Operation supervision method and device, equipment and storage medium

Abstract

The invention provides a homework supervision method and device, equipment and a storage medium, and the method comprises the steps: obtaining homework process data and homework result data of a user, and determining a homework correction result corresponding to the homework result data; based on the homework process data, the homework result data and the corresponding homework correction result, homework process evaluation and homework result evaluation are carried out; the homework evaluation result of the user is determined based on the process evaluation result and the achievement evaluation result obtained through homework process evaluation and homework achievement evaluation, homework process data, homework achievement data and a homework correction result are fused for evaluation, the one-sidedness problem that only the final correctness is concerned in a traditional scheme is solved, and the user experience is improved. The finally determined homework evaluation result can give consideration to the learning attitude and the learning effect, and the real learning state of the user can be reflected more comprehensively and objectively, so that an education basis with higher reference value is provided for teachers and parents, and accurate learning condition diagnosis and personalized tutoring can be realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, and in particular to a homework supervision method, device, equipment and storage medium. BACKGROUND

[0002] Homework is a key link in education and teaching, and students completing homework can consolidate and deepen their understanding and mastery of knowledge points. For a long time, homework correction is mainly manual correction. In recent years, the development of information technology has led to the rise of machine homework correction schemes, such as automatic marking of multiple-choice questions and intelligent scoring of compositions, which have assisted teaching work.

[0003] However, the existing student homework evaluation mode has obvious limitations. First, the evaluation is based on the "score-only theory", which makes it difficult for the evaluation results to fully reflect the comprehensive ability of students. Second, it mainly relies on manual correction by teachers, which is low in efficiency and easily affected by subjective factors. Third, the existing technology cannot accurately determine the root cause of student homework errors, which is not conducive to teachers adjusting teaching strategies. SUMMARY

[0004] The present application provides a homework supervision method, device, equipment and storage medium to solve the problems of the existing technology, such as evaluation based on score-only theory, low efficiency of manual correction, and difficulty in accurately determining the root cause of homework errors.

[0005] The present application provides a homework supervision method, comprising: Obtaining homework process data and homework result data of a user, and determining homework correction results corresponding to the homework result data; Based on the homework process data, the homework result data and the corresponding homework correction results, homework process evaluation and homework result evaluation are performed; Based on the process evaluation results and the result evaluation results obtained by homework process evaluation and homework result evaluation, homework evaluation results of the user are determined.

[0006] According to the homework supervision method provided by the present application, the process evaluation results include homework performance evaluation results; the homework performance evaluation results are determined based on the following steps: Based on the homework process data, the total homework time of the user is determined, as well as the distraction time corresponding to the distraction behavior in the total homework time, and the distraction time in the total homework time is determined based on the proportion of the distraction time in the total homework time, and the answering concentration is determined; Based on the homework process data, the total writing time of the user is determined, as well as the pause time corresponding to the pause action in the total writing time, and the answering coherence is determined based on the proportion of the pause time in the total writing time; Based on the answering concentration and / or the answering coherence, the homework performance evaluation results are determined.

[0007] According to a work supervision method provided by the present invention, the process evaluation result further includes a capability development evaluation result; the capability development evaluation result is determined based on the following steps: Based on the homework results data, the completeness of the user's problem-solving steps is determined, and based on the completeness of the problem-solving steps, the completeness of the user's logical thinking is determined; Based on the work process data and the work correction results, determine the number of self-corrected errors and the total number of errors corresponding to the user's work results data, and determine the comprehensiveness of the user's self-correction based on the ratio of the number of self-corrected errors to the total number of errors. The assessment results of the ability development are determined based on the completeness of logical thinking and / or the comprehensiveness of autonomous error correction.

[0008] According to a job supervision method provided by the present invention, the process evaluation result further includes a learning quality evaluation result; the learning quality evaluation result is determined based on the following steps: Based on the work process data, the user's dwell time on the preset challenge and the number of times the user checks within the total work time are determined, and the work persistence is determined based on the ratio of the dwell time to the standard dwell time. Based on the homework results data, the user's writing standardization is determined, and based on the number of checks and the writing standardization, the homework rigor is determined. Based on the homework grading results, determine the reasonableness score of non-standard answers in the user's homework data, and determine the homework innovation level based on the reasonableness score; The learning quality assessment result is determined based on at least one of the following: the persistence in completing the assignment, the rigor of completing the assignment, and the innovation of completing the assignment.

