of embodied behavior feature; The initial feature baseline represents the i-th type of embodied behavior feature; Indicates the weighting parameter; The current average feature index represents the i-th type of embodied behavioral characteristics; This represents the preset group average reference characteristic.

[0076] The method provided in this invention dynamically and adaptively adjusts the initial feature baseline by acquiring the current average feature index reflecting individual habits and the group average reference feature reflecting the baseline level, thereby calculating a personalized preset feature threshold. This method effectively takes into account the differences in operating habits and proficiency among different learners. Experienced users face more lenient judgment conditions and receive less interference, while novice users face more sensitive judgment conditions and receive more necessary auxiliary intervention. It achieves a highly individualized trigger adjustment mechanism without relying on specific teaching process settings, greatly improving the long-term stability of the state judgment rules and their cross-user scenario adaptability.

[0077] Based on any of the above embodiments, in step 140, the intensity of the guidance intervention is calculated based on the cumulative duration of the abnormal operation state, including: First, obtain the maximum duration threshold for the abnormal operation state.

[0078] Here, the maximum duration threshold refers to a pre-configured reference standard that measures the maximum duration that can be tolerated for a particular abnormal operating state.

[0079] In practical implementation, considering that different types of abnormal operation states have varying degrees of impact on the operation progress, corresponding tolerance upper limit durations can be configured for each abnormal operation state. After determining the specific abnormal type, the preset limit duration parameter corresponding to that abnormal operation state can be directly read and obtained as the maximum duration threshold.

[0080] Alternatively, dynamic maintenance and real-time adjustment can be performed based on the user's historical performance to obtain a dynamic maximum duration threshold for abnormal operation states. In practice, firstly, historical performance data of the user during the training process is extracted, such as historical task scores and historical abnormal trigger frequency—proficiency indicators. Then, based on this historical performance data, the initially configured limit duration parameter is dynamically adjusted, and the dynamically adjusted value is determined as the currently obtained maximum duration threshold. As an example of decreasing the threshold, if historical performance data indicates that the user has high proficiency and excellent historical performance, the maximum duration threshold can be dynamically increased to give them more room for autonomous exploration and error correction; conversely, if historical performance data indicates that the user is a beginner, the maximum duration threshold is decreased to ensure that a higher intensity of intervention is triggered after a shorter period of abnormal operation.

[0081] Then, the ratio of the cumulative duration to the maximum duration threshold is calculated.

[0082] In practice, the cumulative duration of the current abnormal operation state obtained from the aforementioned statistics is divided by the obtained maximum duration threshold, and the corresponding ratio is obtained through division. It should be noted that this ratio reflects the proportion of the current abnormal behavior duration within the set maximum tolerance time interval, demonstrating the dynamic growth relationship in which the intervention intensity gradually increases proportionally with the duration of the abnormality.

[0083] Finally, the intensity of the guidance intervention is determined based on the smaller of the ratio and the preset intensity upper limit.

[0084] Here, the preset intensity upper limit refers to the maximum intervention level set to prevent over-intervention. For example, it can be set to a value of 1, representing the maximum level of system intervention. The guidance intervention intensity here is the specific quantitative indicator that ultimately determines which level of guidance strategy to adopt.

[0085] In practice, the calculated ratio is compared with the preset intensity upper limit value, taking the minimum value. When the cumulative duration is short and the calculated ratio does not exceed the preset intensity upper limit value, the ratio can be directly determined as the actual guidance intervention intensity. However, when the abnormal state lasts long enough, causing the cumulative duration to reach or exceed the maximum duration threshold, resulting in a calculated ratio greater than or equal to the preset intensity upper limit value, truncation is performed, and the preset intensity upper limit value is directly taken as the final guidance intervention intensity. In one embodiment, the guidance intervention intensity can be calculated using the following formula, as shown below: ; In the formula, Indicates the intensity of guidance and intervention; Indicates the cumulative duration of the abnormal operation state; The threshold representing the maximum duration of an abnormal operation state.

