Education data processing method based on coexistence of standard reference and norm reference

Abstract

The invention relates to the technical field of data processing, in particular to an education data processing method based on coexistence of standard reference and norm reference, which comprises the following steps: acquiring real-time access times, accuracy and interaction frequency; determining a first temporary target; determining a second temporary target; determining a processing target; determining a target evaluation index and an analysis evaluation index; and adjusting a reference threshold according to the index deviation and then determining an evaluation grade. According to the invention, through a dual evaluation mechanism fusing standard reference and norm reference, the performance of the student in different learning scenes can be accurately captured, and the learning state of the student can be evaluated in real time and dynamically adjusted. Through step-by-step adjustment of the standard reference threshold and deep application of the clustering analysis model, the limitation of a single evaluation mode can be eliminated, so that the evaluation result is more comprehensive, accurate and personalized, and the problem of low processing equipment identification accuracy caused by poor data objectivity due to a single evaluation standard is effectively solved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to an educational data processing method based on the coexistence of standard reference and norm reference. Background Art

[0002] In modern education, effectively evaluating students' learning performance and engagement has become a key issue in improving educational quality. With the integration of online and offline learning, the diversity and complexity of educational data have increased. Traditional evaluation methods struggle to fully reflect students' individual differences and behavioral characteristics.

[0003] Patent document CN115186014A discloses a data processing method for educational training, which includes: obtaining the frequency of looking around, typical hesitation time and hesitation stability of each child in a target classroom; calculating the attractiveness of the classroom to each child based on the typical hesitation time and hesitation stability of each child in the target classroom; calculating the behavioral difference distance between any two children based on the attractiveness of the classroom to each child, the frequency of looking around and the hesitation time series in the target classroom; classifying each child in the target classroom according to the behavioral difference distance to obtain multiple types; the hesitation time series is the number of children in the classroom. A sequence consisting of each interaction hesitation time; for any type: calculate the spatial distance between each child in this type and other children of this type according to the frequency of looking around of each child in this type and the looking state sequence corresponding to the target time period, calculate the behavioral affiliation corresponding to each child in this type according to the spatial distance, and take the child with the largest behavioral affiliation as the representative child of this type corresponding to the target time period, and take the representative child of this type corresponding to the target time period as the object of recommendation for simulation when the kindergarten teacher reviews the target time later; the looking state sequence corresponding to the target time period is a sequence consisting of the looking states of the corresponding children at each collection moment in the target time period.

[0004] It can be seen that the data processing method used for educational training has the following problems: this method mainly relies on behavioral data within a single classroom, such as the frequency of looking around, the time of hesitation in interaction, etc. These data have subjective biases, which affects the accuracy of the evaluation results. Summary of the Invention

[0005] To this end, the present invention provides an educational data processing method based on the coexistence of standard reference and norm reference, which is used to overcome the problem of low recognition accuracy of processing equipment caused by poor data objectivity due to a single evaluation standard in the existing technology by combining the two evaluation methods of standard reference and norm reference.

[0006] To achieve the above-mentioned object, the present invention provides an educational data processing method based on both standard reference and norm reference, comprising: Obtain the real-time visit count of each target to be evaluated in the online learning platform, the real-time accuracy rate of each target to be evaluated after the test, and the real-time interaction frequency of each target to be evaluated in the offline classroom; Determine a plurality of first temporary targets according to the real-time access times and a preset standard reference threshold; Determining a plurality of second temporary targets according to the real-time accuracy of each of the first temporary targets; determining a plurality of processing targets according to the real-time interaction frequency and the real-time visit count of each second temporary target; Determine a target evaluation index according to the real-time access count, the real-time accuracy rate, and the real-time interaction frequency of each processing target; Determining an analysis evaluation index based on a preset norm reference value, the real-time accuracy of all targets to be evaluated, the total number of real-time visits, the total real-time interaction frequency, and a preset cluster analysis model; Adjusting the preset standard reference threshold according to the target evaluation index and the analysis evaluation index to form an adjusted standard reference threshold; The target evaluation index is determined based on the adjustment standard and a reference threshold, and the evaluation level of the target to be evaluated is determined according to the target evaluation index.

[0007] Furthermore, determining a plurality of first temporary targets according to the real-time access count and a preset standard reference threshold includes: When the number of real-time accesses is less than a preset standard reference threshold, the target to be evaluated is determined to be a first temporary target, and a plurality of first temporary targets are formed.

[0008] Further, determining a plurality of second temporary targets according to the real-time accuracy of each of the first temporary targets includes: When the real-time accuracy rate is greater than a preset accuracy rate threshold, calculating a standard deviation of the real-time accuracy rate within a preset first determined time period to form an accuracy rate fluctuation value; When the accuracy fluctuation value is greater than a preset accuracy fluctuation threshold, the first temporary target is determined to be the second temporary target, and a plurality of second temporary targets are formed.

