Digital management system based on VR technology

Abstract

The invention discloses a digital management system based on a VR technology, and relates to the technical field of digital management, and the system comprises a task execution module which executes a scenario-type decision simulation task according to a task demand in a virtual scene, records the decision behavior and interaction data of a user in real time in the execution process, and sends the decision behavior and interaction data to a server; generating a preliminary feedback report according to the decision behavior of the user and the interaction data; and the execution condition evaluation module is used for extracting task evaluation indexes by adopting data mining on the basis of the preliminary feedback report, evaluating the task execution condition of the user by adopting an analytic hierarchy process on the basis of the task evaluation indexes, and dynamically adjusting the situational decision simulation task in the virtual scene according to an evaluation result. According to the invention, the execution condition evaluation module is utilized to carry out multi-dimensional deep analysis on the decision behavior and interaction data of the user, accurate feedback and guidance are provided, the improvement of personal ability is promoted, and the design and adjustment of subsequent tasks are optimized.

Description

Technical Field

[0001] The present invention relates to the field of digital management technology, in particular to a digital management system based on VR technology. Background Art

[0002] With the rapid development of virtual reality (VR) technology, its application in education, medical care, entertainment, enterprise management and other fields is becoming more and more extensive. In particular, with the in-depth promotion of digitalization and informatization, the traditional management model is gradually developing in the direction of intelligent, immersive interaction. Traditional learning and management methods are usually based on offline centralized learning and written material dissemination. The form is fixed and the interactivity is low, which makes it difficult to fully mobilize the enthusiasm of users to participate. In recent years, with the help of the immersive and interactive characteristics of VR technology, the management system has gradually broken through the limitations of the traditional model. By building virtual scenes and combining vivid situational learning content, it provides users with a more vivid, intuitive and efficient learning and management platform.

[0003] However, the existing management systems based on VR technology still have some shortcomings. First, most systems lack flexibility in content generation and cannot dynamically adjust virtual scenes and learning content according to the user's role and task requirements. This "one-size-fits-all" model makes it difficult to provide personalized learning experiences based on the actual needs of different users, thus affecting the actual teaching effect of the system. Secondly, in terms of the evaluation of user task execution effects, existing technologies usually remain at the level of simple result judgment or basic behavior recording, and lack the ability to comprehensively analyze multi-dimensional indicators such as user decision-making logic and behavior patterns, making it difficult to accurately evaluate and feedback the user's task completion. These technical bottlenecks have hindered the further promotion and optimization of VR management systems. Summary of the invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a digital management system based on VR technology to solve the problems of the existing system in terms of flexibility in content generation and accurate evaluation of user task execution effects.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: The present invention provides a digital management system based on VR technology, which includes: a data acquisition module, which collects user information through VR equipment, pre-processes the user information, and identifies the user role and task requirements according to the pre-processed user information; a scene construction module, which uses PCG to extract knowledge points, scene templates and task materials from a content database according to the user role and task requirements, dynamically generates a virtual scene matching the user role, and loads the virtual scene into the user's VR interface; a task execution module, which executes a scenario-based decision-making simulation task according to the task requirements in the virtual scene, records the user's decision-making behavior and interaction data in real time during the execution process, and generates a preliminary feedback report according to the user's decision-making behavior and interaction data; an execution status evaluation module, which uses data mining to extract task evaluation indicators based on the preliminary feedback report, uses hierarchical analysis method to evaluate the user's task execution based on the task evaluation indicators, and dynamically adjusts the scenario-based decision-making simulation task in the virtual scene according to the evaluation result.

[0007] As a preferred solution of the digital management system based on VR technology described in the present invention, the user information includes the user's identity information, behavior data, decision data and interaction data.

[0008] As a preferred solution of the digital management system based on VR technology described in the present invention, the specific steps of preprocessing user information are as follows: Identify and remove outliers in user information through the Z-score method; Mean filling is used to fill missing values ​​in user information; Filter the user information for noise through a low-pass filter; The label mapping method is used to standardize user information.

[0009] As a preferred solution of the digital management system based on VR technology described in the present invention, the specific steps of identifying user roles and task requirements according to the pre-processed user information are as follows: Based on the pre-processed user information, the user's identity features are extracted by direct mapping, the user's behavior features are extracted by statistical analysis, the user's interaction features are extracted by log analysis, and the user's decision features are extracted by frequency statistics; based on the user's identity features, behavior features, interaction features, and decision features, feature classification is used to classify and organize them to generate a structured feature set; Based on the structured feature set, define the rule logic and classification conditions, and finally build a rule engine. Use the rule engine to analyze users one by one and identify user roles. Based on the collection of structured features, the behavior trajectory analysis method is used to obtain the user's operation path, residence time and action sequence in the VR scene, and the frequency statistics method is combined to analyze the user's task selection preference, and finally identify the task requirements.