[0086] It should be noted that the intensity of the guidance intervention can also be dynamically adjusted based on the actual improvement in the situation. Adjustment methods include: determining the initial intervention intensity as the smaller of the ratio and a preset intensity upper limit; if a guidance intervention record from the previous round exists, obtaining the change in the current abnormal feature within the current behavior sliding window relative to the abnormal feature from the previous round; if the change in abnormal feature indicates improvement in the behavior state, downgrading the initial intervention intensity based on a preset attenuation coefficient, for example, multiplying the initial intervention intensity by a preset attenuation coefficient of 0.7 to obtain the final guidance intervention intensity. ,in This indicates the initial intervention intensity.

[0087] When the changes in the abnormal characteristics indicate that the abnormality persists or intensifies, the initial intervention intensity is upgraded based on a preset increment, for example, by increasing the preset increment by 0.1. The smaller of the upgraded intervention intensity and the preset upper limit is determined as the final guided intervention intensity. ,,in This indicates the initial intervention intensity.

[0088] It should be noted that by introducing a closed-loop adjustment logic based on the previous round of intervention records and the current changes in abnormal characteristics, the intervention intensity is reduced when the state improves and increased when the abnormality intensifies, with the addition of a safety upper limit control. This achieves self-assessment and dynamic closed-loop adjustment of the intervention intensity. This not only ensures that the strong intervention is not maintained for a long time due to sluggishness, but also that the prompts are not removed too early when necessary due to oversensitivity. It fundamentally avoids frequent flashing prompts or repeated false triggers during the interaction process, and maintains the best user immersion experience and intervention effectiveness in the dynamic game.

[0089] The method provided in this invention determines the intensity of guidance intervention by calculating the ratio of the cumulative duration of the abnormal operation state to a maximum duration threshold, and comparing this ratio with the smaller value of a preset intensity upper limit. This method achieves a dynamic linear positive correlation between intervention intensity and the duration of abnormal behavior, allowing the intervention intensity to automatically and smoothly increase as the duration of the abnormal behavior extends, stabilizing at the maximum intervention intensity when a preset time limit is reached. This ensures that subsequent intervention actions follow a smooth transition from suggestion to explicit instruction, effectively avoiding the abrupt interruptions of traditional fixed-intensity feedback and greatly optimizing the learner's immersive experience.

[0090] Based on any of the above embodiments, step 150 includes: When the intensity of the guidance intervention is within a first preset range, the target area for the expected interaction is determined, and spatial guidance information is applied to the target area.

[0091] When the intensity of the guidance intervention is within the second preset range, a virtual demonstration trajectory that is continuously connected to the user's current body position is superimposed in the user's current operating space.

[0092] When the intensity of the guidance intervention is within the third preset range, the user's current operation direction is obtained, and damping or reduction increments are applied to the input components in the current operation direction that deviate from the expected direction.

[0093] The intensity values ​​of the first preset interval, the second preset interval, and the third preset interval increase sequentially.

[0094] It should be noted that the intensity values ​​of the first, second, and third preset intervals increase sequentially, representing the degree of intervention from weak to strong.

[0095] Specifically, when the intensity of the guidance intervention falls within a first preset range, the target area for expected interaction is determined, and spatial guidance information is applied to the target area. The first preset range refers to a numerical range where the intensity of the guidance intervention is at a low level, for example, between 0.1 and 0.4. Here, the target area refers to the virtual three-dimensional spatial range that the user is expected to reach or interact with based on the current learning or operational process. Here, spatial guidance information refers to visual or directional cue elements provided without altering the user's autonomous operational path.

[0096] In practice, when the calculated guidance intervention intensity is determined to fall within a first preset range, the target area where the user should interact is identified by analyzing the current task requirements. Then, spatial highlighting or boundary contour enhancement is applied to this target area to display the aforementioned spatial guidance information. It should be noted that this method provides the user with clear directional cues, serving as a subtle form of suggestive guidance.