[0009] According to a homework supervision method provided by the present invention, the method for determining the focus level based on the proportion of distraction time in the total homework time further includes: Based on the work process data, the user's behavioral performance and emotional state are determined; If the behavior reflects that the user has plagiarized, the emotional state reflects that the user has excessive anxiety, or the focus of answering questions is consistently below the focus threshold, then warning data is generated based on the task process data, the process evaluation results, and the task evaluation results. The warning data will be sent to the bound guardian's terminal and / or teacher's terminal.

[0010] According to the work supervision method provided by the present invention, the early warning data includes early warning data and early warning suggestion data; The warning data includes user images corresponding to the plagiarism behavior and / or user images when experiencing excessive anxiety. The early warning and suggestion data includes at least one of the following: operation process analysis report, operation process improvement suggestions, and supplementary practice exercises.

[0011] The present invention also provides a work supervision device, comprising: The acquisition unit is used to acquire the user's work process data and work result data, and determine the work grading result corresponding to the work result data; The evaluation unit is used to evaluate the work process and the work results based on the work process data, the work result data and their corresponding work correction results. The determining unit is used to determine the user's job evaluation result based on the process evaluation result and the result evaluation result obtained from the job process evaluation and job result evaluation.

[0012] The present invention also provides a work supervision device, including a camera, a touch screen and a processor; The processor is used to acquire the work process data and work result data collected by the camera and the touch screen when the user works on the touch screen, and to determine the work grading result corresponding to the work result data; based on the work process data, the work result data and their corresponding work grading results, to perform work process evaluation and work result evaluation; based on the process evaluation result and result evaluation result obtained from the work process evaluation and work result evaluation, to determine the user's work evaluation result.

[0013] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the job supervision method as described above.

[0014] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the job supervision method as described above.

[0015] The homework supervision method, device, equipment, and storage medium provided by this invention integrate homework process data, homework result data, and homework correction results for evaluation, solving the one-sided problem of traditional solutions that only focus on the final right or wrong. Furthermore, by simultaneously evaluating the homework process and homework result, the final homework evaluation result can take into account both learning attitude and learning effect, and can more comprehensively and objectively reflect the user's true learning status, thereby providing teachers and parents with more valuable educational reference, and thus helping to achieve accurate learning diagnosis and personalized tutoring. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating the work supervision method provided by the present invention; Figure 2 This is a schematic diagram of the work supervision device provided by the present invention; Figure 3 This is a schematic diagram of the operation monitoring equipment provided by the present invention; Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

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

[0019] Homework, as an indispensable part of the education and teaching process, is assigned by teachers to students based on their need to master knowledge points, enabling them to further apply what they have learned to solve problems. By completing homework, students can further consolidate and deepen their understanding of knowledge points, thereby achieving better mastery of them.

[0020] For a long time, homework was mainly graded manually by teachers. In recent years, with the continuous advancement of information technology, various machine-based homework grading solutions have emerged in the education field, such as automatic grading of multiple-choice and true / false questions, and intelligent scoring of essays. These technologies have assisted teaching to some extent.

[0021] However, existing student homework evaluation models still have significant limitations. Firstly, the evaluation focus is overly concentrated on the score of the homework, neglecting the learning behaviors, thinking patterns, and learning qualities during the homework process, leading to a "score-only" tendency and making it difficult to comprehensively reflect students' overall abilities. Secondly, the method of mainly relying on teachers to manually grade homework is not only inefficient but also easily influenced by subjective factors, making it difficult to provide personalized feedback. Thirdly, existing technologies generally lack effective supervision and data collection of the homework process, making it impossible to trace the root cause of errors—for example, whether the errors are due to a lack of knowledge or factors such as distraction and anxiety—which hinders teachers from adjusting their teaching strategies accordingly.