[0010] As a preferred solution of the digital management system based on VR technology described in the present invention, wherein: according to the user role and task requirements, PCG is used to extract knowledge points, scene templates and task materials from the content database, a virtual scene matching the user role is dynamically generated, and the virtual scene is loaded into the user's VR interface. The specific steps are as follows: according to the user role and task requirements, parameterized query is used to extract knowledge points, virtual scene templates and task materials from the content database; The extracted knowledge points, virtual scene templates and task materials are input into the PCG algorithm to dynamically adjust the position, size and color of the elements in the virtual scene to make them conform to the user role's interaction habits and task requirements, and predict the virtual scene scores S of different combinations. i ; Select Virtual Scene Score i The highest combination is used as a virtual scene and loaded into the user's VR interface through a VR interface.

[0011] As a preferred solution of the digital management system based on VR technology described in the present invention, the knowledge points include theoretical knowledge, policies and regulations, historical events, important speeches and practical cases; The task materials include text descriptions, pictures, audio and video resources and interactive content related to the task requirements; The construction process of the virtual scene template includes the following steps: defining the scene goal, scene subject and interaction requirements according to the user role and task requirements; Based on historical knowledge points and historical task materials, the logical framework and theme style of the virtual scene are planned through semantic association analysis; Based on the logical framework and theme style of the virtual scene, use 3D modeling tools to build a virtual scene template.

[0012] As a preferred solution of the digital management system based on VR technology described in the present invention, wherein: according to the task requirements in the virtual scene, the scenario-based decision-making simulation task is executed, the user's decision-making behavior and interaction data are recorded in real time during the execution process, and a preliminary feedback report is generated according to the user's decision-making behavior and interaction data. The specific steps are as follows: the user enters the virtual scene through the VR device, sets multiple scenario-based decision points according to the task requirements, and guides the user through virtual characters, visual prompts and voice prompts; According to the guidance, users perform decision-making tasks in virtual scenes through interactive devices, and the user's decision-making behavior and interaction data are recorded in real time during the execution process; After the user completes each decision, the rule engine analyzes the user's selection correctness, decision logic and behavior patterns in real time based on the recorded decision behavior and interaction data, and generates a preliminary feedback report.

[0013] As a preferred solution of the digital management system based on VR technology described in the present invention, wherein: based on the preliminary feedback report, data mining is used to extract task evaluation indicators, and based on the task evaluation indicators, the hierarchical analysis method is used to evaluate the user's task execution status. The specific steps are as follows: Based on the preliminary feedback report, data mining is used to conduct in-depth analysis of the preliminary feedback report and extract task evaluation indicators; Based on task evaluation indicators, asymmetric interaction effects and dynamic complexity adjustment mechanisms are used to construct a multidimensional relationship model; The interactive relationship between dimensions is analyzed based on the multidimensional relationship model, and a nonlinear function is introduced to characterize the complex relationship between selection correctness, decision logic and behavior pattern. At the same time, a penalty mechanism is used to process the absolute value of the differences between task evaluation indicators. Based on the task evaluation indicators, the hierarchical analysis method is used to evaluate the user's task execution status T.

[0014] As a preferred solution of the digital management system based on VR technology described in the present invention, wherein: the asymmetric interaction effect and dynamic complexity adjustment mechanism are used to construct a multidimensional relationship model, and the specific steps are as follows: Define selection correctness, decision logic and behavior patterns as the core dimensions of the multidimensional relationship model, and use correlation analysis to quantify the degree of correlation between core dimensions and preliminarily identify the positive and negative correlations between core dimensions; Based on the positive and negative correlations among the core dimensions, nonlinear regression method is used to identify the asymmetric interaction effects among the core dimensions. Based on the preliminary feedback reports, principal component analysis was used to extract task complexity factors; Based on the task complexity factor, the weight distribution of the core dimensions is dynamically adjusted through a nonlinear mapping function; Combining the asymmetric interaction effects among core dimensions and the weight distribution of core dimensions, a multidimensional relationship model is finally constructed.

[0015] As a preferred solution of the digital management system based on VR technology described in the present invention, the specific steps of dynamically adjusting the scenario-based decision-making simulation tasks in the virtual scene according to the evaluation results are as follows: Based on the user's historical task execution scores, define a low score threshold T1 and a high score threshold T2; When T≤T1, the task complexity is reduced and the guidance frequency is increased; When T1<T<T2, the task complexity and guidance frequency are maintained; When T≥T2, the task complexity is increased and the guidance frequency is reduced.

[0016] The beneficial effects of the present invention are as follows: through the scenario construction module, personalized virtual scenarios are dynamically generated according to user roles and task requirements, achieving a highly customized learning experience and greatly improving user participation and educational effects; at the same time, the execution evaluation module is used to conduct multi-dimensional in-depth analysis of users' decision-making behaviors and interaction data, providing accurate feedback and guidance, which not only promotes the improvement of personal capabilities, but also optimizes the design and adjustment of subsequent tasks, thereby overcoming the shortcomings of the prior art in content flexibility and user task execution effect evaluation, and significantly enhancing the practicality and teaching effectiveness of the digital management system. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0018] Figure 1 This is a schematic diagram of a digital management system based on VR technology in Example 1.

