Label-based educational data dynamic analysis method and system

Abstract

The invention discloses a label-based education data dynamic analysis method and system, and belongs to the technical field of big data analysis and education business system optimization, and the method comprises the steps: S1, constructing a multi-dimensional label system, and creating an extensible tree-shaped label group structure; the tree-shaped label group comprises administrative level labels and business labels, and the administrative level labels at least comprise a provincial level, a municipal level, a district and county level and a school level; s2, performing data fusion and dynamic calculation, and configuring a dynamic label actuator driven by a threshold value; s3, performing intelligent statistical display, and performing dynamic labeling on the original data through an assembly line processing engine; through the analysis model, generating a multi-dimensional cross analysis result based on the label hierarchical relationship; and S4, dynamic feedback, including real-time early warning and developability evaluation report. According to the method, the statistical requirements of various levels of organizations under multiple dimensions are supported, and the flexibility of data query and statistics is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of big data analysis and education business system optimization, and specifically to a tag-based dynamic analysis method and system for education data. Background Art

[0002] Currently, the comprehensive quality evaluation system is available to all students in primary, middle, and high schools across the province. It has a massive user base, and the scale of student growth data reporting is also extremely large. Against this backdrop, traditional data statistics models have exposed significant limitations and are unable to meet the needs for efficient and rapid statistics and analysis of massive amounts of student business data. Current education data management systems generally suffer from two major flaws:

[0003] First, the statistical dimensions are rigid and have poor scalability. Each analysis requirement (such as the province-wide "dropout warning" being expanded to support only private schools, and the "pass rate of reporting indicators" in each district and county) corresponds to an independently developed functional module. New analysis requirements require the redevelopment of functional modules, which is mainly due to the rigid architecture of traditional system design.

[0004] Second, cross-level data aggregation is inefficient. Predefined analysis dimensions (such as statistical scores by class or subject) cannot be flexibly expanded, and adding new dimensions (such as grouping by school type) requires modifying the database schema and code, which cannot dynamically adapt to changes in the administrative structure.

[0005] Taking a provincial education cloud platform as an example, compiling statistics on the distribution of teacher professional titles requires manually linking data from six independent systems, with an average response time of over 48 hours. While some existing patents cover educational data analysis, they fail to address the issue of multi-level dynamic statistics. Current data queries use fixed tags and lack flexible configuration capabilities. During the code development phase, developers must predetermine the tag settings for business tables and then strictly adhere to established rules for coding. In this model, during the code implementation process, developers must not only constantly monitor the tagging rules applicable to specific business data but also fully implement tagging operations through code. However, during the actual development process, if it is discovered that the existing tagging rules no longer meet business needs, the existing coding logic must be overturned and development work must be restarted. This approach severely lacks flexibility, significantly impacting development efficiency and the smooth progress of the project. Summary of the Invention

[0006] The technical task of the present invention is to address the above shortcomings and provide a label-based dynamic analysis method and system for educational data, which supports the statistical needs of organizations at all levels in multiple dimensions and significantly improves the flexibility of data query and statistics.

[0007] The technical solution adopted by the present invention to solve its technical problem is:

[0008] A tag-based dynamic analysis method for educational data, comprising:

[0009] Step S1: Build a multi-dimensional tag system and create an extensible tree-shaped tag group structure; the tree-shaped tag group includes administrative level tags and business tags, where the administrative level tags include at least four levels: provincial level, municipal level, district level, and school level;

[0010] Step S2: data fusion and dynamic calculation, configuring a threshold-driven dynamic tag executor;

[0011] Step S3: Intelligent statistical display, dynamic labeling of raw data is performed through the pipeline processing engine; through the analysis model, multi-dimensional cross-analysis results are generated based on the label hierarchical relationship;

[0012] Step S4: dynamic feedback, including real-time warning and developmental evaluation reports.

[0013] The present invention builds a flexible and scalable tag management and data statistical analysis system by building a tag middle platform, implementing metadata drive, introducing a behavior tracking system, and establishing a data asset catalog. It supports the statistical needs of organizations at all levels in multiple dimensions and significantly improves the flexibility of data query and statistics. Establish a unified tag management platform to support dynamic tag definition and calculation; implement statistical dimension expansion by configuring metadata instead of modifying the code; perform structured labeling on the entire learning process (such as marking video pause behavior as a "difficulty in understanding" signal); label and catalog existing data assets to improve reuse rate; by systematically solving these problems, the education data system can evolve from "descriptive statistics" to "diagnostic analysis" and "predictive intervention", truly unleashing the potential of smart education.

