Labor education practice teaching effect evaluation method based on big data analysis

Abstract

The present application relates to computer and big data processing technical field, disclose a kind of labor education practical teaching effect evaluation method based on big data analysis.The method includes: collecting student multi-source heterogeneous data;Standardized cleaning and structured conversion;Build three-dimensional index system with labor participation, skill mastery and literacy development as core and calculate each dimension score;Through the multidimensional evaluation model of dynamic weight adjustment to generate comprehensive evaluation score;Based on bias analysis, automatically generate individual diagnosis report, class ability atlas and course optimization suggestions, and feedback to teaching planning module to realize closed-loop optimization.The present application realizes the objective, accurate, traceable quantitative evaluation of labor education effect, improves the scientific and intelligent level of teaching.

Description

Technical Field

[0001] This invention belongs to the field of computer and big data processing technology, specifically relating to a method for evaluating the effectiveness of labor education practice teaching based on big data analysis. Background Technology

[0002] With the deepening of education evaluation reform in the new era, labor education, as an important component of the comprehensive education system encompassing moral, intellectual, physical, aesthetic, and labor education, is receiving increasing attention for the scientific evaluation of its practical teaching effectiveness. Currently, the accelerated development of educational informatization and digital transformation has resulted in educational data exhibiting characteristics of multi-source heterogeneity, massive volume, and high dimensionality, providing new possibilities for constructing a precise and intelligent education evaluation mechanism. Because labor education emphasizes practicality, process, and comprehensiveness, its teaching effectiveness is difficult to effectively measure through traditional paper-and-pencil tests or single indicators. Therefore, it urgently needs to rely on modern information technology to achieve a comprehensive capture and analysis of the learning process and outcomes.

[0003] Educational assessment methods based on big data analytics are gradually becoming a research hotspot. Their core lies in integrating multi-dimensional data such as students' behavioral patterns, interaction records, outputs, and feedback information in real-world work scenarios. Through modeling and data mining, they reveal the inherent patterns of learning effectiveness. This type of method aims to break through the traditional assessment model, which is primarily based on subjective impressions or periodic summaries, and shift towards an evidence-based, data-driven dynamic evaluation paradigm. This allows for a more objective reflection of students' depth of work participation, skill acquisition levels, and overall competence development.

[0004] While some existing educational assessment systems have attempted to incorporate data collection and statistical analysis functions, the following problems still exist: The data sources are singular, mostly limited to attendance records or teacher ratings, and fail to effectively integrate multimodal information such as classroom observation, work evaluation, peer review, and environmental perception. The evaluation indicator system is static and rigid, lacking a structured description and quantitative representation of the core competencies of labor education; The analysis model is superficial, only performing simple summaries or averages, failing to delve into the potential relationships and causal paths between data from different dimensions, resulting in evaluation results that lack explanatory power and guidance.

[0005] Especially in the context of large-scale, routine labor education, the aforementioned problems severely restrict the dynamic adjustment and personalized intervention of teaching plans, making it difficult to support education administrators and teachers in making precise decisions. Therefore, there is an urgent need for a method to evaluate the effectiveness of labor education practice teaching that can deeply integrate multi-source data, construct a scientific indicator system, and achieve intelligent correlation analysis. Summary of the Invention

[0006] This invention provides a method for evaluating the effectiveness of labor education practice teaching based on big data analysis. It constructs a data collection system covering the entire process, all elements, and all subjects, integrates multi-source heterogeneous data, establishes a three-dimensional quantitative evaluation index system with labor participation, skill mastery, and quality development as the core, and adopts a multi-modal feature fusion and dynamic weight adjustment mechanism to conduct objective, accurate, and traceable quantitative evaluation of the effectiveness of labor education practice teaching. At the same time, it generates targeted teaching planning optimization suggestions, thereby achieving closed-loop feedback and continuous improvement of labor education quality.