[0009] Furthermore, determining a number of processing targets based on the real-time interaction frequency and the real-time visit count of each second temporary target includes: Calculating a standard deviation of the real-time interaction frequency within a preset second determined time period to form an interaction frequency fluctuation value; Calculating a standard deviation of the real-time access count within the preset second determined time period to form an access count fluctuation value; A number of processing targets are determined according to the interaction frequency fluctuation value and the visit number fluctuation value.

[0010] Furthermore, determining several processing targets based on the interaction frequency fluctuation value and the visit number fluctuation value includes: Drawing a change curve according to the interaction frequency fluctuation value to form an interaction change curve; Drawing a change curve according to the fluctuation value of the number of visits to form an access change curve; Calculating the cosine similarity between the interaction change curve and the access change curve to form a change synchronization degree; When the change synchronization degree is less than a preset synchronization degree threshold, determining that the second temporary target is an impact target; Non-influencing targets among all the targets to be evaluated are selected as processing targets to form a plurality of processing targets.

[0011] Furthermore, determining a target evaluation index according to the real-time access count, the real-time accuracy, and the real-time interaction frequency of each processing target includes: Normalizing the real-time access times to form a normalized access time; Normalizing the real-time accuracy to form a normalized accuracy; Normalizing the real-time interaction frequency to form a normalized interaction frequency; The normalized number of visits, the normalized accuracy, the normalized interaction frequency, the preset number of visits weight, the preset accuracy weight, and the preset interaction frequency weight are weighted and summed to form a target evaluation index.

[0012] Furthermore, determining the analysis evaluation index based on a preset norm reference value, the real-time accuracy of all targets to be evaluated, the total number of real-time visits, the total real-time interaction frequency, and a preset cluster analysis model includes: Sorting all the real-time accuracy rates and determining the ranking value of each target to be evaluated; Calculating the sum of the ranking value and the preset norm reference value to obtain a ranking range lower limit, calculating the difference between the ranking value and the preset norm reference value to obtain a ranking range upper limit, thereby forming a ranking range; An analysis evaluation index is determined according to the real-time accuracy, the real-time number of visits, and the real-time interaction frequency of all the targets to be evaluated within the ranking range.

[0013] Furthermore, determining the analysis evaluation index according to the real-time accuracy, the real-time number of visits, and the real-time interaction frequency of all the targets to be evaluated within the ranking range includes: The real-time accuracy, the real-time number of visits, and the real-time interaction frequency of all the targets to be evaluated within the ranking range are input into the preset cluster analysis model to obtain the analysis evaluation index.

[0014] Furthermore, adjusting the preset standard reference threshold according to the target evaluation index and the analysis evaluation index to form the adjusted standard reference threshold includes: Calculating the relative deviation between the target evaluation index and the analysis evaluation index to form an index deviation; When the index deviation is greater than the preset deviation threshold, the preset standard reference threshold is adjusted according to the relative deviation between the index deviation and the preset deviation threshold and a preset adjustment coefficient to form an adjusted standard reference threshold.

[0015] Furthermore, determining the evaluation level of the target to be evaluated according to the target evaluation index includes: The evaluation level corresponding to the target evaluation index is searched in the preset evaluation index table to obtain the evaluation level of the target to be evaluated.

[0016] Compared with the existing technology, the beneficial effect of the present invention is that, by integrating the dual evaluation mechanism of standard reference and norm reference, it can accurately capture students' performance in different learning scenarios, evaluate their learning status in real time and make dynamic adjustments. Through the gradual adjustment of the standard reference threshold and the in-depth application of the cluster analysis model, the limitations of a single evaluation method can be eliminated, making the evaluation results more comprehensive, accurate and personalized. In addition, this method can not only promptly identify the weak links in students' learning process, but also dynamically adjust the threshold and evaluation criteria based on actual performance, thereby improving the flexibility and adaptability of educational evaluation, and effectively solving the problem of low recognition accuracy of processing equipment caused by poor data objectivity due to a single evaluation standard.

[0017] Furthermore, by comparing real-time visit counts against pre-set thresholds, we can effectively screen out those targets that may be experiencing insufficient engagement, thereby promptly identifying potential low activity issues. By pre-calibrating these low-visit targets, we can provide a basis for subsequent adjustments to teaching strategies and resource optimization, ensuring a more accurate and targeted evaluation process.

[0018] Furthermore, by comparing accuracy fluctuations against a preset accuracy fluctuation threshold, we can identify individuals with unstable performance. Even if their overall accuracy is high, large fluctuations may indicate unstable learning. Fluctuation analysis provides a more comprehensive assessment than relying solely on static accuracy, helping to identify potential problems and optimize the learner's progress. This approach facilitates a more objective analysis of learners' performance and provides data support for educational decision-making.