[0019] Figure 2 This is a flowchart of preprocessing user information in Example 1. DETAILED DESCRIPTION

[0020] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0021] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0022] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0023] Example 1, reference Figure 1 and Figure 2, which is the first embodiment of the present invention, and provides a digital management system based on VR technology, comprising the following steps: The data collection module collects user information through VR devices, pre-processes the user information, and identifies user roles and task requirements based on the pre-processed user information.

[0024] User information includes user identity information, behavior data, decision data, and interaction data.

[0025] It should be noted that the above content has been obtained with the user's consent and is used for legitimate purposes; Furthermore, user information is collected through a variety of sensors and interactive interfaces integrated in VR devices. Identity information is input through the system when the user registers or logs in, or obtained synchronously with an external database. Behavioral data is collected in real time by tracking dynamic behaviors such as user movement, operation frequency, and dwell time in the virtual scene. Decision data comes from the choices and solutions made by users in situational decision-making simulation tasks, and these data are obtained through interaction records during task execution. Interaction data includes interactions between users and the virtual environment and other users, such as conversation content, collaborative activities, and responses to prompts, which are captured by monitoring users' interactive behaviors and their operations in the VR interface.

[0026] Identify and remove outliers in user information through the Z-score method; It should be noted that the Z-score method is used to identify and eliminate outliers in user information. The Z-score of each user information is calculated, that is, the standard deviation between the current user information and the average value. If the absolute value of the Z-score of a certain user information exceeds the preset threshold, the data point is considered to be an outlier and is eliminated.

[0027] Mean filling is used to fill missing values ​​in user information; It should be noted that mean filling is used to fill missing values ​​in user information. For data fields with missing values, the average of all non-missing values ​​in the field is calculated, and this average is used as a substitute for the missing values.

[0028] Filter the user information for noise through a low-pass filter; It should be noted that the user information is noise filtered through a low-pass filter. The characteristic of the low-pass filter that allows signals below the cutoff frequency to pass while signals above the frequency are attenuated is utilized to process the behavioral data in the user information, remove high-frequency noise interference, and retain useful information components.

[0029] The label mapping method is used to standardize user information.

[0030] It should be noted that the label mapping method is used to standardize user information. Based on a pre-defined label system, various labels in the original user information are converted into numerical values ​​or classification codes under a unified standard to enhance the consistency of user information from different sources or formats.

[0031] Based on the pre-processed user information, the user's identity features are extracted by direct mapping, the user's behavior features are extracted by statistical analysis, the user's interaction features are extracted by log analysis, and the user's decision features are extracted by frequency statistics; It should be noted that, based on the pre-processed user information, the specific attributes in the identity information such as name, gender, age, etc. are directly mapped to the predefined identity feature categories through the direct mapping method to form structured identity feature data; the user's operation frequency, residence time, movement path and other behavioral data are quantitatively analyzed through the statistical analysis method, and indicators such as average residence time and most frequently visited areas are calculated to extract the user's behavioral characteristics; the user's interaction records in the virtual scene are parsed through the log analysis method, including the interaction with virtual objects, the communication content with other users, etc., to identify the interaction mode and preference, and then extract the user's interaction characteristics; the frequency of the user's choices in the decision-making simulation task is counted through the frequency statistics method to determine the decision type that the user tends to take and its changing trend, so as to extract the user's decision characteristics.

[0032] Based on the user's identity characteristics, behavior characteristics, interaction characteristics and decision-making characteristics, the feature classification method is used to classify and organize them to generate a structured feature set; It should be noted that based on the user's identity characteristics, behavioral characteristics, interaction characteristics and decision-making characteristics, the feature classification method is used for classification and organization. First, a series of standard classification dimensions are defined, which cover key aspects such as identity attributes, behavioral patterns, interaction types and decision-making tendencies; then, the various characteristics of each user are matched and classified according to the corresponding standard classification dimensions to ensure that the same type of features are integrated together to form a unified feature set; in this process, for each type of feature, the corresponding statistical values ​​or indicators are calculated, such as average values, frequency distributions, etc., to quantitatively represent the specific performance of the features; finally, a structured feature set is generated.

[0033] Based on the structured feature set, define the rule logic and classification conditions, and finally build a rule engine. Use the rule engine to analyze users one by one and identify user roles. It should be noted that, based on the structured feature set, a series of rule logics and classification conditions are first defined. These rules and conditions are formulated based on identity features, behavior features, interaction features, and decision features, aiming to accurately capture the unique attributes of different user roles. Subsequently, these rule logics and classification conditions are programmed into the rule engine, which analyzes each user's structured feature set one by one to identify the user role. Based on the collection of structured features, the behavior trajectory analysis method is used to obtain the user's operation path, residence time and action sequence in the VR scene, and the frequency statistics method is combined to analyze the user's task selection preference, and finally identify the task requirements.