[0014] Furthermore, the tree-shaped tag group specifically includes:

[0015] Administrative dimension label group: nested in four levels: province → city → district → county → school;

[0016] Business dimension tag group: including three categories: student development, teacher growth, and teaching management;

[0017] Composite label group: Cartesian product combination of administrative dimension and business dimension.

[0018] Furthermore, the step S1 constructs tag grouping, supports the creation of tag groups of various dimensions, and supports the maintenance of various business tags under each tag group, including:

[0019] Academic ability labels, including subject knowledge mastery (such as "mathematics and geometry - proficient"), higher-order thinking (critical thinking, innovative thinking scores);

[0020] Behavioral development labels, including learning engagement (video playback rate, homework time), collaboration ability (group project contribution value);

[0021] Psychological quality labels, including growth mindset (psychological assessment results) and stress tolerance (speed of recovery from exam mistakes);

[0022] Practice innovation label, including scientific research results (patents / papers), social practice (volunteer service hours, project leadership);

[0023] Construct label groups and use a label generation mechanism that combines dynamic and manual methods, including:

[0024] Rule engine: Configure 300+ basic rules (e.g., "A+ for three consecutive operations indicates high learning stability");

[0025] Manual maintenance: supports administrators to maintain various data labels according to various maintenance rules including educational authorities, school types, and student types.

[0026] Furthermore, the data fusion is to integrate multi-source heterogeneous data, including:

[0027] Structured data, including transcripts, attendance records, physical test data, etc.

[0028] Semi-structured data, including electronic portfolios, project reports, etc.;

[0029] Unstructured data, including classroom videos (behavior recognition) and forum discussions (sentiment analysis);

[0030] Threshold condition settings include:

[0031] Exact matching conditions (e.g., teacher title = "senior teacher");

[0032] Dynamic calculation conditions (e.g., student mobility > 20% of the district or county average);

[0033] Associated matching conditions (such as the admission rate of the head teacher's class meets the standard).

[0034] Furthermore, the step S3 includes:

[0035] (1) Growth trajectory tracking:

[0036] Time series analysis: Use the DTW algorithm to match similar growth patterns and identify deviations from the trajectory;

[0037] Milestone prediction: predicting the probability of achieving excellence based on a survival analysis model (e.g., “the probability of achieving programming proficiency within 3 months is 78%”);

[0038] (2) Group comparison analysis:

[0039] Clustering model: Discover hidden student groups (such as the "high-potential unstimulated group") based on the OPTICS algorithm;

[0040] Attribution of differences: Applying the causal forest model to identify key factors influencing group differences;

[0041] (3) Personalized diagnostic engine:

[0042] Knowledge graph reasoning: Identify weaknesses in capabilities and related knowledge points (e.g., “weak function application → impacts physical mechanics analysis”);

[0043] Intervention simulation: Verify the effectiveness of recommended measures through counterfactual analysis (e.g., “Increasing lab hours by 20% can improve comprehension by 35%”).

[0044] Furthermore, the multidimensional cross-analysis results support:

[0045] Hierarchical drill-down analysis, including drilling down from provincial summaries to individual student records;

[0046] Cross-label group correlation analysis, including the correlation between teachers' academic qualifications and the grades of the classes they teach;

[0047] Time series comparative analysis, including changes in district and county education resource investment over the past three years;

[0048] The statistical result output module includes:

[0049] Standardized data interface (JSON / XML format);

[0050] Interactive visualization component (supports 12 types of charts including heat maps and Sankey diagrams);

[0051] Customized report generator (automatically formatted according to educational institution templates).

[0052] Furthermore, the real-time warning includes multi-level warning and graded intervention. The multi-level warning rules are as follows:

[0053] Primary warning: homework submission is delayed more than 3 times;

[0054] Intermediate warning: Knowledge mastery has declined for two consecutive weeks;

[0055] Advanced warning: psychological stress index > threshold and social interaction decreases sharply;

[0056] The graded intervention automatically triggers plans including push notifications for remedial exercises and reminders for teacher home visits;

[0057] Set up a label quality monitoring module to trigger an alert when the following situations occur:

[0058] (1) The coverage of a single tag is lower than the set threshold (e.g. <5%);

[0059] (2) The deviation of cross-level label statistics exceeds 20%;

[0060] (3) Data source field changes lead to label invalidation;

[0061] The developmental evaluation report shall include:

[0062] Three-dimensional radar chart: shows the dynamic balance of academic / practical / psychological aspects;

[0063] Growth motivation analysis: visualize the ratio of intrinsic motivation (interest-driven) to extrinsic motivation (pressure to enter higher education);

[0064] Adaptive recommendations: Generate personalized development plans (e.g., “It is recommended to participate in robotics competitions to strengthen engineering thinking”).