[0022] In response, this invention provides a method for monitoring work, which aims to break away from the single-dimensional evaluation model by analyzing work process data and work result data. It incorporates considerations such as learning behavior and thinking patterns into the evaluation system, thereby achieving effective monitoring and comprehensive evaluation of the entire user's work process. This allows for tracing the root causes of errors and providing accurate personalized feedback, comprehensively reflecting the user's overall capabilities.

[0023] Figure 1 This is a flowchart illustrating the work supervision method provided by the present invention. This method can be applied to work supervision equipment, which can be an electronic device with display and human-computer interaction functions, such as a smart tablet, learning machine, or learning robot. Figure 1 As shown, the method includes: Step 110: Obtain the user's work process data and work result data, and determine the work grading result corresponding to the work result data; Step 120: Based on the work process data, work result data and their corresponding work correction results, conduct work process evaluation and work result evaluation; Step 130: Based on the process evaluation results and outcome evaluation results obtained from the work process evaluation and work outcome evaluation, determine the user's work evaluation result.

[0024] Specifically, when a user is working on the monitoring device, the device collects real-time data about the user's work process using built-in sensor components such as a camera, microphone, and touchscreen sensors. This data not only records the time dimension of the user's work, such as the start and end times, total duration, total writing time, and pauses during writing, but also focuses on recording the user's behavior and emotional state during the work process. For example, facial expressions captured by the camera (such as frowning or biting a pen indicating anxiety), head posture (such as frequent head turning or prolonged gaze away from the work area), and body movements (such as fiddling with stationery), as well as audio data collected by the microphone. This data can dynamically reflect the user's concentration and emotions during the work process.

[0025] At the same time, the device will also acquire the user's work result data, that is, the final content presented after the work is completed. Here, the work result data can be obtained by taking a picture of the work interface on the touch screen and performing optical character recognition; it can also be obtained by directly acquiring the electronic handwriting data written by the user with a stylus on the touch screen, or by acquiring the data directly input by the user with writing tools such as a mouse or keyboard. This embodiment of the invention does not specifically limit this.

[0026] After obtaining the homework results data, in this embodiment of the invention, the device's built-in intelligent model, such as a pre-trained and deployed homework grading model, or by directly calling a multimodal large model, a large language model, or an artificial intelligence large model, can be used to analyze the homework results data to achieve homework grading and obtain the homework grading results. These results reflect the correctness, scores, and knowledge point coverage of each homework question in the homework results data.

[0027] Furthermore, the user's work process and work results can be evaluated using work process data, work result data, and work correction results, respectively, to obtain corresponding process evaluation results and result evaluation results.

[0028] Specifically, the purpose of this process assessment is to transform unstructured process data into quantifiable evaluation indicators to measure users' non-intellectual factors. For example, analyzing eye movements and gestures in the process data can calculate the percentage of distraction time, thus assessing the user's focus; analyzing the frequency and duration of pauses in writing can assess the coherence of answers; facial expressions can assess the user's emotional stability when facing difficult problems; analyzing problem-solving steps in the results can assess the completeness and comprehensiveness of the user's logical thinking; and analyzing grading results, using scores for non-standard answers, can assess the user's creativity. Through these analyses, process assessment results reflecting the user's learning attitude and ability can be obtained.

[0029] Correspondingly, homework performance evaluation can be conducted by combining homework results data and grading results. This evaluation is no longer limited to simple right or wrong judgments, but rather involves in-depth analysis of the user's knowledge mastery rate and error repetition rate by analyzing the content of the homework results data, such as the total number of questions, and the error patterns in the grading results. For example, the number of questions answered correctly and the total number of questions can be identified to calculate the knowledge mastery rate. Similarly, the number of repeated errors and the total number of errors can be identified to calculate the user's error repetition rate. This analysis yields an evaluation result that reflects the depth and detail of the user's knowledge mastery.

[0030] After this, the process evaluation results and outcome evaluation results can be integrated to generate the final assignment evaluation result. That is, different weights can be assigned to the process evaluation results and outcome evaluation results. For example, the weight ratio of process and outcome can be dynamically adjusted according to the different focuses of quality education, and a comprehensive score (such as 0-10 points) and / or comprehensive grade (such as excellent, good, improved, need improvement, etc.) can be obtained through weighted calculation.