[0007] This invention provides a method for evaluating the effectiveness of practical teaching in labor education based on big data analysis, including: Collect multi-source data of students in the process of labor education practice. The multi-source data includes student identity information, behavior trajectory data, task completion records, teacher evaluation data, peer evaluation data, image or text data of the results, classroom interaction logs and external environment perception data. The multi-source data is standardized, cleaned, timestamp aligned, and structured to form a time-series dataset in a unified format. Based on the aforementioned time-series dataset, the labor participation index, skill mastery index, and quality development index were calculated respectively. The labor participation index was weighted and synthesized by attendance frequency, task response timeliness, collaborative interaction density, and continuous investment duration. The skill mastery index was weighted and synthesized by task completion accuracy, compliance with operating procedures, proficiency in tool use, and result quality score. The quality development index was weighted and synthesized by responsibility awareness, compliance with safety regulations, contribution to teamwork, and depth of reflection log. The three indicator values ​​are input into a preset multi-dimensional evaluation model. This multi-dimensional evaluation model uses the analytic hierarchy process to determine the initial weights and dynamically adjusts the weights of each dimension in combination with the historical evaluation results in the sliding window, and outputs a comprehensive evaluation score. Based on the deviation analysis of the comprehensive assessment score and its components, a quantitative assessment report is generated, which includes individual diagnostic reports, class competency maps, and curriculum improvement suggestions. The quantitative evaluation report is fed back to the teaching planning decision-making module, which automatically generates the target adjustment plan, task difficulty gradient configuration plan and resource allocation suggestions for the next cycle of labor education practice teaching based on the preset optimization rule library.

[0008] Preferably, multi-source data is collected from students during labor education practice, including: The attendance records are generated by obtaining students' sign-in and departure timestamps through the campus card system or facial recognition terminals. By deploying an Internet of Things (IoT) sensor network in the workplace, the system can collect real-time data on students' operational sequences, tool usage frequency, and dwell time at designated workstations. Receive student task progress logs, photos of interim results, and self-reflection texts via mobile terminal application; The teacher-side management platform receives structured scores from teachers regarding students' behavior, skill application, and attitudes during the work process. The peer review module collects quantitative scores from group members on each other's cooperation, initiative, and contribution based on a preset evaluation scale. The physical products of labor are automatically photographed using optical imaging equipment, and their integrity, cleanliness, and process standardization characteristics are extracted using image recognition algorithms. The natural language processing engine was used to perform semantic analysis on the students' reflection log texts, extracting keyword frequency, sentiment index, and cognitive depth indicators.

[0009] Preferably, the multi-source data undergoes standardization cleaning, timestamp alignment, and structure transformation, including: Remove data fields with a missing rate greater than a specified value; Continuous numerical data is subjected to min-max normalization, and categorical data is subjected to one-hot encoding transformation. Align all data records using a unified time base; Unstructured text data is transformed into fixed-dimensional semantic embedding vectors after word segmentation, stop word removal, and word vector mapping. After the image data is processed by the convolutional neural network feature extractor, a visual feature vector is output. All processed data are aggregated by student identifiers to form a multi-channel time series matrix indexed by time.

[0010] Preferably, the calculation of the labor participation index value includes: Attendance frequency is defined as the ratio of the actual number of working days to the number of working days that should have been taken within the statistical period. The task response time is defined as the average delay time from task release to the student's first operation, which is mapped to a value between 0 and 1 using an inverse proportional function. Collaborative interaction density is defined as the number of times a student interacts effectively with other members per unit of time. Effective interactions include voice dialogue, tool transfer, or collaborative operation. The duration of continuous input is defined as the proportion of the longest continuous operation without interruption during a single work process to the total time. After weighted summation, the scores are compressed to a 0-100 scale using the Sigmoid function to obtain the labor participation index value.

[0011] Preferably, the calculation of the skill mastery index value includes: The accuracy of task completion is quantified by automatically comparing the consistency between the task objective and the actual output using edit distance or pixel similarity algorithms. The compliance rate with operational norms is calculated by matching the student's operation sequence with a preset action rule library and calculating the percentage of compliant steps. Tool proficiency is assessed by constructing a proficiency index based on the number of times the tool is used, the frequency of switching between tools, and the number of misoperations. The quality score of the results is a weighted average of teacher scores, peer scores, and image feature scores; After weighted summation, the results are linearly mapped to a 0-100 point scale to obtain the skill mastery index value.

[0012] Preferably, the calculation of the literacy development index value includes: The sense of responsibility is quantified by the average score of the "proactively taking on responsibilities" and "persistently completing tasks" dimensions in teacher evaluations; Safety compliance is obtained by negatively mapping the number of violations detected by IoT sensors; Teamwork contribution is calculated from the average value of the "helpful" and "effective communication" items in the peer evaluation. The depth of the reflection log is comprehensively evaluated by the text complexity, the number of causal reasoning sentences, and the density of metacognitive vocabulary calculated by the natural language processing engine. After weighted summation and normalization to a 0-100 point scale, the index value of literacy development is obtained.