[0065] The present invention also claims protection for a tag-based dynamic analysis system for educational data, comprising:

[0066] Multi-dimensional tag system building module, used to create an extensible tree-shaped tag group structure;

[0067] The data fusion and dynamic calculation module is used to configure the tag executor associated with the data source, set the field mapping rules and threshold conditions, and realize the integration of multi-source heterogeneous data and the setting of threshold conditions;

[0068] Intelligent statistical display module, used to generate multi-dimensional cross-analysis results based on the label hierarchical relationship through analysis models;

[0069] Dynamic feedback module, used to achieve real-time early warning and generate developmental evaluation reports;

[0070] The system specifically realizes dynamic analysis of educational data through the above-mentioned method.

[0071] The present invention also claims a tag-based dynamic analysis device for educational data, comprising: at least one memory and at least one processor;

[0072] The at least one memory is configured to store a machine-readable program;

[0073] The at least one processor is configured to call the machine-readable program to implement the above method.

[0074] The present invention also claims protection for a computer-readable medium having computer instructions stored thereon, which are capable of implementing the above method when executed by a processor.

[0075] Compared with the prior art, the tag-based dynamic analysis method and system of educational data of the present invention has the following beneficial effects:

[0076] The present invention significantly improves the performance and user experience of the education data management system. The tag-based dynamic analysis method of education data can achieve the following significant technical effects by constructing a flexible and intelligent tag management system.

[0077] 1. Improved data processing efficiency: This invention establishes a standardized labeling system to categorize and label educational data, enabling rapid data retrieval and classification. Compared to traditional educational data processing methods, this avoids the disordered storage and repeated processing of large amounts of data, significantly shortens data preprocessing time, and improves educational data processing efficiency. It can quickly analyze massive amounts of educational data, providing timely data support for educational decision-making.

[0078] 2. Optimizing Data Analysis Accuracy: A tag-based dynamic analysis mechanism enables real-time adjustment of analysis dimensions and weights based on changing educational scenarios and needs. Through dynamic tag updates and correlation analysis, key information and underlying patterns in educational data can be accurately captured, avoiding information omissions or misjudgments caused by static analysis models. This improves the accuracy and reliability of educational data analysis results, providing a more precise basis for developing personalized education plans and evaluating teaching quality.

[0079] 3. Enhanced Data Adaptability: This invention is compatible with educational data from various sources and formats. Whether it's structured performance data or unstructured instructional videos or student feedback text, all can be uniformly managed and analyzed through the tagging system. Furthermore, in response to the continuous emergence of new data types and analysis needs in the education field, the tagging system has good scalability and can quickly incorporate new tag categories, making this invention more adaptable and versatile, and widely applicable to various educational scenarios.

[0080] 4. Holographic Data Representation: This system maps heterogeneous data such as student behavior, academic performance, and psychological assessments to a standardized labeling system, creating a 360-degree digital profile of students and addressing traditional data silos. Technical Indicators: The average time required to add district and county-level statistical dimensions has been reduced from 72 hours to 4 hours. Data utilization has increased from 30% in traditional systems to over 85%, supporting real-time correlation analysis across more than 10 data sources.

[0081] 5. Dynamic Feature Engineering: Automatically generates high-level composite labels (e.g., "critical thinking = logical reasoning score x project practice participation") based on a rule engine and AI model, overcoming the limitations of manually defined features. Application scenario: Identifying 12 implicit learning patterns (e.g., "nighttime efficiency" and "fragmented learning") overlooked by traditional methods.