[0097] Furthermore, when the guidance intervention intensity falls within the second preset range, a virtual demonstration trajectory continuously connected to the user's current body position is superimposed on the user's current operating space. The second preset range refers to a numerical range where the guidance intervention intensity is at a moderate level, for example, between 0.4 and 0.7. Here, the current operating space refers to the 3D virtual environment in which the user is currently engaging in embodied interaction. The virtual demonstration trajectory refers to a virtual action path with visual representations such as semi-transparency, generated to guide the user to perform the correct action.

[0098] In practice, when the intervention intensity increases and falls within a second preset range, the coordinates of the user's current body position, such as their hands, are extracted. Combined with the center point of the desired operating area, algorithms such as vector synthesis and interpolation are used to calculate and generate a virtual demonstration trajectory that smoothly and continuously connects to the current body position. This trajectory is then displayed non-intrusively overlaid on the current operating space. It should be noted that this method provides users with an intuitive and concrete demonstration of movement trends while maintaining their autonomy.

[0099] In one embodiment, taking the user's current body position as the user's current hand position as an example, the virtual demonstration trajectory continuously connected to the user's current body position can be calculated by the following formula, as shown below: ; In the formula, Indicates a virtual demonstration trajectory; Indicates the user's current hand position; Indicates the guiding strength coefficient; This indicates the desired location of the center point of the operating area.

[0100] In addition, when the intensity of the guidance intervention is within the third preset range, the user's current operation direction is obtained, and damping or reduction of the input component that deviates from the expected direction in the current operation direction is applied.

[0101] The third preset interval refers to a numerical range where the intensity of the guidance intervention is at a relatively high level, such as an intensity value between 0.7 and 1.0. Here, the current operation direction refers to the physical vector direction of the user's real-time limb movement or gaze movement. The damping or reduction increment refers to the operational parameters at the underlying physical simulation engine or interaction logic level that resist or reduce the user's erroneous movement components.

[0102] In practice, when the intervention intensity reaches the third preset interval of the highest level, the user's current operation direction is monitored in real time and compared with the expected correct operation direction through vector decomposition to extract the erroneous input component that deviates from the expected direction. Then, without freezing or locking the user's overall operation permissions, a reverse damping force is applied to the deviating input component or its displacement increment is directly reduced. Thus, the user's operation trend can be semi-forcefully corrected while maintaining continuous interaction.

[0103] The method provided in this invention maps sequentially increasing first, second, and third preset intervals to different intervention strategies, such as applying spatial guidance information, overlaying virtual demonstration trajectories, and applying damping or weakening increments. It executes corresponding levels of interactive feedback based on the intensity of the guidance intervention, achieving a smooth transition from mild visual cues and moderate explicit actions to severe physical operational constraints. All mapping feedback is directly overlaid within the original interactive channel, rather than using abrupt mode switching such as pop-ups or pauses. This effectively achieves correction and collaborative operation while maintaining the continuity of interaction and immersive learning within the metaverse environment, greatly enhancing the acceptability and effectiveness of the system guidance.

[0104] Based on any of the above embodiments Figure 2 This is one of the structural schematic diagrams of the metaverse learning behavior guidance device provided by the present invention, such as... Figure 2 As shown, the device as a whole constitutes an intelligent learning enhancement architecture that integrates perception and dynamic collaborative decision-making. It mainly includes a behavior signal acquisition module, a behavior pattern analysis module, a guidance trigger module, a guidance mapping module, and a behavior feedback and adjustment module that form a closed loop, which are connected in sequence.

[0105] Specifically, the behavior signal acquisition module is used to acquire multimodal embodied operation signals of the user in a preset behavior sliding window in real time at a fixed sampling period in a metaverse or virtual training environment. These signals include, but are not limited to, continuous time series data streams such as target pointing data, operation rhythm data, and hand movement trajectory data.

[0106] The behavior pattern analysis module is connected to the behavior signal acquisition module. It is used to receive sliding window data and perform in-depth embodied behavior feature extraction, quantifying indicators such as operation hesitation features, repeated attempts features, path deviation features, and rhythm abnormal features. Then, based on these embodied behavior features and preset rules, it determines the macroscopic operation state of the user in the current sliding window.