[0019] Furthermore, by analyzing fluctuations in interaction frequency and visit counts, we can identify targets with unstable performance. This allows us to precisely locate targets requiring intervention, avoiding judgments based solely on static data. This improves assessment accuracy and intervention effectiveness, ensuring that learning targets receive appropriate attention and support.

[0020] Furthermore, by quantifying the fluctuations in interaction frequency and number of visits and calculating the degree of synchronization of their changes, we can effectively distinguish targets with different performance fluctuations and mark targets that do not meet the synchronization requirements as influencing targets, thereby optimizing subsequent processing and improving the accuracy of evaluation results.

[0021] Furthermore, the weighted summation method can integrate multi-dimensional data and assign a reasonable weight to each indicator, making the target evaluation index more accurate and objective. Through this index, a comprehensive assessment of the targets can be carried out, which helps decision makers identify and prioritize those targets with outstanding performance.

[0022] Furthermore, by combining ranking and ranking calculations with norm-referenced values, we can objectively and comprehensively assess the performance of each evaluation target, avoiding bias caused by a single criterion and ensuring a more accurate and comprehensive analysis and evaluation index. Furthermore, by leveraging the support of cluster analysis models, we can effectively improve the accuracy and reliability of evaluations across multiple dimensions.

[0023] Furthermore, the application of cluster analysis models can effectively integrate multiple evaluation dimensions to provide a comprehensive and accurate evaluation index. By combining real-time data, it can dynamically reflect the actual performance of the target, avoiding the limitations of a single evaluation standard, and providing a scientific basis for decision-making.

[0024] Furthermore, based on the difference between the actual performance of the target and the evaluation index, the standard reference threshold is dynamically adjusted to ensure that the evaluation system is more accurate and flexible. By adjusting the threshold, it is possible to cope with deviations in different situations, improve the accuracy and adaptability of the evaluation, and thus optimize the decision-making process.

[0025] Furthermore, by searching the evaluation index table, we can efficiently and accurately assign the corresponding evaluation level to the evaluated target, reduce the interference of human factors, and ensure the consistency and objectivity of the evaluation. At the same time, the entire evaluation process has the advantages of standardization and automation, which helps to improve the transparency and efficiency of the evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 Flowchart of the educational data processing method based on the coexistence of standard reference and norm reference in this embodiment; Figure 2 A logic decision diagram for determining the first temporary target for this embodiment; Figure 3A logic decision diagram for determining the second temporary target for this embodiment; Figure 4 A logical decision diagram for determining the impact targets for this embodiment. DETAILED DESCRIPTION

[0027] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention.

[0028] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0029] See also Figure 1 As shown, it is a flow chart of the educational data processing method based on the coexistence of standard reference and norm reference in this embodiment; This embodiment provides a method for processing educational data based on both standard reference and norm reference, including: Obtain the real-time visit count of each target to be evaluated in the online learning platform, the real-time accuracy rate of each target to be evaluated after the test, and the real-time interaction frequency of each target to be evaluated in the offline classroom; Determine a plurality of first temporary targets according to the real-time access times and a preset standard reference threshold; Determining a plurality of second temporary targets according to the real-time accuracy of each of the first temporary targets; determining a plurality of processing targets according to the real-time interaction frequency and the real-time visit count of each second temporary target; Determine a target evaluation index according to the real-time access count, the real-time accuracy rate, and the real-time interaction frequency of each processing target; Determining an analysis evaluation index based on a preset norm reference value, the real-time accuracy of all targets to be evaluated, the total number of real-time visits, the total real-time interaction frequency, and a preset cluster analysis model; Adjusting the preset standard reference threshold according to the target evaluation index and the analysis evaluation index to form an adjusted standard reference threshold; The target evaluation index is determined based on the adjustment standard and a reference threshold, and the evaluation level of the target to be evaluated is determined according to the target evaluation index.

[0030] Real-time data on each evaluation objective is typically obtained through a data interface that integrates the online learning platform, testing system, and classroom interaction tools. Online learning platform visits are automatically recorded and updated in real time based on student login behavior and course browsing history. Post-test accuracy is calculated by connecting to the testing system interface, collecting student performance data for each test in real time. Offline classroom interaction frequency is collected through interactive devices within the classroom (such as smart blackboards, classroom management systems, and student response systems), including student participation data such as the number of questions answered, the number of questions asked, and the frequency of hand raises. This data is monitored and stored in real time through a unified data collection platform, ensuring the accuracy and timeliness of each indicator for subsequent analysis and processing.