[0082] 6. Supporting scientific education decision-making: This platform provides visual analysis results, transforming complex education data into intuitive and easy-to-understand charts and reports, enabling decision-makers to quickly grasp the current status and development trends of education. Furthermore, based on historical data and dynamic analysis models, it can predict future educational development, assisting decision-makers in formulating scientific and rational policies and plans, and improving the scientific nature and foresight of education decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] Figure 1 It is a diagram illustrating a multi-dimensional knowledge base management method associated with smart tags provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0084] An embodiment of the present invention provides a tag-based dynamic analysis method for educational data, comprising:

[0085] Step S1: Build a multi-dimensional tag system and create an extensible tree-shaped tag group structure; the tree-shaped tag group includes administrative level tags and business tags, where the administrative level tags include at least four levels: provincial level, municipal level, district level, and school level;

[0086] Step S2: Data fusion and dynamic calculation, configuring the tag executor associated with the data source, setting the field mapping rules and threshold conditions;

[0087] Step S3: Intelligent statistical display, dynamic labeling of raw data is performed through the pipeline processing engine; through the analysis model, multi-dimensional cross-analysis results are generated based on the label hierarchical relationship;

[0088] Step S4: dynamic feedback, including real-time warning and developmental evaluation reports.

[0089] The tree-shaped tag group specifically includes:

[0090] Administrative dimension label group: nested in four levels: province → city → district → county → school;

[0091] Business dimension tag group: including three categories: student development, teacher growth, and teaching management;

[0092] Composite label group: Cartesian product combination of administrative dimension and business dimension.

[0093] The threshold condition setting includes:

[0094] Exact matching conditions (e.g., teacher title = "senior teacher");

[0095] Dynamic calculation conditions (e.g., student mobility > 20% of the district or county average);

[0096] Associated matching conditions (such as the admission rate of the head teacher's class meets the standard).

[0097] The multidimensional cross-analysis results support:

[0098] Hierarchical drill-down analysis, including drilling down from provincial summaries to individual student records;

[0099] Cross-label group correlation analysis, including the correlation between teachers' academic qualifications and the grades of the classes they teach;

[0100] Time series comparative analysis, including changes in education resource investment in districts and counties over the past three years.

[0101] Set up a label quality monitoring module to trigger an alert when the following situations occur:

[0102] (1) The coverage of a single tag is lower than the set threshold (e.g. <5%);

[0103] (2) The deviation of cross-level label statistics exceeds 20%;

[0104] (3) Data source field changes lead to label invalidation;

[0105] The statistical result output module includes:

[0106] Standardized data interface (JSON / XML format);

[0107] Interactive visualization component (supports 12 types of charts including heat maps and Sankey diagrams);

[0108] Customized report generator (automatically formatted according to educational institution templates).

[0109] This method builds a multidimensional labeling system, integrates multi-source educational data, and utilizes dynamic analysis techniques to achieve innovative real-time evaluation, trend prediction, and personalized intervention of educational data. Its core lies in breaking through the limitations of traditional statistical data analysis and constructing an intelligent analysis system that is "data-driven, dynamically iterative, and holographically aware." The specific implementation of this method is as follows:

[0110] 1. Build a multi-dimensional tag system and create an extensible tree-shaped tag group structure. This includes:

[0111] 1. Classification of holographic labels.

[0112] Build tag groups, support the creation of tag groups of various dimensions, and support the maintenance of various business tags under each tag group, such as:

[0113] (1) Academic ability labels: subject knowledge mastery (e.g., “mathematics and geometry – proficiency”), higher-order thinking (critical thinking, innovative thinking scores);

[0114] (2) Behavioral development labels: learning engagement (video playback rate, homework time), collaboration ability (group project contribution value);

[0115] (3) Psychological quality labels: growth mindset (psychological assessment results), stress tolerance (speed of recovery from exam mistakes);

[0116] (4) Practical innovation labels: scientific research results (patents / papers), social practice (volunteer service hours, project leadership).

[0117] 2. A label generation mechanism that combines dynamic and manual methods.

[0118] (1) Rule engine: Configure 300+ basic rules (e.g., “A+ for three consecutive operations indicates high learning stability”).

[0119] (2) Manual maintenance: supports administrators to maintain various data labels according to educational authorities, school types, student types, etc.

[0120] 2. Data fusion and dynamic calculation, configure threshold-driven dynamic label executor.

[0121] Multi-source heterogeneous data integration:

[0122] Structured data: transcripts, attendance records, physical test data, etc.

[0123] Semi-structured data: electronic portfolios, project reports, etc.;

[0124] Unstructured data: classroom videos (behavior recognition), forum discussions (sentiment analysis), etc.

[0125] 3. Intelligent statistical display, analysis model generates multi-dimensional cross-analysis results.