[0013] Preferably, the multi-dimensional evaluation model adopts a three-layer structure: The first layer is the indicator layer, which includes three primary indicators: labor participation rate, skill mastery rate, and quality development rate. The second layer is the criteria layer; The third layer is the data layer, which corresponds to the various observation variables collected in the original data.

[0014] Preferably, the initial weights are determined by the expert Delphi method and fixed in the weight allocation table; The dynamic weight adjustment mechanism uses historical data from the prescribed evaluation period as a window to calculate the correlation coefficient between each primary indicator and the achievement of the final teaching objective, and redistributes the weights according to the magnitude of the correlation coefficient to ensure that the evaluation focus shifts dynamically with the teaching stage.

[0015] Preferably, the generated quantitative evaluation report includes: Calculate the absolute and relative deviations of each indicator value from the class average within the current evaluation period; Items with a deviation greater than a preset threshold are marked as dimensions to be optimized. Based on a pre-set diagnostic rule base, a corresponding causal hypothesis is matched for each dimension to be optimized. The causal hypothesis includes insufficient instructional design, uneven resource allocation, differences in students' basic knowledge, or external interference factors. Generate individual-level diagnostic summaries, including strengths, weaknesses, and improvement suggestions; Aggregate all class data, create a 3D radar chart to show the overall ability distribution, and mark outlier areas; Provide suggestions for course optimization, including adjusting task types, strengthening teacher allocation, or supplementing safety training.

[0016] Preferably, the teaching planning decision module performs the following operations: If the labor participation index value is less than the standard deviation of the class mean, it is recommended to increase the fun of the task design or introduce an incentive mechanism. If the skill mastery index is generally low in the operation of specific tools, it is recommended to add a special training module in the next cycle. If the compliance rate with safety regulations in the competency development section continues to decline, it is recommended to increase the frequency of pre-job safety drills and update warning signs. Based on the distribution of comprehensive evaluation scores, the system automatically categorizes individuals into three ability groups: high, medium, and low, and generates differentiated task packages for each group. The resource allocation suggestions will be synchronized to the academic affairs management system, triggering processes such as material procurement, venue reservation, or teacher scheduling.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention constructs a multi-source data collection system covering the entire process of labor education, breaking through the limitations of traditional reliance on subjective questionnaires or single performance evaluations, and realizing the objectivity and comprehensiveness of assessment data; 2. A three-dimensional quantitative indicator system with labor participation, skill mastery, and quality development as its core has been established, enabling the precise measurement of abstract labor quality. 3. A dynamic weight adjustment mechanism is introduced to enable the assessment model to adapt to the key objectives of different teaching stages, thereby improving the timeliness and relevance of the assessment. 4. Through deviation analysis and rule matching, individual diagnoses and course optimization suggestions are automatically generated, forming a closed-loop teaching optimization mechanism of "assessment-feedback-improvement"; 5. The overall system architecture supports large-scale concurrent processing and can be deployed on regional education cloud platforms, providing macro-level decision-making support for education administrative departments and improving the scientific, refined, and intelligent level of labor education practice teaching. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the overall technical solution architecture of the present invention; Figure 2 This is a schematic diagram illustrating the core principle framework of the three-dimensional quantitative evaluation index system and dynamic weight adjustment mechanism in this invention; Figure 3This is a logical flowchart of the multi-source heterogeneous data acquisition and preprocessing stage in this invention. Figure 4 This is a flowchart illustrating the logical process of calculating the three core indicators of labor participation, skill mastery, and literacy development in this invention. Figure 5 This is a logical flowchart of the quantitative evaluation report generation and deviation analysis stages in this invention; Figure 6 This is a schematic diagram of the multi-level interaction relationship and data flow between the teaching planning decision module and the external academic affairs system in this invention. Detailed Implementation

[0019] refer to Figures 1 to 6 This invention provides a method for evaluating the effectiveness of labor education practice teaching based on big data analysis. This method constructs a data collection system covering the entire process, all elements, and all subjects, integrates multi-source heterogeneous data, establishes a three-dimensional quantitative evaluation index system with labor participation, skill mastery, and quality development as the core, and adopts a multi-modal feature fusion and dynamic weight adjustment mechanism to conduct objective, accurate, and traceable quantitative evaluation of the effectiveness of labor education practice teaching. At the same time, it generates targeted teaching planning optimization suggestions, thereby achieving closed-loop feedback and continuous improvement of labor education quality.