[0034] It should be explained that, based on a structured feature set, the behavioral trajectory analysis method is used to track in detail the user's operation path, dwell time and action sequence in the virtual reality scene, and the user's activity pattern is accurately depicted by recording and analyzing each movement, interaction and pause point of the user in the VR environment; at the same time, the frequency statistics method is combined to conduct frequency statistics on different tasks chosen by the user to perform, and the task types that the user tends to participate in and their preferred order are identified; on this basis, the user's behavioral trajectory and task selection frequency are comprehensively considered, and the user's behavioral data is deeply analyzed to determine their specific task requirements and points of interest, and finally the task requirements are identified.

[0035] The scene construction module uses PCG to extract knowledge points, scene templates, and task materials from the content database according to user roles and task requirements, dynamically generates a virtual scene that matches the user role, and loads the virtual scene into the user's VR interface.

[0036] Based on user roles and task requirements, parameterized queries are used to extract knowledge points, virtual scene templates, and task materials from the content database; It should be noted that the content database is a resource library that stores a wealth of data and information, including theoretical knowledge, policies and regulations, historical events, important speeches, practical cases and other content; these materials come from official documents, historical archives, academic research results and practical experience summaries, and are included in the database after being sorted and digitized, providing solid content support for the digital management system based on VR technology, ensuring that users can obtain authoritative, comprehensive and in-depth learning resources.

[0037] The extracted knowledge points, virtual scene templates and task materials are input into the PCG algorithm to dynamically adjust the position, size and color of the elements in the virtual scene to make them conform to the user role's interaction habits and task requirements, and predict the virtual scene scores S of different combinations. i , the expression is: in, Si is the virtual scene score of the i-th combination, A i is the matching degree of knowledge points in the i-th combination, B i is the adaptation degree of the virtual scene template in the i-th combination, C i is the matching degree of the task materials in the ith combination, and e is the natural base number; Furthermore, the matching degree A of the knowledge points in the i-th combination i , the expression is: Among them, W 1 Represents the semantic features of the current task, W 2 The semantic content of the current knowledge point; The degree of adaptation of the virtual scene template in the i-th combination mode B i , the expression is: Among them, P 1 is the interactive behavior characteristic parameter of the user in the virtual scene, P 2 It is the design behavior characteristic parameter in the virtual scene template.

[0038] The matching degree of task materials is calculated by the correlation between the user's task requirements and the content of task materials. Specifically, according to the user's task selection preferences and task requirements, the feature vectors of the text description, pictures, audio and video content of the task materials are extracted, and the matching degree between the task materials and user requirements is quantified by combining the semantic matching model (such as TF-IDF or BERT), and finally C is obtained. i .

[0039] It should be noted that the elements in the virtual scene include knowledge points, virtual scene templates and task materials.

[0040] Interaction habits refer to the behavioral preferences and operation modes that users show when interacting with the system in a virtual environment, including the frequency of use of interactive devices, operation methods, visual focus, action paths, dwell time and other characteristics. These habits can be obtained by collecting and analyzing the user's behavior data in the VR scene. For example, the user's operation path and dwell time can be recorded through the behavior trajectory analysis method, and the user's operation preferences can be analyzed in combination with the frequency statistics method to obtain the user's interaction habits.

[0041] It should also be noted that based on the extracted knowledge points, virtual scene templates and task materials, they are input into the programmatic content generation algorithm, and the position, size, color and other attributes of the elements in the virtual scene are dynamically adjusted to ensure that these elements can highly match the user role's interaction habits and task requirements; for example, for users who focus on theoretical learning, the design of the text display area can be optimized to enhance the reading experience, while for users who prefer practical operations, the functionality of the interactive components is enhanced. In this process, the algorithm will evaluate the virtual scene effects under different configurations, predict their adaptability and user experience satisfaction, and select the best combination based on the calculated scores to construct the final virtual scene.

[0042] Select Virtual Scene Score i The highest combination is used as a virtual scene and loaded into the user's VR interface through a VR interface.

[0043] Knowledge points include theoretical knowledge, policies and regulations, historical events, important speeches and practical cases; Task materials include text descriptions, pictures, audio and video resources, and interactive content related to task requirements; The construction process of the virtual scene template includes the following steps: Define scenario goals, scenario subjects, and interaction requirements based on user roles and task requirements; It should be noted that the scenario goals, scenario subjects and interaction requirements are defined according to the user roles and task requirements. First, the user's specific role, such as learner or educator, is clarified, and their specific task requirements in the virtual scenario, such as theoretical learning or practical operations, are determined; then, based on this information, specific scenario goals are set, such as in-depth understanding of the development process of a certain historical period, and at the same time, the main content of the scenario is determined, such as the reproduction of key historical events or the role-playing of important figures, and the user's interaction needs in the scenario are planned in detail, such as obtaining more information by clicking or participating in simulated decision-making.

[0044] Based on historical knowledge points and historical task materials, the logical framework and theme style of the virtual scene are planned through semantic association analysis; It should be explained that based on historical knowledge points and historical task materials, the logical framework and theme style of the virtual scene are planned through semantic association analysis, and semantic association technology is used to conduct in-depth analysis of a large amount of historical data to find out the internal connections between different knowledge points. For example, the policy background of a certain period is connected with the important events of the time to construct a logically clear content context; on this basis, the theme style of the virtual scene is determined, such as historical review or policy interpretation from a modern perspective, to ensure that the content is both educational and attractive.