[0126] (1) Growth trajectory tracking:

[0127] Time series analysis: Use the DTW algorithm to match similar growth patterns and identify deviations from the trajectory;

[0128] Milestone prediction: Predict the probability of achieving excellence based on a survival analysis model (e.g., “the probability of achieving programming proficiency within 3 months is 78%”).

[0129] (2) Group comparison analysis:

[0130] Clustering model: OPTICS algorithm discovers hidden student groups (such as the "high potential unmotivated group")

[0131] Attribution of differences: Using the causal forest model to identify key factors influencing group differences

[0132] (3) Personalized diagnostic engine:

[0133] Knowledge graph reasoning: Identify weaknesses in capabilities and related knowledge points (e.g., “weak function application → impacts physical mechanics analysis”);

[0134] Intervention simulation: Verify the effectiveness of recommended measures through counterfactual analysis (e.g., “Increasing lab hours by 20% can improve comprehension by 35%”).

[0135] 4. Dynamic feedback mechanism.

[0136] 1. Real-time early warning system.

[0137] (1) Multi-level warning rules, including:

[0138] Primary warning: homework submission is delayed more than 3 times;

[0139] Intermediate warning: Knowledge mastery has declined for two consecutive weeks;

[0140] Advanced warning: psychological stress index > threshold and social interaction decreases sharply;

[0141] (2) Graded intervention: automatically triggering 10 types of plans, including push notifications for remedial exercises and reminders for teacher home visits.

[0142] 2. Developmental evaluation report.

[0143] (1) Three-dimensional radar chart: showing the dynamic balance of academics, practice and psychology.

[0144] (2) Growth motivation analysis: Visualize the ratio of intrinsic motivation (interest-driven) to extrinsic motivation (pressure to enter higher education).

[0145] (3) Adaptive suggestions: Generate personalized development plans (e.g., “It is recommended to participate in robotics competitions to strengthen engineering thinking”).

[0146] In smart education systems, traditional unlabeled or fixed-label query statistical data has many problems and drawbacks, which directly affect the value mining and application of educational data. The main drawbacks and detailed analysis are as follows:

[0147] (1) The data dimensions are rigid and the analysis perspective is single.

[0148] Predefined dimension limits:

[0149] Problem: Only fixed-dimensional statistics are supported (such as by class, subject, and time), and flexible combination analysis is not possible.

[0150] Example: It is impossible to simultaneously analyze cross-dimensional requirements such as "the performance of girls using tablets in science lab classes."

[0151] Consequences: Hidden association patterns (such as the correlation between device type and subject performance) are difficult to discover.

[0152] Static indicator system:

[0153] Performance: The indicator calculation method is hard-coded in the system (e.g. "pass rate = number of passed students / total number of students").

[0154] Limitations: The system needs to be redeveloped when derivative indicators such as "progress of left-behind children" need to be calculated.

[0155] (2) Data value density is low.

[0156] Unstructured data waste:

[0157] Problem: Unstructured data such as video viewing behavior and forum discussions cannot be effectively quantified.

[0158] Data: 90% of the behavioral data of an online education platform was not included in the analysis due to lack of labels.

[0159] Deep feature loss:

[0160] Comparison between traditional statistics and labeling analysis:

[0161] Traditional statistics: only record the "correct answer rate";

[0162] Labeled analysis: Statistics can be used to calculate "problem-solving time distribution → cognitive load index" and "error type → knowledge weaknesses".

[0163] (3) Inefficient response to demand.

[0164] Long development cycle:

[0165] Process: Business requirements → Requirements review → Development → Testing → Launch (average time 3-6 weeks).

[0166] Cost: Adding a new analysis module to a county-level education cloud platform costs an average of 15 man-days.

[0167] The cost of change is high.

[0168] Case: When the Ministry of Education's new regulations require the addition of the "physical health standard compliance rate" indicator, the database structure needs to be modified and the original system functionality needs to be reconstructed.

[0169] The comparison between this method and traditional labeling statistics is as follows:

[0170] Table 1 Effect comparison

[0171] Comparison Dimension Traditional unlabeled statistics Dynamic labeling statistics in this embodiment Dimensional flexibility Fixed dimensions, modification requires development Support for ad-hoc dimension combinations Data utilization Can only analyze structured data Can mine unstructured data such as text and video Response speed Changing requirements takes weeks Business personnel can configure new indicators in real time Depth of analysis Surface polymerization results Drill down to individual behavior patterns System scalability New requirements lead to accumulation of technical debt By configuring the extension, the code remains basically unchanged

[0172] This dynamic analysis method innovatively employs specific tagging rules to manage statistical queries of business data. The system dynamically maintains tag types and business tags, supporting flexible and diverse filtering and statistical analysis operations within the tag system. This allows students to accurately query business data based on the user's selected tag type and business tag value, achieving efficient and accurate data statistics.