[0107] The guidance triggering module receives the status judgment result from the behavior pattern analysis module. When the operation status corresponding to multiple consecutive behavior sliding windows is the same abnormal operation status, it extracts the cumulative duration of the abnormal operation status and calculates the proportionally increasing guidance intervention intensity accordingly. This enables real-time intervention before the error is formed and avoids false triggering caused by accidental errors.

[0108] The guidance mapping module matches and executes the corresponding guidance strategy based on the intensity of the output guidance intervention to guide the user's learning behavior. By seamlessly overlaying spatial guidance information, virtual demonstration trajectories, or applying physical operation damping into the user's original interaction channel, it achieves immersive feedback without interrupting the current action flow.

[0109] The behavior feedback and adjustment module receives subsequent data after the guidance is executed, and its output is fed back to the aforementioned behavior pattern analysis module. This module is used to recalculate behavioral characteristics and states to evaluate the intervention effect as the guidance process continues to intervene. This allows for dynamic adjustment of the guidance intervention intensity and adaptive updates to the user's baseline characteristic parameters, thus forming an adaptive closed-loop adjustment mechanism with long-term stability and cross-scenario adaptability.

[0110] Based on any of the above embodiments Figure 3 This is the second structural schematic diagram of the metaverse learning behavior guidance device provided by the present invention, as shown below. Figure 3 As shown, the device includes: Acquisition unit 310 acquires the user's specific operation signal within the behavior sliding window; Feature extraction unit 320 extracts features from the embodied operation signal within the behavior sliding window to obtain embodied behavior features; The state determination unit 330 determines the user's operation state under the behavior sliding window based on the specific behavioral characteristics. The intervention intensity calculation unit 340 calculates the guidance intervention intensity based on the cumulative duration of the abnormal operation state when the operation states corresponding to multiple consecutive behavior sliding windows are all the same abnormal operation state. The guidance unit 350 guides the user's learning behavior based on a guidance strategy corresponding to the intensity of the guidance intervention.

[0111] The device provided in this invention acquires embodied operation signals within a behavior sliding window and extracts embodied behavior features to determine continuous operation states. When continuous abnormal operation states are identified, the intensity of guidance intervention is dynamically calculated based on the cumulative duration, and finally, a corresponding guidance strategy is matched to guide learning behavior. This method achieves a proactive and seamless immersive behavior guidance mechanism through dynamic evaluation and adaptive intervention of the user's continuous operation states. This overcomes the shortcomings of traditional result-oriented or node-triggered methods, such as intervention lag and interaction breaks, achieving smooth, accurate, and timely correction before errors occur. Consequently, it significantly improves skill training efficiency and the user's immersive learning experience in the metaverse environment.

[0112] Based on any of the above embodiments, the personal operation signal includes target pointing data, operation rhythm data, and hand movement trajectory data; The feature extraction unit is specifically used for: Based on the target pointing data, the operation target is determined, and the dwell time of the user near the operation target before the key operation is obtained in the current behavior sliding window, as well as the reference dwell time of similar historical operations. Based on the dwell time and the reference dwell time, the operation hesitation characteristics are determined. Based on the operation rhythm data, the total number of independent operation actions and the number of repetitions of the same operation within the current behavior sliding window are obtained. Based on the number of repetitions and the total number of independent operation actions, the repeated trial feature is determined. Based on the hand movement trajectory data, the distance parameter between the sampling point of the user's hand movement trajectory within the current behavior sliding window and the center point of the desired operation area is obtained, and the path deviation feature is determined based on the distance parameter. Based on the operation rhythm data, the time interval parameter between adjacent operations within the current behavior sliding window is extracted, and rhythm anomaly characteristics are determined based on the dispersion of the time interval parameter. The embodied behavior feature is obtained based on at least one of the operation hesitation feature, the repeated attempt feature, the path deviation feature, and the rhythm abnormality feature.

[0113] Based on any of the above embodiments, the state determination unit is specifically used for: Obtain the preset feature threshold corresponding to the embodied behavioral features; The embodied behavioral features are compared with the corresponding preset feature thresholds to determine any abnormal deviations in the embodied behavioral features. Based on the abnormal deviation and the preset state determination rules, the user's operation state under the behavior sliding window is determined.