[0031] The preset standard reference threshold is a benchmark used to determine whether a student's learning behavior meets basic standards. It is usually determined by historical data on the platform or the expected learning goals, and is usually set between 8 and 12 visits. In this example, it is set to 10 visits to ensure that students' learning engagement on the platform reaches a certain level, preventing low-engagement students from being mistakenly rated as good students.

[0032] The preset norm reference value is a benchmark ranking value used to evaluate the relative performance of the target to be evaluated. It depends on the overall distribution of the target group to be evaluated, especially the ranking of the targets in the group. It is usually set to a ranking value of a preset percentage of the total target group to ensure that the relative position of the evaluated target in the group meets the predetermined standards. It is usually set to a ranking value of the top 10% or 30% of the total target group. In this embodiment, the preset norm reference value is set to a ranking value of 20% of the total target group. This can ensure moderate screening and stratification during system evaluation, avoid over-concentration or over-dispersed evaluation results, and thus optimize the ranking and assessment of targets.

[0033] The default clustering analysis model is the KAMILA clustering algorithm model. The KAMILA clustering analysis model is a multimodal data analysis method based on a clustering algorithm. The KAMILA clustering algorithm combines the K-means clustering method with multi-instance learning. It can process multidimensional data and effectively perform cluster analysis on different types of modal data. Specifically, the KAMILA clustering algorithm represents each data instance as multiple sub-instances (i.e., multiple features or data points) and uses the relationships between each sub-instance to perform clustering. This method not only effectively clusters structured data but also handles uncertainty and noise in the data. It is particularly suitable for scenarios in the education field where comprehensive evaluation requires simultaneous consideration of multiple evaluation indicators (such as test scores, interaction frequency, etc.). In this embodiment, the KAMILA clustering model can simultaneously process data from different data sources (such as the number of visits to the online learning platform, test accuracy, and classroom interaction frequency) to form a multi-dimensional evaluation system. Through multi-instance learning, it can better handle data with uncertainty and noise, ensuring that the clustering results are more robust. Through cluster analysis, it can take into account both standard-referenced and norm-referenced evaluation methods, solve the problem of fragmented evaluation standards in traditional methods, and provide more comprehensive and objective evaluation results.

[0034] First, real-time data on the target to be evaluated in different learning scenarios is obtained, including the number of visits to the online learning platform, the accuracy rate after the test, and the frequency of interaction in offline classes. A first provisional target is determined by setting a standard reference threshold, and a second provisional target is further selected based on the real-time accuracy rate. Subsequently, the processing target is determined by combining the interaction frequency and number of visits, and the target evaluation index is calculated based on this data. Next, an analysis evaluation index is generated using the norm reference value and cluster analysis model, and the standard reference threshold is then adjusted. Finally, based on the adjusted standard reference threshold, the final evaluation grade of the target to be evaluated is determined.

[0035] By integrating a dual assessment mechanism of criterion-referenced and norm-referenced learning, it is possible to accurately capture students' performance in different learning scenarios, assess their learning status in real time, and dynamically adjust it. By gradually adjusting the criterion-referenced threshold and deeply applying cluster analysis models, the limitations of a single evaluation method can be eliminated, making the evaluation results more comprehensive, accurate, and personalized. Furthermore, this method not only promptly identifies weak links in students' learning processes but also dynamically adjusts thresholds and assessment criteria based on actual performance, improving the flexibility and adaptability of educational assessments and effectively addressing the problem of low recognition accuracy of processing equipment caused by the lack of data objectivity due to a single assessment criterion.

[0036] Please continue reading Figure 2 As shown, it is a logic decision diagram for determining the first temporary target in this embodiment; Specifically, determining a plurality of first temporary targets according to the real-time access count and a preset standard reference threshold includes: When the number of real-time accesses is less than the preset standard reference threshold, the target to be evaluated is determined to be a first temporary target, and a plurality of first temporary targets are formed.

[0037] Based on a comparison of real-time visit counts with a preset standard reference threshold, if a target's real-time visit count falls below the threshold, it is designated as a first provisional target. This determination allows the system to filter out targets with insufficient visit counts and analyze and process them separately, allowing for further evaluation of their performance based on other metrics.

[0038] By comparing real-time visit counts against pre-set thresholds, we can effectively screen out targets that may be experiencing insufficient engagement, thereby promptly identifying potential low activity issues. Preemptively identifying these low-visit targets provides a basis for subsequent instructional strategy adjustments and resource optimization, ensuring a more accurate and targeted evaluation process.