[0173] Taking into account the complexity and diversity of business scenarios in actual applications, and to ensure the adaptability and flexibility of the system, this method introduces the concept of dynamically configuring label rules, and by combining technologies such as rule engines, metadata management, real-time calculations, and front-end data display, builds a flexible and scalable label management and data statistical analysis system. In order to better meet the differentiated needs of various business scenarios, this method focuses on designing a set of highly differentiated and customized label rule systems. Within this rule framework, it is possible to accurately screen business data tables that need to be processed by labels, flexibly set label categories and business label values, and also provide a rich set of label setting methods. Subsequently, various related business data operations can be carried out smoothly, supporting the statistical needs of organizations at all levels under multiple dimensions, and significantly improving the flexibility of data query and statistics.

[0174] An embodiment of the present invention also provides a label-based education data dynamic analysis system, which specifically implements education data dynamic analysis through the label-based education data dynamic analysis method described in the above embodiment.

[0175] The system includes:

[0176] 1. A multi-dimensional tag system building module, used to create an extensible tree-shaped tag group structure. The tree-shaped tag group contains administrative level tags and business tags, where administrative level tags include at least four levels: provincial, municipal, district, and school. The multi-dimensional tag system building module includes:

[0177] (1) Holographic label classification: Construct label grouping, support the creation of label groups of various dimensions, and support the maintenance of various business labels under each label group, such as:

[0178] Academic ability labels: subject knowledge mastery (such as "mathematics and geometry - proficiency"), higher-order thinking (critical thinking, innovative thinking scores);

[0179] Behavioral development labels: learning engagement (video playback rate, homework time), collaboration ability (group project contribution value);

[0180] Psychological quality labels: growth mindset (psychological assessment results), stress tolerance (speed of recovery from exam mistakes);

[0181] Practice innovation tags: scientific research results (patents / papers), social practice (volunteer service hours, project leadership).

[0182] (2) Dynamic and manual label generation mechanism:

[0183] Rule Engine: Configure 300+ basic rules (e.g., "A+ for three consecutive operations indicates high learning stability").

[0184] Manual maintenance: supports administrators to maintain various data labels according to education authorities, school types, student types, etc.

[0185] 2. Data fusion and dynamic calculation module, used to configure the tag executor associated with the data source, set field mapping rules and threshold conditions, and realize multi-source heterogeneous data integration and threshold condition setting.

[0186] Multi-source heterogeneous data integration includes:

[0187] Structured data: transcripts, attendance records, physical test data, etc.

[0188] Semi-structured data: electronic portfolios, project reports, etc.;

[0189] Unstructured data: classroom videos (behavior recognition), forum discussions (sentiment analysis), etc.

[0190] The threshold condition setting includes:

[0191] Exact matching conditions (e.g., teacher title = "senior teacher");

[0192] Dynamic calculation conditions (e.g., student mobility > 20% of the district or county average);

[0193] Associated matching conditions (such as the admission rate of the head teacher's class meets the standard).

[0194] 3. Intelligent statistical display module, used to generate multi-dimensional cross-analysis results based on the label hierarchical relationship through the analysis model. It includes:

[0195] (1) Growth trajectory tracking:

[0196] Time series analysis: Use the DTW algorithm to match similar growth patterns and identify deviations from the trajectory;

[0197] Milestone prediction: Predict the probability of achieving excellence based on a survival analysis model (e.g., “the probability of achieving programming proficiency within 3 months is 78%”).

[0198] (2) Group comparison analysis:

[0199] Clustering model: OPTICS algorithm discovers hidden student groups (such as the "high potential unmotivated group")

[0200] Attribution of differences: Using the causal forest model to identify key factors influencing group differences

[0201] (3) Personalized diagnostic engine:

[0202] Knowledge graph reasoning: Identify weaknesses in capabilities and related knowledge points (e.g., “weak function application → impacts physical mechanics analysis”);

[0203] Intervention simulation: Verify the effectiveness of recommended measures through counterfactual analysis (e.g., “Increasing lab hours by 20% can improve comprehension by 35%”).