[0114] Based on any of the above embodiments, the state determination unit is further specifically used for: Obtain the current average feature index of the user for the embodied behavioral feature within the historical operation period, as well as the group average reference feature and initial feature baseline corresponding to the embodied behavioral feature; Based on the current average feature index of the embodied behavioral characteristics and the group average reference features, the initial feature baseline is adjusted to obtain the preset feature threshold corresponding to the embodied behavioral characteristics.

[0115] Based on any of the above embodiments, the intervention intensity calculation unit is specifically used for: Obtain the maximum duration threshold for the abnormal operation state; Calculate the ratio of the cumulative duration to the maximum duration threshold; The intensity of the guidance intervention is determined based on the smaller of the ratio and the preset intensity upper limit.

[0116] Based on any of the above embodiments, the guiding unit is specifically used for: When the intensity of the guidance intervention is within a first preset range, the target area for the expected interaction is determined, and spatial guidance information is applied to the target area. When the intensity of the guidance intervention is within the second preset range, a virtual demonstration trajectory that is continuously connected to the user's current body position is superimposed in the user's current operating space; When the intensity of the guidance intervention is within the third preset range, the user's current operation direction is obtained, and damping or reduction increments are applied to the input components in the current operation direction that deviate from the expected direction. The intensity values ​​of the first preset interval, the second preset interval, and the third preset interval increase sequentially.

[0117] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4As shown, the electronic device may include a processor 410, a communication interface 420, a memory 430, and a communication bus 440, wherein the processor 410, the communication interface 420, and the memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions in the memory 430 to execute a metaverse learning behavior guidance method. This method includes: acquiring the embodied operation signal of the user within a behavior sliding window; extracting features from the embodied operation signal within the behavior sliding window to obtain embodied behavior features; determining the user's operation state within the behavior sliding window based on the embodied behavior features; calculating the guidance intervention intensity based on the cumulative duration of the abnormal operation state when multiple consecutive behavior sliding windows correspond to the same abnormal operation state; and guiding the user's learning behavior based on a guidance strategy corresponding to the guidance intervention intensity.

[0118] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0119] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the metaverse learning behavior guidance method provided by the above methods. The method includes: acquiring the embodied operation signal of a user within a behavior sliding window; extracting features from the embodied operation signal within the behavior sliding window to obtain embodied behavior features; determining the user's operation state under the behavior sliding window based on the embodied behavior features; calculating the guidance intervention intensity based on the cumulative duration of the abnormal operation state when the operation states corresponding to multiple consecutive behavior sliding windows are the same abnormal operation state; and guiding the user's learning behavior based on the guidance strategy corresponding to the guidance intervention intensity.

[0120] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the metaverse learning behavior guidance method provided by the above methods. The method includes: acquiring embodied operation signals of a user within a behavior sliding window; extracting features from the embodied operation signals within the behavior sliding window to obtain embodied behavior features; determining the user's operation state under the behavior sliding window based on the embodied behavior features; calculating the guidance intervention intensity based on the cumulative duration of the abnormal operation state when multiple consecutive operation states corresponding to the behavior sliding windows are the same abnormal operation state; and guiding the user's learning behavior based on a guidance strategy corresponding to the guidance intervention intensity.

[0121] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0122] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0123] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A metaverse-based learning behavior guidance method, characterized in that, include: Obtain the user's specific action signals within the behavior slider window; Feature extraction is performed on the embodied operation signal within the behavior sliding window to obtain embodied behavior features; Based on the specific behavioral characteristics, the user's operational state under the behavioral sliding window is determined; When multiple consecutive sliding windows of the behavior correspond to the same abnormal operation state, the intensity of the guidance intervention is calculated based on the cumulative duration of the abnormal operation state. Based on the guidance strategy corresponding to the intensity of the guidance intervention, the user's learning behavior is guided.