[0039] Please continue reading Figure 3 As shown, it is a logic decision diagram for determining the second temporary target in this embodiment; Specifically, determining a plurality of second temporary targets according to the real-time accuracy of each of the first temporary targets includes: When the real-time accuracy rate is greater than a preset accuracy rate threshold, calculating a standard deviation of the real-time accuracy rate within a preset first determined time period to form an accuracy rate fluctuation value; When the accuracy fluctuation value is greater than a preset accuracy fluctuation threshold, the first temporary target is determined to be the second temporary target, and a plurality of second temporary targets are formed.

[0040] The preset accuracy threshold is a standard value for judging whether a learning objective is performed well. It depends on the requirements of the educational scenario and the distribution of historical data. It is usually set between 80% and 90%. In this embodiment, it is set to 85%, which helps to screen out those objectives that basically meet the learning requirements and ensure that subsequent analysis can focus on those low-accuracy objectives that require more attention.

[0041] The preset first determination time length is the time period used to calculate the real-time accuracy fluctuation value, which depends on the evaluation cycle and stability requirements of the target performance. It is usually set between 10 minutes and 30 minutes. In this embodiment, it is set to 15 minutes, which can balance the evaluation accuracy and the changes in real-time data, and ensure that the performance trend of the target can be reflected in a timely manner during the evaluation process.

[0042] The preset accuracy fluctuation threshold is a parameter used to evaluate the stability of target performance. It is set based on the historical performance fluctuations of the target group and is usually set between 5% and 15%. In this embodiment, it is set to 10%. This helps to identify targets with high accuracy but large fluctuations, avoids the neglect of targets with high accuracy but instability, and ensures that learners with inconsistent performance are screened out for intervention.

[0043] After the first temporary target is selected, further analysis is performed based on each target's real-time accuracy. If the real-time accuracy exceeds the preset accuracy threshold, the system calculates the standard deviation of the accuracy over a certain period of time to determine the accuracy fluctuation value. If this fluctuation value exceeds the preset accuracy fluctuation threshold, the target is identified as the second temporary target. This process helps screen out targets with significant performance fluctuations, facilitating more in-depth evaluation and adjustment.

[0044] By comparing accuracy fluctuations against preset accuracy fluctuation thresholds, we can identify individuals with unstable performance. Even if their overall accuracy is high, large fluctuations may indicate unstable learning. Fluctuation analysis provides a more comprehensive assessment than relying solely on static accuracy, helping to identify potential problems and optimize the learner's progress. This approach facilitates a more objective analysis of learners' performance and provides data support for educational decision-making.

[0045] Specifically, determining a number of processing targets according to the real-time interaction frequency and the real-time visit count of each second temporary target includes: Calculating a standard deviation of the real-time interaction frequency within a preset second determined time period to form an interaction frequency fluctuation value; Calculating a standard deviation of the real-time access count within the preset second determined time period to form an access count fluctuation value; A number of processing targets are determined according to the interaction frequency fluctuation value and the visit number fluctuation value.

[0046] The preset second determined time length is the time period used to calculate the real-time interaction frequency and visit number fluctuation values. It depends on the collection cycle of the interaction data and the requirements of the target activity frequency. It is usually set between 20 minutes and 60 minutes. In this embodiment, it is set to 30 minutes, which can cover a relatively complete interaction process, ensure that the calculation of the fluctuation value is more comprehensive and accurate, and help analyze the performance fluctuations of the target.

[0047] The setting of the preset second determination time period helps reduce the interference of short-term data fluctuations and ensures that the calculation of fluctuation values is more stable and accurate, thereby providing a more comprehensive and reliable basis for target evaluation and improving the accuracy of analysis results and the effectiveness of decision-making.

[0048] First, by calculating the standard deviation of the real-time interaction frequency and the real-time visit count within a predetermined second time period, the interaction frequency fluctuation value and the visit count fluctuation value are obtained, respectively. Next, based on the changes in these two fluctuation values, it is determined which second temporary targets exhibit significant fluctuations. Based on the magnitude of the fluctuation values, targets requiring further processing are selected to form a number of processing targets.

[0049] By analyzing fluctuations in interaction frequency and visit counts, we can identify targets with unstable performance. This allows us to precisely locate targets requiring intervention, avoiding judgments based solely on static data. This improves assessment accuracy and intervention effectiveness, ensuring that learning targets receive appropriate attention and support.

[0050] Please continue reading Figure 4 As shown, it is a logical decision diagram for determining the impact target in this embodiment; Specifically, determining several processing goals based on the interaction frequency fluctuation value and the visit number fluctuation value includes: Drawing a change curve according to the interaction frequency fluctuation value to form an interaction change curve; Drawing a change curve according to the fluctuation value of the number of visits to form an access change curve; Calculating the cosine similarity between the interaction change curve and the access change curve to form a change synchronization degree; When the change synchronization degree is less than a preset synchronization degree threshold, determining that the second temporary target is an impact target; Non-influencing targets among all the targets to be evaluated are selected as processing targets to form a plurality of processing targets.