[0020] The method includes the following steps: S1 collects multi-source data from students during their labor education practice. S2, standardize and clean the multi-source data, align the timestamps, and perform structure transformation to form a time-series dataset in a unified format; S3. Based on the time-series dataset, calculate the labor participation index, skill mastery index, and literacy development index respectively. S4. Input the three indicator values ​​into the preset multi-dimensional evaluation model. The multi-dimensional evaluation model uses the analytic hierarchy process to determine the initial weights and dynamically adjusts the weights of each dimension in combination with the historical evaluation results in the sliding window, and outputs the comprehensive evaluation score. S5. Based on the deviation analysis of the comprehensive evaluation score and its components, generate a quantitative evaluation report that includes individual diagnostic reports, class competency maps, and curriculum improvement suggestions. S6. The quantitative evaluation report is fed back to the teaching planning decision module, which automatically generates the target adjustment plan, task difficulty gradient configuration plan and resource allocation suggestions for the next cycle of labor education practice teaching based on the preset optimization rule library.

[0021] In step S1, multi-source data of students during labor education practice is collected. The multi-source data includes student identity information, behavior trajectory data, task completion records, teacher evaluation data, peer evaluation data, images or text data of the results, classroom interaction logs, and external environment perception data.

[0022] Specifically, attendance records are generated by obtaining students' sign-in and departure timestamps through the campus card system or facial recognition terminals. By deploying an Internet of Things (IoT) sensor network in the workplace, the system can collect real-time data on students' operational sequences, tool usage frequency, and dwell time at designated workstations. Receive student task progress logs, photos of interim results, and self-reflection texts via mobile terminal application; The teacher-side management platform receives structured scores from teachers regarding students' behavior, skill application, and attitudes during the work process. The peer review module collects quantitative scores from group members on each other's cooperation, initiative, and contribution based on a preset evaluation scale. The physical products of labor are automatically photographed using optical imaging equipment, and their integrity, cleanliness, and process standardization characteristics are extracted using image recognition algorithms. The natural language processing engine was used to perform semantic analysis on the students' reflection log texts, extracting keyword frequency, sentiment index, and cognitive depth indicators.

[0023] All data acquisition devices are connected to the campus LAN and use an encrypted socket communication protocol based on the Transmission Control Protocol to ensure the security and integrity of data transmission.

[0024] Data collection frequency is set according to data type: behavioral trajectories and operation sequences are sampled every 5 seconds, image and text data are uploaded when task nodes are triggered, and evaluation data is entered within 24 hours after the task ends.

[0025] In step S2, the multi-source data undergoes standardization cleaning, timestamp alignment, and structure transformation to form a time-series dataset with a unified format. The specific execution process is as follows: First, remove data fields with a missing rate greater than the specified value of 30%, and fill in fields with a missing rate less than the specified value using forward filling or linear interpolation. Secondly, continuous numerical data is subjected to min-max normalization to map it to the 0 to 1 interval, and categorical data is subjected to one-hot encoding to generate binary feature vectors. Thirdly, all data records are aligned with a unified time base, and the time granularity is set to 5 minutes, that is, all events are aggregated according to a 5-minute time window, and if there is no data in a certain window, it is marked as null. Subsequently, the unstructured text data was segmented, stop words were removed, and word vectors were mapped to transform it into fixed-dimensional semantic embedding vectors. The word vector model adopted a Chinese pre-trained model finely tuned on the educational corpus, with an output dimension of 300. The image data was processed by a convolutional neural network feature extractor, which output a 1280-dimensional visual feature vector. This convolutional neural network feature extractor was based on a residual network architecture and completed transfer learning on a public labor education image dataset. Finally, all processed data are aggregated by student identifier to form a multi-channel time series matrix indexed by time. Each row corresponds to a time window, and each column corresponds to a feature channel. The matrix dimension is T×D, where T is the number of time windows and D is the total number of feature dimensions.

[0026] In step S3, based on the time-series dataset, the labor participation index, skill mastery index, and literacy development index are calculated respectively. The labor participation index is a weighted composite of attendance frequency, task response timeliness, collaborative interaction density, and continuous investment duration.

[0027] Attendance frequency is defined as the ratio of actual working days to required working days within a statistical period, ranging from 0 to 1; task response time is defined as the average delay time from task assignment to the student's first operation, expressed as an inverse proportional function. Mapped to values ​​between 0 and 1 This refers to the "average delay time from task release to the student's first operation" corresponding to the task response time. To adjust the parameters, it is set to 0.02; the collaborative interaction density is defined as the number of times a student interacts effectively with other members per unit time. The effective interaction includes voice dialogue, tool transfer or collaborative operation, which is jointly determined by the IoT sensor and the audio analysis module; the continuous input time is defined as the proportion of the longest continuous time of uninterrupted operation in a single work process to the total time.