[0204] The multidimensional cross-analysis results support:

[0205] Hierarchical drill-down analysis, including drilling down from provincial summaries to individual student records;

[0206] Cross-label group correlation analysis, including the correlation between teachers' academic qualifications and the grades of the classes they teach;

[0207] Time series comparative analysis, including changes in education resource investment in districts and counties over the past three years.

[0208] The statistical result output module includes:

[0209] Standardized data interface (JSON / XML format);

[0210] Interactive visualization component (supports 12 types of charts including heat maps and Sankey diagrams);

[0211] Customized report generator (automatically formatted according to educational institution templates).

[0212] 4. Dynamic feedback module, used to achieve real-time early warning and generate developmental evaluation reports. Including:

[0213] (1) Real-time warning:

[0214] Multi-level warning rules, including:

[0215] Primary warning: homework submission is delayed more than 3 times;

[0216] Intermediate warning: Knowledge mastery has declined for two consecutive weeks;

[0217] Advanced warning: psychological stress index > threshold and social interaction decreases sharply;

[0218] Graded intervention: automatically triggers 10 types of plans, including remedial exercise push and teacher home visit reminders.

[0219] (2) Developmental evaluation report, including:

[0220] 3D radar chart: Shows the dynamic balance between academics, practice, and psychology.

[0221] Growth motivation analysis: Visualize the ratio of intrinsic motivation (interest-driven) to extrinsic motivation (pressure to enter higher education).

[0222] Adaptive recommendations: Generate personalized development plans (e.g., “It is recommended to participate in robotics competitions to strengthen engineering thinking”).

[0223] Set up a label quality monitoring module to trigger an alert when the following situations occur:

[0224] (1) The coverage of a single tag is lower than the set threshold (e.g. <5%);

[0225] (2) The deviation of cross-level label statistics exceeds 20%;

[0226] (3) Changes to data source fields cause labels to become invalid.

[0227] An embodiment of the present invention further provides a tag-based dynamic analysis device for educational data, comprising: at least one memory and at least one processor;

[0228] The at least one memory is configured to store a machine-readable program;

[0229] The at least one processor is used to call the machine-readable program to implement the tag-based dynamic analysis method of educational data described in the above embodiment.

[0230] An embodiment of the present invention further provides a computer-readable medium having computer instructions stored thereon, which, when executed by a processor, causes the processor to execute the tag-based dynamic analysis method for educational data described in the above embodiment. Specifically, a system or device equipped with a storage medium can be provided, on which software program code implementing the functions of any of the above embodiments is stored, and a computer (or CPU or MPU) of the system or device can be caused to read and execute the program code stored in the storage medium.

[0231] In this case, the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, and thus the program code and the storage medium storing the program code constitute part of the present invention.

[0232] Examples of storage media for providing program code include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), magnetic tapes, non-volatile memory cards, and ROMs. Alternatively, the program code can be downloaded from a server computer via a communication network.

[0233] In addition, it should be clear that the functions of any of the above embodiments can be achieved not only by executing the program code read by the computer, but also by enabling the operating system operating on the computer to complete part or all of the actual operations based on the instructions of the program code.

[0234] In addition, it can be understood that the program code read from the storage medium is written into the memory provided in the expansion board inserted into the computer or into the memory provided in the expansion unit connected to the computer, and then based on the instructions of the program code, the CPU installed on the expansion board or expansion unit is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above embodiments.

[0235] The present invention has been shown and described in detail above through the accompanying drawings and preferred embodiments. However, the present invention is not limited to these disclosed embodiments. Based on the above multiple embodiments, those skilled in the art can know that the code review methods in the above different embodiments can be combined to obtain more embodiments of the present invention, and these embodiments are also within the scope of protection of the present invention.

Claims

1. A tag-based dynamic analysis method for educational data, characterized in that: include: Step S1: Build a multi-dimensional tag system and create an extensible tree-shaped tag group structure; The tree-shaped label group includes administrative level labels and business labels, wherein the administrative level labels include at least four levels: provincial level, municipal level, district and county level, and school level; Step S2: data fusion and dynamic calculation, configuring a threshold-driven dynamic tag executor; Step S3: Intelligent statistical display, dynamic labeling of raw data is performed through the pipeline processing engine; through the analysis model, multi-dimensional cross-analysis results are generated based on the label hierarchical relationship; Step S4: dynamic feedback, including real-time warning and developmental evaluation reports.