2. The metaverse learning behavior guidance method according to claim 1, characterized in that, The embodied operation signal includes target pointing data, operation rhythm data, and hand movement trajectory data; The step of extracting features from the embodied operation signals within the behavior sliding window to obtain embodied behavior features includes: Based on the target pointing data, the operation target is determined, and the dwell time of the user near the operation target before the key operation is obtained in the current behavior sliding window, as well as the reference dwell time of similar historical operations. Based on the dwell time and the reference dwell time, the operation hesitation characteristics are determined. Based on the operation rhythm data, the total number of independent operation actions and the number of repetitions of the same operation within the current behavior sliding window are obtained. Based on the number of repetitions and the total number of independent operation actions, the repeated trial feature is determined. Based on the hand movement trajectory data, the distance parameter between the sampling point of the user's hand movement trajectory within the current behavior sliding window and the center point of the desired operation area is obtained, and the path deviation feature is determined based on the distance parameter. Based on the operation rhythm data, the time interval parameter between adjacent operations within the current behavior sliding window is extracted, and rhythm anomaly characteristics are determined based on the dispersion of the time interval parameter. The embodied behavior feature is obtained based on at least one of the operation hesitation feature, the repeated attempt feature, the path deviation feature, and the rhythm abnormality feature.

3. The metaverse learning behavior guidance method according to claim 2, characterized in that, The step of determining the user's operation state under the behavior sliding window based on the specific behavioral characteristics includes: Obtain the preset feature threshold corresponding to the embodied behavioral features; The embodied behavioral features are compared with the corresponding preset feature thresholds to determine any abnormal deviations in the embodied behavioral features. Based on the abnormal deviation and the preset state determination rules, the user's operation state under the behavior sliding window is determined.

4. The metaverse learning behavior guidance method according to claim 3, characterized in that, The step of obtaining the preset feature threshold corresponding to the embodied behavioral feature includes: Obtain the current average feature index of the user for the embodied behavioral feature within the historical operation period, as well as the group average reference feature and initial feature baseline corresponding to the embodied behavioral feature; Based on the current average feature index of the embodied behavioral characteristics and the group average reference features, the initial feature baseline is adjusted to obtain the preset feature threshold corresponding to the embodied behavioral characteristics.

5. The metaverse learning behavior guidance method according to any one of claims 1 to 4, characterized in that, The intensity of the guided intervention is calculated based on the cumulative duration of the abnormal operation state, including: Obtain the maximum duration threshold for the abnormal operation state; Calculate the ratio of the cumulative duration to the maximum duration threshold; The intensity of the guidance intervention is determined based on the smaller of the ratio and the preset intensity upper limit.

6. The metaverse learning behavior guidance method according to any one of claims 1 to 4, characterized in that, The guidance strategy based on the intensity of the guidance intervention, which guides the user's learning behavior, includes: When the intensity of the guidance intervention is within a first preset range, the target area for the expected interaction is determined, and spatial guidance information is applied to the target area. When the intensity of the guidance intervention is within the second preset range, a virtual demonstration trajectory that is continuously connected to the user's current body position is superimposed in the user's current operating space; When the intensity of the guidance intervention is within the third preset range, the user's current operation direction is obtained, and damping or reduction increments are applied to the input components in the current operation direction that deviate from the expected direction. The intensity values ​​of the first preset interval, the second preset interval, and the third preset interval increase sequentially.

7. A metaverse learning behavior guidance device, characterized in that, include: The acquisition unit acquires the user's specific operation signals within the behavior sliding window; The feature extraction unit extracts features from the embodied operation signal within the behavior sliding window to obtain embodied behavior features; The state determination unit determines the user's operation state under the behavior sliding window based on the specific behavioral characteristics. The intervention intensity calculation unit calculates the guidance intervention intensity based on the cumulative duration of the abnormal operation state when the operation states corresponding to multiple consecutive behavior sliding windows are all the same abnormal operation state. The guidance unit guides the user's learning behavior based on a guidance strategy corresponding to the intensity of the guidance intervention.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the metaverse learning behavior guidance method as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the metaverse learning behavior guidance method as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the metaverse learning behavior guidance method as described in any one of claims 1 to 6.

Images