[0045] Based on the logical framework and theme style of the virtual scene, use 3D modeling tools to build a virtual scene template.

[0046] It should be explained that, based on the logical framework and theme style of the virtual scene, 3D modeling tools are used to build virtual scene templates, and appropriate 3D modeling software is selected to design and create virtual environment templates according to the previously determined logical framework and theme style. For example, a conference room or battlefield of a certain historical period can be reproduced, accurate to every detail such as the placement of tables and chairs, wall decorations, etc.; the task execution module executes situational decision-making simulation tasks according to the task requirements in the virtual scene, records the user's decision-making behavior and interaction data in real time during the execution process, and generates a preliminary feedback report based on the user's decision-making behavior and interaction data.

[0047] Users enter a virtual scene through VR devices, set multiple situational decision points according to task requirements, and guide users through virtual characters, visual prompts, and voice prompts; It should be noted that after the user enters the virtual scene through the VR device, multiple situational decision points are set at specific locations according to the preset task requirements. For example, in a scene that recreates a historical meeting, when the user approaches the moment of discussing a key policy, a decision point will be triggered, requiring the user to choose to support or oppose the policy. In this process, virtual characters are used to simulate real people to ask questions or express opinions, and visual prompts such as text information or indicator arrows on the screen, as well as voice prompts such as background explanations or character dialogues, are used to provide users with detailed background knowledge and operation guidance.

[0048] According to the guidance, users perform decision-making tasks in virtual scenes through interactive devices, and the user's decision-making behavior and interaction data are recorded in real time during the execution process; It should be noted that decision-making behavior includes the choices made by users in virtual scenes for specific situations, the actions taken, and their logical consistency. Interaction data includes all interaction records between users and virtual environments, such as clicks, gesture controls, dwell time, and communication content with other virtual characters.

[0049] After the user completes each decision, the rule engine analyzes the user's selection correctness, decision logic and behavior patterns in real time based on the recorded decision behavior and interaction data, and generates a preliminary feedback report.

[0050] It should be noted that once the user completes the task at a decision point, the rule engine is immediately started for in-depth analysis based on the recorded decision-making behavior and interaction data. For example, the correctness of the user's choice is evaluated by comparing the user's support for a proposal at the discussion meeting with the standard answer (if it is for educational purposes) or best practice cases; further, the logical consistency of the user in continuous decision-making is analyzed, such as whether the same position is maintained or whether there is a reasonable reason for change; at the same time, the user's behavior pattern is examined, such as whether he actively explores different options or tends to make decisions quickly. Finally, a preliminary feedback report is generated based on the above analysis results.

[0051] The preliminary feedback report includes an assessment of the correctness of the user's decision, analysis of the decision logic, evaluation of behavior patterns, and personalized improvement suggestions and learning resource recommendations based on the above analysis results.

[0052] The execution evaluation module uses data mining to extract task evaluation indicators based on the preliminary feedback report, and uses the hierarchical analysis method based on the task evaluation indicators to evaluate the user's task execution. According to the evaluation results, the scenario-based decision-making simulation tasks in the virtual scene are dynamically adjusted.

[0053] Based on the preliminary feedback report, data mining is used to conduct in-depth analysis of the preliminary feedback report to extract task evaluation indicators; It should be noted that based on the preliminary feedback report, data mining technology is used to conduct in-depth analysis of the data in the report, first identifying and extracting indicators closely related to task completion, such as user decision accuracy, response time, path selection frequency, etc.; cluster analysis is used to distinguish different types of user behavior patterns, and regression analysis is used to determine which factors most significantly affect task performance, so as to extract key task evaluation indicators from a large amount of feedback data; Based on task evaluation indicators, asymmetric interaction effects and dynamic complexity adjustment mechanisms are used to construct a multidimensional relationship model; Based on the multidimensional relationship model, the interaction between dimensions is analyzed, and nonlinear functions are introduced to characterize the complex relationship between selection correctness, decision logic and behavior patterns. At the same time, a penalty mechanism is used to process the absolute value of the differences between task evaluation indicators. Based on the task evaluation indicators, the hierarchical analysis method is used to evaluate the user's task execution status T, which is expressed as: Among them, P represents the accuracy of the user when performing the scenario-based decision-making simulation task, L represents the rationality of the user's decision logic, D represents the standardization of the user's behavior pattern, k represents the nonlinear contribution of the user's behavior pattern, δ represents the task complexity factor, ∈ is a small integer, and T is the user's task execution score.

[0054] Furthermore, selection correctness refers to whether the user's choice of the target option or answer is correct when completing the task, reflecting the user's understanding and judgment ability of the task goal. For example, in a learning system, when a user answers a question, if he chooses the correct answer from four options, the selection correctness is 100%; if the answer is wrong, the selection correctness will decrease.