[0028] The four sub-indicators were assigned initial weights of 0.3, 0.25, 0.25, and 0.2, respectively. After weighted summation, the results were compressed to a 0-1 percentage scale using the Sigmoid function to obtain the labor participation index value. The calculation formula is as follows: ; , , , These are the normalized values ​​of the four sub-indicators. , , , , This is the bias term, set to 0.5.

[0029] The skill mastery index is a weighted composite of task completion accuracy, compliance with operational standards, tool proficiency, and output quality score. Task completion accuracy is determined by the system automatically comparing the consistency between the task objective and the actual output; for text-based tasks, an edit distance algorithm is used, and for physical task tasks, a pixel similarity algorithm is used, with the result normalized to 0-1. Compliance with operational standards is calculated by matching student operation sequences against a pre-set action rule base and determining the percentage of compliant steps. Tool proficiency... A proficiency index is constructed based on the number of tool calls, switching frequency, and number of misoperations. The formula is as follows: , For the effective number of calls, Total duration This represents the number of erroneous operations. The attenuation coefficient is set to 0.1. The achievement quality score is a weighted average of teacher scores, peer scores, and image feature scores, with weights of 0.5, 0.3, and 0.2, respectively. The four sub-indicators are assigned initial weights of 0.4, 0.2, 0.2, and 0.2, respectively, and after weighted summation, they are linearly mapped to a 0-1 percentage scale to obtain the skill mastery index value. .

[0030] The quality development index is a weighted composite of the performance of responsibility awareness, compliance with safety regulations, contribution to teamwork, and the depth of reflection log.

[0031] The sense of responsibility is quantified by the average score of dimensions such as "proactively taking on responsibilities" and "persistently completing tasks" in teacher evaluations; compliance with safety regulations The number of violations detected by IoT sensors is obtained through negative mapping, as shown in the formula: , For the number of violations, The preset upper limit is set to 5 times; the team collaboration contribution is calculated from the average of items such as "helpful" and "effective communication" in the peer evaluation; the depth of the reflection log is comprehensively evaluated by the text complexity, the number of causal reasoning sentences and the metacognitive vocabulary density calculated by the natural language processing engine, and the weighted average of the three is normalized.

[0032] The four sub-indicators mentioned above were assigned initial weights of 0.3, 0.25, 0.25, and 0.2, respectively. After weighted summation and normalization to a 0-1 percentage system, the index value of literacy development was obtained. .

[0033] In step S4, the three indicator values ​​are... , , The input is fed into a pre-defined multi-dimensional evaluation model, which employs a three-layer structure: The first layer is the indicator layer, which includes three primary indicators: labor participation rate, skill mastery rate, and quality development rate. The second layer is the criteria layer, with each primary indicator having four secondary sub-indicators, for a total of 12. The third layer is the data layer, which corresponds to the various observation variables collected in the original data.

[0034] Initial weights were determined using the expert Delphi method and fixed in the weight allocation table. The dynamic weight adjustment mechanism uses historical data from the specified evaluation period (the most recent 30 periods) as a window to calculate the correlation coefficient between each primary indicator and the achievement of the final teaching objective, and then reallocates weights according to the magnitude of the correlation coefficient. Let the weight of the current indicator within the window be... The overall evaluation score for each cycle is , No. Labor participation index value for each period Skill mastery index value Cultivation of basic literacy indicators The Pearson correlation coefficient, an indicator of labor participation, is... Pearson correlation coefficient, a measure of skill mastery Pearson correlation coefficient, an indicator of the degree of development of literacy. Calculated separately as follows: ; ; ; This represents the average value of the labor participation rate index over 30 periods. This is the average of the comprehensive evaluation scores over 30 periods. This represents the average of the skill mastery index over 30 periods. This represents the average value of the literacy development index over 30 cycles.

[0035] Dynamic weighting of labor participation index Dynamic weighting of skills mastery indicators Dynamic weighting of literacy development indicators Updated according to proportional scaling principle: ; ; ; Overall assessment score The value ranges from 0 to 100.