[0051] The preset synchronization threshold is a standard value used to determine the synchronization between the interaction frequency fluctuation value and the visit number fluctuation value. It depends on the specific evaluation scenario and data characteristics and is usually set between 0.6 and 0.8. In this embodiment, it is set to 0.75, which can effectively screen out targets with large differences in interaction frequency and visit number fluctuations, ensuring that the actual situation can be accurately reflected when processing targets, thereby improving the reliability and practicality of target evaluation.

[0052] First, we plot change curves based on the interaction frequency fluctuations and visit count fluctuations, generating an interaction change curve and a visit change curve. We then calculate the cosine similarity of these two change curves to determine the degree of change synchronization. If the degree of change synchronization is less than a preset synchronization threshold, the second temporary target is identified as an influencing target. Finally, we select non-influencing targets from the targets to be evaluated as processing targets, forming a number of processing targets.

[0053] By quantifying the fluctuations in interaction frequency and number of visits and calculating the degree of synchronization of their changes, we can effectively distinguish targets with different performance fluctuations and mark targets that do not meet the synchronization requirements as influencing targets, thereby optimizing subsequent processing and improving the accuracy of evaluation results.

[0054] Specifically, determining the target evaluation index according to the real-time access count, the real-time accuracy, and the real-time interaction frequency of each processing target includes: Normalizing the real-time access times to form a normalized access time; Normalizing the real-time accuracy to form a normalized accuracy; Normalizing the real-time interaction frequency to form a normalized interaction frequency; The normalized number of visits, the normalized accuracy, the normalized interaction frequency, the preset number of visits weight, the preset accuracy weight, and the preset interaction frequency weight are weighted and summed to form a target evaluation index.

[0055] The preset visit count weight is used to indicate the degree of influence of the visit count on the target evaluation index. It depends on the business needs of the target and the importance of the visit count. It is usually set between 0 and 1. In this embodiment, it is set to 0.3, which can reflect the importance of the visit count while avoiding over-reliance on a single indicator.

[0056] The preset accuracy weight is used to indicate the degree of influence of the accuracy on the target evaluation index. It depends on the accuracy requirement of the target task and is usually set between 0 and 1. In this embodiment, it is set to 0.4, which can emphasize the importance of accuracy and ensure that accuracy is fully considered during task evaluation.

[0057] The preset interaction frequency weight is used to indicate the degree of influence of the interaction frequency on the target evaluation index. It depends on the actual significance of the interaction behavior to the target and is usually set between 0 and 1. In this embodiment, it is set to 0.3, which can balance the influence of the interaction frequency and ensure that the interaction factor in the target evaluation is not ignored.

[0058] First, we weight the number of real-time visits, real-time accuracy, and real-time interaction frequency, taking into account their respective preset weights. Then, we calculate the target evaluation index for each processing target through weighted summation, comprehensively reflecting the target's performance across various indicators.

[0059] The weighted summation method can integrate multi-dimensional data and assign a reasonable weight to each indicator, making the target evaluation index more accurate and objective. Through this index, a comprehensive assessment of the targets can be carried out, which helps decision makers identify and prioritize those targets with outstanding performance.

[0060] Specifically, determining the analysis evaluation index based on a preset norm reference value, the real-time accuracy of all targets to be evaluated, the total number of real-time visits, the total real-time interaction frequency, and a preset cluster analysis model includes: Sorting all the real-time accuracy rates and determining the ranking value of each target to be evaluated; Calculating the sum of the ranking value and the preset norm reference value to obtain a ranking range lower limit, calculating the difference between the ranking value and the preset norm reference value to obtain a ranking range upper limit, thereby forming a ranking range; An analysis evaluation index is determined according to the real-time accuracy, the real-time number of visits, and the real-time interaction frequency of all the targets to be evaluated within the ranking range.

[0061] First, the real-time accuracy of all targets to be evaluated is ranked according to the preset norm reference value to determine the ranking of each target. Next, the sum of each ranking value and the preset norm reference value is calculated to obtain the lower limit of the ranking range, and the difference between them is calculated to determine the upper limit, forming the ranking range. Within this range, the analysis and evaluation index is finally determined by combining the real-time accuracy, number of visits, and interaction frequency of each target.

[0062] By combining ranking and ranking calculations with norm-referenced values, we can objectively and comprehensively assess the performance of each evaluation target, avoiding bias caused by a single criterion and ensuring a more accurate and comprehensive analysis and evaluation index. Furthermore, by leveraging the support of cluster analysis models, we can effectively improve the accuracy and reliability of evaluations across multiple dimensions.