[0055] Decision logic refers to whether the operation steps performed by the user in the process of completing the task are coherent and reasonable, and whether they conform to the task rules and expected paths. For example, in a simulated learning scenario, the user needs to unlock the knowledge point module first, then complete the question-answering task, and finally submit the summary. If the user completes the steps, the decision logic is reasonable; if the user skips some steps or the operation order is mixed up, the decision logic does not meet expectations.

[0056] Behavior patterns refer to the operating habits and interaction methods of users in the process of completing tasks, including click order, dwell time, and attention path. For example, in the system, a user is accustomed to reading the task requirements first, then browsing the knowledge points, completing the questions and submitting the summary, while another user may randomly click on different modules. These are all different manifestations of user behavior patterns.

[0057] It should be noted that, based on the multidimensional relationship model analysis, it is necessary to start from multiple dimensions and clarify the complex interactive relationship between the dimensions during the task execution process. For example, when the user completes the task, the different dimensions may not be simply linearly related, but have nonlinear and dynamic interactive characteristics, which requires the model to accurately characterize these relationships. The complex interactions between these dimensions are accurately characterized by introducing nonlinear functions, such as using specific mathematical expressions to describe how the probability of users making correct decisions in different scenarios changes with the changes in their logical rationality and behavioral norms; then, a penalty mechanism is used to process the absolute value of the differences between task evaluation indicators to ensure that any performance that deviates from the expected standard will be quantified and reflected in the final score. For example, when there is a gap between the user's actual performance and the ideal state, the gap will be magnified to emphasize the need for improvement; then, based on task evaluation indicators such as accuracy P, logical rationality L and behavioral pattern norms D, the hierarchical analysis method is used to calculate the user's comprehensive task execution score T.

[0058] It should also be noted that the rationality of the decision logic (L) is measured by analyzing the user's operation sequence during the task execution process, whether the task rules are followed, and the coherence of the completion path. Specifically, an ideal task execution path is set as a benchmark, and the actual operation process of the user is compared, such as whether the task is completed according to the logic of "first learning the knowledge points, then answering the questions, and finally submitting the summary". If the user's operation is completely in line with the expected path, the logical rationality can be given the maximum value; if there is skipping, repeated operations or logical confusion (such as jumping directly to the answering link without learning the knowledge points), the score will be reduced according to the degree of deviation. Finally, by quantitatively analyzing the behavioral records of the user's operation steps, the rationality is converted into a numerical value, for example, by calculating the similarity between the operation steps and the ideal path (such as a 0-1 standardized score) to determine the value of L.

[0059] The normativeness (D) of the behavior pattern is measured by the user's operating habits, click behavior, and stability of the behavior path during the task execution process. For example, analyze the user's stay time on the interface, whether the clicked area is concentrated in the key modules (such as task requirements or knowledge point content), and whether there are too many invalid clicks or long stagnation. If the user's behavior shows a high degree of concentration and the operation is concentrated on the task-related module, it can be considered that the normativeness of the behavior pattern is high; if the user frequently clicks on irrelevant areas or there are too many irrelevant operations, the normativeness is low. By recording the user's click behavior trajectory and operation log, combined with statistical indicators (such as invalid click ratio, dwell time distribution, etc.), the normativeness of the behavior pattern is converted into a numerical value. For example, the normativeness can be calculated by analyzing the proportion of clicks in the key area to the total number of clicks, and normalizing it to finally get the value of D.

[0060] Asymmetric interaction effects and dynamic complexity adjustment mechanisms are used to construct a multidimensional relationship model. The specific steps are as follows: selection correctness, decision logic and behavior pattern are defined as the core dimensions of the multidimensional relationship model, and correlation analysis is used to quantify the degree of correlation between core dimensions, and preliminarily identify the positive and negative correlations between core dimensions; It should be noted that the correctness of selection, decision logic and behavior pattern are defined as the core dimensions of the multidimensional relationship model, and the correlation analysis is used to quantify the degree of correlation between the core dimensions. The Pearson correlation coefficient between each pair of core dimensions is calculated to determine whether they are positively or negatively correlated. For example, in a specific task, the correlation coefficient between the correctness of selection and the decision logic is positive, indicating that there is a positive association between the two; while the correlation coefficient between the correctness of selection and the behavior pattern is negative, indicating that there may be a reverse influence.

[0061] Based on the positive and negative correlations among the core dimensions, nonlinear regression method is used to identify the asymmetric interaction effects among the core dimensions. It should be noted that first, a nonlinear regression model containing core dimensions and their interaction terms is set. For example, when analyzing the relationship between choice correctness, decision logic and behavior patterns, the product term of choice correctness and decision logic and their respective square terms are introduced to capture possible nonlinear effects. Then, the model is fitted using historical data, and the model parameters are optimized by minimizing the error between the predicted value and the actual value. In this process, it may be found that when the user's choice correctness is improved to a certain threshold, the improvement speed of their decision logic is significantly accelerated, while this acceleration effect is not obvious at a lower level of choice correctness, indicating the existence of an asymmetric interaction effect. Based on the preliminary feedback reports, principal component analysis was used to extract task complexity factors; It should be noted that all task evaluation indicators are first standardized, then the covariance matrix is ​​constructed and its eigenvalues ​​and eigenvectors are calculated, and the eigenvectors corresponding to the first few largest eigenvalues ​​are selected as the main components, which represent the main variation direction of the data. For example, when analyzing the performance of users in multiple tasks, one or more factors reflecting the overall complexity of the task can be extracted through principal component analysis.