[0036] In step S5, based on the deviation analysis of the comprehensive evaluation score and its components, a quantitative evaluation report is generated, which includes individual diagnostic reports, class competency maps, and course improvement suggestions.

[0037] The specific implementation process includes: calculating the absolute deviations of the labor participation index, skill mastery index, and literacy development index from the class average level within the current assessment period. The relative deviations of the labor participation index, skills mastery index, and literacy development index are respectively... μ , , These are the class averages for the labor participation rate, skills mastery rate, and character development rate indicators, respectively. , , These are the standard deviations of labor participation, skills mastery, and literacy development indicators; indicators with deviations greater than a preset threshold are identified, with the threshold set at 1.5 times the absolute value, and marked as dimensions to be optimized. Based on a pre-defined diagnostic rule base, corresponding causal hypotheses are matched for each dimension to be optimized. These hypotheses include inadequate instructional design, uneven resource allocation, differences in students' basic abilities, or external interference factors. Individual-level diagnostic summaries are generated, including strengths, weaknesses, and improvement suggestions. All class data is aggregated to create a 3D radar chart displaying the overall ability distribution, with the three axes corresponding to... , , Mark outlier areas; provide optimization suggestions at the course level, including adjusting task types, strengthening teacher allocation, or supplementing safety training.

[0038] In step S6, the quantitative evaluation report is fed back to the teaching planning decision module, which automatically generates the target adjustment plan, task difficulty gradient configuration plan and resource allocation suggestions for the next cycle of labor education practice teaching based on the preset optimization rule library.

[0039] Specific measures include: if the labor participation index value is less than one standard deviation from the class average, it is recommended to increase the fun of the task design or introduce an incentive mechanism based on points. If the skill mastery index is generally low in the operation of specific tools, it is recommended to add a special training module in the next cycle, such as wood sawing and circuit welding. If the compliance rate with safety regulations in the competency development section continues to decline, it is recommended to increase the frequency of pre-job safety drills and update warning signs. Based on the distribution pattern of the comprehensive assessment scores, the system automatically divides individuals into three ability groups: high, medium, and low. The high-achieving group (…) ), middle group ( ), low group ( And generate differentiated task packages for each group, with high-group tasks focusing on innovative and complex tasks, and low-group tasks focusing on basic norms and repeated training; The resource allocation suggestions are synchronized to the academic affairs management system, triggering material procurement, venue reservation or teacher scheduling processes. For example, when the damage rate of woodworking tools is greater than 20%, a procurement list is automatically generated.

[0040] The implementation of the above method relies on a complete system architecture. This system includes a multi-source data acquisition unit, a data preprocessing unit, a three-dimensional index calculation unit, a dynamic evaluation modeling unit, an evaluation report generation unit, and a teaching plan optimization unit.

[0041] The multi-source data acquisition unit includes an identity recognition subunit, an Internet of Things sensing subunit, a mobile terminal interaction subunit, a teacher evaluation input subunit, a peer evaluation subunit, an optical imaging subunit, and a natural language processing subunit. Each subunit is connected to the central data aggregation node via the campus LAN.

[0042] The data preprocessing unit incorporates a data cleaning rule engine, a time alignment processor, a normalization calculator, a text vectorization module, and an image feature extractor. The image feature extractor uses a pre-trained model based on a residual network to output fixed-dimensional visual feature vectors.

[0043] The three-dimensional indicator calculation unit contains three parallel computing channels, corresponding to labor participation, skill mastery, and literacy development, respectively. Each channel has an embedded sub-indicator calculator and a weighted synthesizer, and the weight parameters are dynamically loaded by an external configuration interface.

[0044] The dynamic evaluation modeling unit includes a static weight storage, a historical data cache, a correlation analyzer, and a weight redistributor. The correlation analyzer uses the Pearson correlation coefficient algorithm, and the weight redistributor updates the weights of each dimension according to the scaling principle to ensure that the total weight is always 1.

[0045] The assessment report generation unit integrates a deviation detector, a rule matching engine, a text generator, and a visualization plotter. The text generator uses a template filling mechanism to convert structured data into natural language descriptions. The visualization plotter uses polar coordinates to draw 3D radar charts and supports aggregated display by class, grade, or semester.

[0046] The teaching planning optimization unit includes a rule base, a task generator, a resource configuration suggestion tool, and an interface adapter. The rule base stores 120 preset optimization strategies, each of which includes trigger conditions, action instructions, and priority indicators. The interface adapter provides application programming interfaces to interface with the academic affairs management, material procurement, and human resources systems.