2. The tag-based dynamic analysis method for educational data according to claim 1, characterized in that: The tree-shaped tag group specifically includes: Administrative dimension label group: nested in four levels: province → city → district → county → school; Business dimension tag group: including three categories: student development, teacher growth, and teaching management; Composite label group: Cartesian product combination of administrative dimension and business dimension.

3. A tag-based dynamic analysis method for educational data according to claim 1 or 2, characterized in that: Step S1 constructs tag grouping, supports the creation of tag groups of various dimensions, and supports the maintenance of various business tags under each tag group, including: Academic ability labels, including subject knowledge mastery and higher-order thinking; Behavioral development labels, including learning engagement and collaboration skills; Psychological quality labels, including growth mindset and stress tolerance; Practice innovation label, including scientific research results and social practice; The tag grouping is constructed by using a tag generation mechanism that combines dynamic and manual methods, including: Rule engine: configure 300+ basic rules; Manual maintenance: supports administrators to maintain various data labels according to various maintenance rules including educational authorities, school types, and student types.

4. The tag-based dynamic analysis method for educational data according to claim 1, characterized in that: The data fusion is to integrate multi-source heterogeneous data, including: Structured data, including transcripts, attendance records, and physical test data; Semi-structured data, including electronic portfolios and project reports; Unstructured data, including classroom videos and forum discussions; Threshold condition settings include: Exact matching conditions; Dynamic calculation conditions; Associate matching conditions.

5. The tag-based dynamic analysis method for educational data according to claim 1, characterized in that: The step S3 includes: (1) Growth trajectory tracking: Time series analysis: Use the DTW algorithm to match similar growth patterns and identify deviations from the trajectory; Milestone prediction: predicting the probability of achieving excellence based on the survival analysis model; (2) Group comparison analysis: Clustering model: Discover hidden student groups based on the OPTICS algorithm; Attribution of differences: Applying the causal forest model to identify key factors influencing group differences; (3) Personalized diagnostic engine: Knowledge graph reasoning: locate capability shortcomings and related knowledge points; Intervention simulation: Verifying the effectiveness of proposed measures through counterfactual analysis.

6. A tag-based dynamic analysis method for educational data according to claim 1 or 5, characterized in that: The multidimensional cross-analysis results support: Hierarchical drill-down analysis, including drilling down from provincial summaries to individual student records; Cross-label group correlation analysis, including the correlation between teachers' academic qualifications and the grades of the classes they teach; Time series comparative analysis, including changes in district and county education resource inputs; The statistical result output module includes: Standardized data interface; Interactive visualization components; Customizable report generator.

7. The tag-based dynamic analysis method for educational data according to claim 1, characterized in that: The real-time warning includes multi-level warning and graded intervention. The multi-level warning rules are as follows: Primary warning: homework submission is delayed more than 3 times; Intermediate warning: Knowledge mastery has declined for two consecutive weeks; Advanced warning: psychological stress index > threshold and social interaction decreases sharply; The graded intervention automatically triggers plans including push notifications for remedial exercises and reminders for teacher home visits; Set up a label quality monitoring module to trigger an alert when the following situations occur: (1) The coverage of a single tag is lower than the set threshold; (2) The deviation of cross-level label statistics exceeds 20%; (3) Data source field changes lead to label invalidation; The developmental evaluation report shall include: Three-dimensional radar chart: shows the dynamic balance of academic / practical / psychological aspects; Growth motivation analysis: visualize the ratio of intrinsic motivation to extrinsic motivation; Adaptive recommendations: Generate personalized development plans.

8. A tag-based education data dynamic analysis system, characterized in that: include: Multi-dimensional tag system building module, used to create an extensible tree-shaped tag group structure; The data fusion and dynamic calculation module is used to configure the tag executor associated with the data source, set the field mapping rules and threshold conditions, and realize the integration of multi-source heterogeneous data and the setting of threshold conditions; Intelligent statistical display module, used to generate multi-dimensional cross-analysis results based on the label hierarchical relationship through analysis models; Dynamic feedback module, used to achieve real-time early warning and generate developmental evaluation reports; The system specifically implements dynamic analysis of educational data through the method described in any one of claims 1 to 7.

9. A tag-based dynamic analysis device for educational data, characterized in that: include: at least one memory and at least one processor; The at least one memory is configured to store a machine-readable program; The at least one processor is configured to call the machine-readable program to implement the method according to any one of claims 1 to 7.

10. A computer-readable medium, characterized in that The computer readable medium stores computer instructions, which, when executed by a processor, can implement the method according to any one of claims 1 to 7.

Images