[0062] Based on the task complexity factor, the weight distribution of the core dimensions is dynamically adjusted through a nonlinear mapping function; It should be noted that, first, the overall difficulty level of the current task is determined based on the task complexity factor extracted by principal component analysis, for example, three levels of simple, medium and complex are distinguished; then, an initial weight is set for each core dimension (selection correctness, decision logic, behavior pattern), and a nonlinear mapping function based on the task complexity factor is introduced to adjust these weights. This function, such as the sigmoid function or the tanh function, can smoothly change the weight distribution ratio according to different values ​​of the task complexity factor; specifically, when dealing with high-complexity tasks, if it is found that increasing the weight of the behavior pattern dimension can more effectively improve user performance, the weight of the behavior pattern is increased through the mapping function, while the weight of the selection correctness and decision logic is appropriately reduced; on the contrary, in low-complexity tasks, it may be necessary to emphasize the weight of the selection correctness to ensure the consolidation of basic skills; for example, in the simulation of a complex team collaboration task, as the task complexity factor increases, the nonlinear mapping function will automatically increase the weight of the behavior pattern and decision logic, and reduce the dependence on the selection correctness, thereby ensuring that in task environments of different complexities, the evaluation can more accurately reflect the actual ability and adaptability of the user.

[0063] Combining the asymmetric interaction effects among core dimensions and the weight distribution of core dimensions, a multidimensional relationship model is finally constructed.

[0064] Based on the user's historical task execution scores, define a low score threshold T1 and a high score threshold T2; When T≤T1, the task complexity is reduced and the guidance frequency is increased; Furthermore, for example, when a user first comes into contact with a virtual learning module, the content of the decision-making simulation task is simplified, and prompts and guidance are frequently provided to help the user become familiar with the operating procedures and background knowledge more quickly.

[0065] When T1<T<T2, the task complexity and guidance frequency are maintained; Furthermore, for example, when the user already has a certain foundation and can skillfully complete tasks of the current difficulty, the complexity of the existing task is kept unchanged, while continuing to provide appropriate prompts and support at the same frequency to ensure the user's continued participation and steady progress.

[0066] When T≥T2, the task complexity is increased and the guidance frequency is reduced.

[0067] Furthermore, for users who demonstrate high proficiency and confidence, the difficulty of the task can be increased in subsequent tasks, such as introducing more variables or higher-level decision points, and the number and frequency of prompts can be gradually reduced to encourage users to think and solve problems independently.

[0068] In summary, the present invention achieves a highly customized learning experience and greatly improves user participation and educational effects through: a scenario construction module that dynamically generates personalized virtual scenarios according to user roles and task requirements; at the same time, an execution evaluation module is used to conduct a multi-dimensional in-depth analysis of users' decision-making behaviors and interaction data, providing accurate feedback and guidance, which not only promotes the improvement of personal capabilities, but also optimizes the design and adjustment of subsequent tasks, thereby overcoming the shortcomings of the prior art in content flexibility and user task execution effect evaluation, and significantly enhancing the practicality and teaching effectiveness of the digital management system.

[0069] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A digital management system based on VR technology, characterized by: include, The data collection module collects user information through VR devices, pre-processes the user information, and identifies user roles and task requirements based on the pre-processed user information; The scene construction module uses PCG to extract knowledge points, scene templates, and task materials from the content database according to the user role and task requirements, dynamically generates a virtual scene that matches the user role, and loads the virtual scene into the user's VR interface; The task execution module executes the scenario-based decision-making simulation task according to the task requirements in the virtual scene, records the user's decision-making behavior and interaction data in real time during the execution process, and generates a preliminary feedback report based on the user's decision-making behavior and interaction data; The execution evaluation module uses data mining to extract task evaluation indicators based on the preliminary feedback report, and uses the hierarchical analysis method based on the task evaluation indicators to evaluate the user's task execution. According to the evaluation results, the scenario-based decision-making simulation tasks in the virtual scene are dynamically adjusted.

2. The digital management system based on VR technology as claimed in claim 1, characterized in that: The user information includes the user's identity information, behavior data, decision data and interaction data.

3. The digital management system based on VR technology as claimed in claim 2, characterized in that: The specific steps of preprocessing the user information are as follows: Identify and remove outliers in user information through the Z-score method; Mean filling is used to fill missing values ​​in user information; Filter the user information for noise through a low-pass filter; The label mapping method is used to standardize user information.