[0047] The entire system runs on a regional education cloud platform, supporting thousands of concurrent student data processing requests and processing an average of 1 billion records per day. The system adopts a microservice architecture, with each unit deployed independently and elastically scalable. Asynchronous communication is achieved through message queues, ensuring high availability and low-latency response. Data storage utilizes a distributed time-series database, supporting efficient time-range queries and aggregation calculations. Security mechanisms encompass encrypted data transmission, access control, and operation log auditing, meeting the Level 3 requirements of the National Education Information System Security Protection System.

[0048] Through the coordinated operation of the above methods and systems, a fundamental shift has been achieved in the evaluation of the effectiveness of labor education practice teaching from subjective experience judgment to objective data-driven evaluation. This has solved the technical problems of traditional evaluation methods, such as narrow coverage, strong lag, and lack of personalized feedback, and improved the scientific, refined, and intelligent level of labor education.

[0049] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0050] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for evaluating the effectiveness of practical teaching in labor education based on big data analysis, characterized in that: include: Collect multi-source data of students in the process of labor education practice. The multi-source data includes student identity information, behavior trajectory data, task completion records, teacher evaluation data, peer evaluation data, image or text data of the results, classroom interaction logs and external environment perception data. The multi-source data is standardized, cleaned, timestamp aligned, and structured to form a time-series dataset in a unified format. Based on the aforementioned time-series dataset, the labor participation index, skill mastery index, and quality development index were calculated respectively. The labor participation index was weighted and synthesized by attendance frequency, task response timeliness, collaborative interaction density, and continuous investment duration. The skill mastery index was weighted and synthesized by task completion accuracy, compliance with operating procedures, proficiency in tool use, and result quality score. The quality development index was weighted and synthesized by responsibility awareness, compliance with safety regulations, contribution to teamwork, and depth of reflection log. The three indicator values ​​are input into a preset multi-dimensional evaluation model. This multi-dimensional evaluation model uses the analytic hierarchy process to determine the initial weights and dynamically adjusts the weights of each dimension in combination with the historical evaluation results in the sliding window, and outputs a comprehensive evaluation score. Based on the deviation analysis of the comprehensive assessment score and its components, a quantitative assessment report is generated, which includes individual diagnostic reports, class competency maps, and curriculum improvement suggestions. The quantitative evaluation report is fed back to the teaching planning decision-making module, which automatically generates the next cycle of labor education practice teaching goal adjustment plan, task difficulty gradient configuration plan and resource allocation suggestions based on the preset optimization rule library.

2. The method for evaluating the effectiveness of labor education practical teaching based on big data analysis according to claim 1, characterized in that, Collect multi-source data from students during labor education practices, including: The attendance records are generated by obtaining students' sign-in and departure timestamps through the campus card system or facial recognition terminals. By deploying an Internet of Things (IoT) sensor network in the workplace, the system can collect real-time data on students' operational sequences, tool usage frequency, and dwell time at designated workstations. Receive student task progress logs, photos of interim results, and self-reflection texts via mobile terminal application; The teacher-side management platform receives structured scores from teachers regarding students' behavior, skill application, and attitudes during the work process. The peer review module collects quantitative scores from group members on each other's cooperation, initiative, and contribution based on a preset evaluation scale. The physical products of labor are automatically photographed using optical imaging equipment, and their integrity, cleanliness, and process standardization characteristics are extracted using image recognition algorithms. The natural language processing engine was used to perform semantic analysis on the students' reflection log texts, extracting keyword frequency, sentiment index, and cognitive depth indicators.

3. The method for evaluating the effectiveness of labor education practical teaching based on big data analysis according to claim 2, characterized in that, The multi-source data undergoes standardization cleaning, timestamp alignment, and structure transformation, including: Remove data fields with a missing rate greater than a specified value; Continuous numerical data is subjected to min-max normalization, and categorical data is subjected to one-hot encoding transformation. Align all data records using a unified time base; Unstructured text data is transformed into fixed-dimensional semantic embedding vectors after word segmentation, stop word removal, and word vector mapping. After the image data is processed by the convolutional neural network feature extractor, a visual feature vector is output. All processed data are aggregated by student identifiers to form a multi-channel time series matrix indexed by time.