[0063] Specifically, determining the analysis evaluation index according to the real-time accuracy, the real-time number of visits, and the real-time interaction frequency of all the targets to be evaluated within the ranking range includes: The real-time accuracy, the real-time number of visits, and the real-time interaction frequency of all the targets to be evaluated within the ranking range are input into the preset cluster analysis model to obtain the analysis evaluation index.

[0064] According to the determined ranking range, the real-time accuracy, real-time number of visits and real-time interaction frequency of all targets to be evaluated within the ranking range are used as input and brought into the preset cluster analysis model. The model will comprehensively evaluate the performance of each target based on these input data and output an analysis evaluation index.

[0065] The application of cluster analysis models can effectively integrate multiple evaluation dimensions to provide a comprehensive and accurate evaluation index. By combining real-time data, it can dynamically reflect the actual performance of the target, avoiding the limitations of a single evaluation standard, and providing a scientific basis for decision-making.

[0066] Specifically, adjusting the preset standard reference threshold according to the target evaluation index and the analysis evaluation index to form the adjusted standard reference threshold includes: Calculating the relative deviation between the target evaluation index and the analysis evaluation index to form an index deviation; When the index deviation is greater than the preset deviation threshold, the preset standard reference threshold is adjusted according to the relative deviation between the index deviation and the preset deviation threshold and a preset adjustment coefficient to form an adjusted standard reference threshold.

[0067] The preset deviation threshold is used to determine whether the difference between the target evaluation index and the analysis evaluation index is significant, thereby deciding whether the standard reference threshold needs to be adjusted. It depends on the accuracy requirements of the evaluation system and the diversity of the target group. It is usually set between 0.05 and 0.1. In this embodiment, it is set to 0.08 to balance accuracy and flexibility, and avoid excessive adjustment caused by too small a threshold, which affects stability.

[0068] First, the relative deviation between the target evaluation index and the analytical evaluation index is calculated to obtain the index deviation. When the index deviation exceeds the preset deviation threshold, the preset standard reference threshold is adjusted based on the index deviation, the relative deviation of the deviation threshold, and the preset adjustment coefficient, thereby forming a new adjusted standard reference threshold.

[0069] Based on the difference between the actual performance of the target and the evaluation index, the standard reference threshold is dynamically adjusted to ensure that the evaluation system is more accurate and flexible. By adjusting the threshold, it is possible to cope with deviations in different situations, improve the accuracy and adaptability of the evaluation, and thus optimize the decision-making process.

[0070] Specifically, determining the evaluation level of the target to be evaluated according to the target evaluation index includes: The evaluation level corresponding to the target evaluation index is searched in the preset evaluation index table to obtain the evaluation level of the target to be evaluated.

[0071] The preset evaluation index table is a pre-set standardized table that records the correspondence between different target evaluation indices and corresponding evaluation levels. It is designed according to the performance level and evaluation criteria of the target, generally set within an appropriate index range, and is usually set according to the needs of the actual application scenario.

[0072] The evaluation grade refers to the evaluation result grade determined by the value of the target evaluation index, which is usually divided from low to high according to performance. The evaluation grade is set based on the target evaluation criteria, performance requirements and business needs.

[0073] In this embodiment, the preset evaluation index table is:

[0074] First, a query is performed based on the target evaluation index, searching a pre-set evaluation index table. Based on the target evaluation index value, the corresponding evaluation grade is found and assigned to the target to be evaluated. This query method can quickly generate a specific evaluation grade for each target.

[0075] By searching the evaluation index table, we can efficiently and accurately assign the corresponding evaluation level to the target to be evaluated, reduce the interference of human factors, and ensure the consistency and objectivity of the evaluation. At the same time, the entire evaluation process has the advantages of standardization and automation, which helps to improve the transparency and efficiency of the evaluation.

[0076] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

[0077] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that the present invention is susceptible to various modifications and variations. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A method for processing educational data based on both standard reference and norm reference, characterized in that: include: Obtain the real-time visit count of each target to be evaluated in the online learning platform, the real-time accuracy rate of each target to be evaluated after the test, and the real-time interaction frequency of each target to be evaluated in the offline classroom; Determine a plurality of first temporary targets according to the real-time access times and a preset standard reference threshold; Determining a plurality of second temporary targets according to the real-time accuracy of each of the first temporary targets; determining a plurality of processing targets according to the real-time interaction frequency and the real-time visit count of each second temporary target; Determine a target evaluation index according to the real-time access count, the real-time accuracy rate, and the real-time interaction frequency of each processing target; Determining an analysis evaluation index based on a preset norm reference value, the real-time accuracy of all targets to be evaluated, the total number of real-time visits, the total real-time interaction frequency, and a preset cluster analysis model; Adjusting the preset standard reference threshold according to the target evaluation index and the analysis evaluation index to form an adjusted standard reference threshold; The target evaluation index is determined based on the adjustment standard and a reference threshold, and the evaluation level of the target to be evaluated is determined according to the target evaluation index.