4. The digital management system based on VR technology as claimed in claim 3, characterized in that: The specific steps of identifying user roles and task requirements based on pre-processed user information are as follows: Based on the pre-processed user information, the user's identity features are extracted by direct mapping, the user's behavior features are extracted by statistical analysis, the user's interaction features are extracted by log analysis, and the user's decision features are extracted by frequency statistics; based on the user's identity features, behavior features, interaction features, and decision features, feature classification is used to classify and organize them to generate a structured feature set; Based on the structured feature set, define the rule logic and classification conditions, and finally build a rule engine. Use the rule engine to analyze users one by one and identify user roles. Based on the collection of structured features, the behavior trajectory analysis method is used to obtain the user's operation path, residence time and action sequence in the VR scene, and the frequency statistics method is combined to analyze the user's task selection preference, and finally identify the task requirements.

5. The digital management system based on VR technology as claimed in claim 4, characterized in that: According to the user role and task requirements, PCG is used to extract knowledge points, scene templates and task materials from the content database, dynamically generate a virtual scene that matches the user role, and load the virtual scene into the user's VR interface. The specific steps are as follows: Based on user roles and task requirements, parameterized queries are used to extract knowledge points, virtual scene templates, and task materials from the content database; The extracted knowledge points, virtual scene templates and task materials are input into the PCG algorithm to dynamically adjust the position, size and color of the elements in the virtual scene to make them conform to the user role's interaction habits and task requirements, and predict the virtual scene scores S of different combinations. i ; Select Virtual Scene Score i The highest combination is used as a virtual scene and loaded into the user's VR interface through a VR interface.

6. The digital management system based on VR technology as claimed in claim 5, characterized in that: The knowledge points include theoretical knowledge, policies and regulations, historical events, important speeches and practical cases; The task materials include text descriptions, pictures, audio and video resources and interactive content related to the task requirements; The construction process of the virtual scene template includes the following steps: Define scenario goals, scenario subjects, and interaction requirements based on user roles and task requirements; Based on historical knowledge points and historical task materials, the logical framework and theme style of the virtual scene are planned through semantic association analysis; Based on the logical framework and theme style of the virtual scene, use 3D modeling tools to build a virtual scene template.

7. The digital management system based on VR technology as claimed in claim 5, characterized in that: According to the task requirements in the virtual scene, the scenario-based decision-making simulation task is executed, the user's decision-making behavior and interaction data are recorded in real time during the execution process, and a preliminary feedback report is generated according to the user's decision-making behavior and interaction data. The specific steps are as follows: Users enter a virtual scene through VR devices, set multiple situational decision points according to task requirements, and guide users through virtual characters, visual prompts, and voice prompts; According to the guidance, users perform decision-making tasks in virtual scenes through interactive devices, and the user's decision-making behavior and interaction data are recorded in real time during the execution process; After the user completes each decision, the rule engine analyzes the user's selection correctness, decision logic and behavior patterns in real time based on the recorded decision behavior and interaction data, and generates a preliminary feedback report.

8. The digital management system based on VR technology as claimed in claim 7, characterized in that: Based on the preliminary feedback report, data mining is used to extract task evaluation indicators, and the hierarchical analysis method is used to evaluate the user's task execution based on the task evaluation indicators. The specific steps are as follows: Based on the preliminary feedback report, data mining is used to conduct in-depth analysis of the preliminary feedback report and extract task evaluation indicators; Based on task evaluation indicators, asymmetric interaction effects and dynamic complexity adjustment mechanisms are used to construct a multidimensional relationship model; The interactive relationship between dimensions is analyzed based on the multidimensional relationship model, and a nonlinear function is introduced to characterize the complex relationship between selection correctness, decision logic and behavior pattern. At the same time, a penalty mechanism is used to process the absolute value of the differences between task evaluation indicators. Based on the task evaluation indicators, the hierarchical analysis method is used to evaluate the user's task execution status T.

9. The digital management system based on VR technology as claimed in claim 8, characterized in that: The asymmetric interaction effect and dynamic complexity adjustment mechanism are used to construct a multidimensional relationship model. The specific steps are as follows: Define selection correctness, decision logic and behavior patterns as the core dimensions of the multidimensional relationship model, and use correlation analysis to quantify the degree of correlation between core dimensions and preliminarily identify the positive and negative correlations between core dimensions; Based on the positive and negative correlations among the core dimensions, nonlinear regression method is used to identify the asymmetric interaction effects among the core dimensions. Based on the preliminary feedback reports, principal component analysis was used to extract task complexity factors; Based on the task complexity factor, the weight distribution of the core dimensions is dynamically adjusted through a nonlinear mapping function; Combining the asymmetric interaction effects among core dimensions and the weight distribution of core dimensions, a multidimensional relationship model is finally constructed.

10. The digital management system based on VR technology as claimed in claim 9, characterized in that: The specific steps of dynamically adjusting the scenario-based decision-making simulation task in the virtual scene according to the evaluation results are as follows: Based on the user's historical task execution scores, define a low score threshold T1 and a high score threshold T2; When T≤T1, the task complexity is reduced and the guidance frequency is increased; When T1<T<T2, the task complexity and guidance frequency are maintained; When T≥T2, the task complexity is increased and the guidance frequency is reduced.

Images