4. The method for evaluating the effectiveness of labor education practical teaching based on big data analysis according to claim 3, characterized in that, The calculation of the labor participation index value includes: Attendance frequency is defined as the ratio of the actual number of working days to the number of working days that should have been taken within the statistical period. The task response time is defined as the average delay time from task release to the student's first operation, which is mapped to a value between 0 and 1 using an inverse proportional function. Collaborative interaction density is defined as the number of times a student interacts effectively with other members per unit of time. Effective interactions include voice dialogue, tool transfer, or collaborative operation. The duration of continuous input is defined as the proportion of the longest continuous operation without interruption during a single work process to the total time. After weighted summation, the scores are compressed to a 0-100 scale using the Sigmoid function to obtain the labor participation index value.

5. The method for evaluating the effectiveness of labor education practical teaching based on big data analysis according to claim 4, characterized in that, The calculation of the skill mastery index value includes: The accuracy of task completion is quantified by automatically comparing the consistency between the task objective and the actual output using edit distance or pixel similarity algorithms. The compliance rate with operational norms is calculated by matching the student's operation sequence with a preset action rule library and calculating the percentage of compliant steps. Tool proficiency is assessed by constructing a proficiency index based on the number of times the tool is used, the frequency of switching between tools, and the number of misoperations. The quality score of the results is a weighted average of teacher scores, peer scores, and image feature scores; After weighted summation, the results are linearly mapped to a 0-100 point scale to obtain the skill mastery index value.

6. The method for evaluating the effectiveness of labor education practical teaching based on big data analysis according to claim 5, characterized in that, The calculation of the literacy development index value includes: The sense of responsibility is quantified by the average score of the "proactively taking on responsibilities" and "persistently completing tasks" dimensions in teacher evaluations; Safety compliance is obtained by negatively mapping the number of violations detected by IoT sensors; Teamwork contribution is calculated from the average value of the "helpful" and "effective communication" items in the peer evaluation. The depth of the reflection log is comprehensively evaluated by the text complexity, the number of causal reasoning sentences, and the density of metacognitive vocabulary calculated by the natural language processing engine. After weighted summation and normalization to a 0-100 point scale, the index value of literacy development is obtained.

7. The method for evaluating the effectiveness of labor education practical teaching based on big data analysis according to claim 6, characterized in that, The multi-dimensional evaluation model adopts a three-layer structure: The first layer is the indicator layer, which includes three primary indicators: labor participation rate, skill mastery rate, and quality development rate. The second layer is the criteria layer; The third layer is the data layer, which corresponds to the various observation variables collected in the original data.

8. The method for evaluating the effectiveness of labor education practical teaching based on big data analysis according to claim 7, characterized in that, The initial weights were determined using the expert Delphi method and fixed in the weight allocation table; The dynamic weight adjustment mechanism uses historical data from the prescribed evaluation period as a window to calculate the correlation coefficient between each primary indicator and the achievement of the final teaching objective, and redistributes the weights according to the magnitude of the correlation coefficient to ensure that the evaluation focus shifts dynamically with the teaching stage.

9. The method for evaluating the effectiveness of labor education practical teaching based on big data analysis according to claim 8, characterized in that, Generating the quantitative assessment report includes: Calculate the absolute and relative deviations of each indicator value from the class average within the current evaluation period; Items with a deviation greater than a preset threshold are marked as dimensions to be optimized. Based on a pre-set diagnostic rule base, a corresponding causal hypothesis is matched for each dimension to be optimized. The causal hypothesis includes insufficient instructional design, uneven resource allocation, differences in students' basic knowledge, or external interference factors. Generate individual-level diagnostic summaries, including strengths, weaknesses, and improvement suggestions; Aggregate all class data, create a 3D radar chart to show the overall ability distribution, and mark outlier areas; Provide suggestions for course optimization, including adjusting task types, strengthening teacher allocation, or supplementing safety training.

10. The method for evaluating the effectiveness of labor education practical teaching based on big data analysis according to claim 9, characterized in that, The instructional planning decision-making module performs the following operations: If the labor participation index value is less than the standard deviation of the class mean, it is recommended to increase the fun of the task design or introduce an incentive mechanism. If the skill mastery index is generally low in the operation of specific tools, it is recommended to add a special training module in the next cycle. If the compliance rate with safety regulations in the competency development section continues to decline, it is recommended to increase the frequency of pre-job safety drills and update warning signs. Based on the distribution of comprehensive evaluation scores, the system automatically categorizes individuals into three ability groups: high, medium, and low, and generates differentiated task packages for each group. The resource allocation suggestions will be synchronized to the academic affairs management system, triggering processes such as material procurement, venue reservation, or teacher scheduling.

Images