2. The educational data processing method based on both standard reference and norm reference according to claim 1, characterized in that: Determining a plurality of first temporary targets according to the real-time access count and a preset standard reference threshold includes: When the number of real-time accesses is less than a preset standard reference threshold, the target to be evaluated is determined to be a first temporary target, and a plurality of first temporary targets are formed.

3. The educational data processing method based on both standard reference and norm reference according to claim 2, characterized in that: Determining a plurality of second temporary targets according to the real-time accuracy of each of the first temporary targets includes: When the real-time accuracy rate is greater than a preset accuracy rate threshold, calculating a standard deviation of the real-time accuracy rate within a preset first determined time period to form an accuracy rate fluctuation value; When the accuracy fluctuation value is greater than a preset accuracy fluctuation threshold, the first temporary target is determined to be the second temporary target, and a plurality of second temporary targets are formed.

4. The educational data processing method based on both standard reference and norm reference according to claim 3 is characterized in that: Determining a number of processing targets based on the real-time interaction frequency and the real-time visit count of each second temporary target includes: Calculating a standard deviation of the real-time interaction frequency within a preset second determined time period to form an interaction frequency fluctuation value; Calculating a standard deviation of the real-time access count within the preset second determined time period to form an access count fluctuation value; A number of processing targets are determined according to the interaction frequency fluctuation value and the visit number fluctuation value.

5. The educational data processing method based on both standard reference and norm reference according to claim 4 is characterized in that: Determining several processing objectives based on the interaction frequency fluctuation value and the visit number fluctuation value includes: Drawing a change curve according to the interaction frequency fluctuation value to form an interaction change curve; Drawing a change curve according to the fluctuation value of the number of visits to form an access change curve; Calculating the cosine similarity between the interaction change curve and the access change curve to form a change synchronization degree; When the change synchronization degree is less than a preset synchronization degree threshold, determining that the second temporary target is an impact target; Non-influencing targets among all the targets to be evaluated are selected as processing targets to form a plurality of processing targets.

6. The educational data processing method based on both standard reference and norm reference according to claim 5, characterized in that: Determining a target evaluation index according to the real-time access count, the real-time accuracy, and the real-time interaction frequency of each processing target includes: Normalizing the real-time access times to form a normalized access time; Normalizing the real-time accuracy to form a normalized accuracy; Normalizing the real-time interaction frequency to form a normalized interaction frequency; The normalized number of visits, the normalized accuracy, the normalized interaction frequency, the preset number of visits weight, the preset accuracy weight, and the preset interaction frequency weight are weighted and summed to form a target evaluation index.

7. The educational data processing method based on both standard reference and norm reference according to claim 6, characterized in that: Determining the analysis evaluation index based on a preset norm reference value, the real-time accuracy of all targets to be evaluated, the total number of real-time visits, the total real-time interaction frequency, and a preset cluster analysis model includes: Sorting all the real-time accuracy rates and determining the ranking value of each target to be evaluated; Calculating the sum of the ranking value and the preset norm reference value to obtain a ranking range lower limit, calculating the difference between the ranking value and the preset norm reference value to obtain a ranking range upper limit, thereby forming a ranking range; An analysis evaluation index is determined according to the real-time accuracy, the real-time number of visits, and the real-time interaction frequency of all the targets to be evaluated within the ranking range.

8. The educational data processing method based on both standard reference and norm reference according to claim 7, characterized in that: Determining the analysis evaluation index based on the real-time accuracy, the real-time number of visits, and the real-time interaction frequency of all the targets to be evaluated within the ranking range includes: The real-time accuracy, the real-time number of visits, and the real-time interaction frequency of all the targets to be evaluated within the ranking range are input into the preset cluster analysis model to obtain the analysis evaluation index.

9. The educational data processing method based on both standard reference and norm reference according to claim 8, characterized in that: Adjusting the preset standard reference threshold according to the target evaluation index and the analysis evaluation index to form the adjusted standard reference threshold includes: Calculating the relative deviation between the target evaluation index and the analysis evaluation index to form an index deviation; When the index deviation is greater than the preset deviation threshold, the preset standard reference threshold is adjusted according to the relative deviation between the index deviation and the preset deviation threshold and a preset adjustment coefficient to form an adjusted standard reference threshold.

10. The educational data processing method based on both standard reference and norm reference according to claim 9, characterized in that: Determining the evaluation level of the target to be evaluated according to the target evaluation index includes: The evaluation level corresponding to the target evaluation index is searched in the preset evaluation index table to obtain the evaluation level of the target to be evaluated.

Images