[0031] In addition, the results of homework assessment can also be presented as a multi-dimensional comprehensive assessment profile or a recommendation report. This report not only includes the mastery of knowledge points based on the outcome assessment results, such as strong mathematical calculation ability, but also includes feedback on learning behavior, thinking style and learning quality based on the process assessment results, such as a significant decrease in concentration in the second half of the homework, thus forming an objective assessment of the user's overall homework.

[0032] The homework supervision method provided by this invention integrates homework process data, homework result data, and homework correction results for evaluation, solving the one-sided problem of traditional solutions that only focus on the final right or wrong. Furthermore, by taking a two-pronged approach of evaluating the homework process and homework result, the final homework evaluation result can take into account both learning attitude and learning effect, and can more comprehensively and objectively reflect the user's true learning status. This provides teachers and parents with more valuable educational reference, thereby helping to achieve accurate learning diagnosis and personalized tutoring.

[0033] Based on the above embodiments, the process evaluation results include job performance evaluation results; the job performance evaluation results are determined based on the following steps: Based on the data from the homework process, the user's total homework time and the distraction time corresponding to distraction behaviors within the total homework time are determined. Based on the proportion of distraction time in the total homework time, the user's focus level is determined. Based on the data from the writing process, the user's total writing time and the pause duration corresponding to the pause actions within the total writing time are determined, and the continuity of the answer is determined based on the proportion of the pause duration in the total writing time. The assessment results for assignment performance are determined based on the student's focus and / or fluency in answering questions.

[0034] Specifically, process evaluation results may include job performance evaluation results, which can be determined through the following process: First, the total duration of the user's current task can be determined from the timestamp information of the task process data, that is, the total time from the start to the end. Next, during this period, visual analysis algorithms (such as facial landmark detection, body motion recognition, etc.) or motion sensor data can be used to identify the user's distracting behaviors. Distracting behaviors here refer to non-learning-oriented actions exhibited by the user during the task, such as frequently turning their head to talk, taking their gaze away from the task area for extended periods, fiddling with stationery, getting up from their seat, or sleeping on the desk. By statistically analyzing the duration of these distracting behaviors, the distraction duration can be obtained.

[0035] Then, the percentage of distraction time in the total task time can be calculated. This percentage directly reflects the user's attention loss, and the user's focus level in answering questions can be determined based on this percentage.

[0036] For example, inverse proportional logic can be used for calculation, such as the following formula: Answering focus rate = (Total homework time - Distraction time) / Total homework time × 10; The answering focus level is calculated, and at this point, the answering focus level is a quantitative indicator of 0-10 points.

[0037] For example, when the percentage is below a preset threshold, such as 5%, the focus on answering questions can be judged as high; conversely, when the percentage is above the threshold, such as 20%, the focus on answering questions can be judged as low.

[0038] Simultaneously, the writing trajectory characteristics in the task process data can be analyzed to calculate the user's answer continuity. Specifically, this can involve first determining the user's total writing time. It should be noted that this total writing time can be equivalent to the total task time, or it can be the cumulative time the user spends writing. Next, pauses during the writing process can be detected. These pauses refer to abnormal stillness that occurs during what should be continuous writing, such as the pen tip remaining suspended in the air for more than a preset time, such as 10 seconds, or an excessively long time interval between two pen strokes. The duration of these pauses is obtained by summing up the times of these pause segments.

[0039] Then, the percentage of pause time within the total writing time can be calculated. This percentage objectively reflects the fluency or hesitation of the user's thinking while answering questions. Based on this percentage, the user's answer coherence can be determined.

[0040] For example, the coherence of the answer can be calculated using the following formula: Answer coherence = (Total writing time - Pause time) / Total writing time × 10.

[0041] A high degree of fluency in answering questions usually means that the user has a good grasp of the knowledge points and thinks smoothly; while a low degree of fluency in answering questions, i.e., a high percentage of pauses, may indicate that the user is stuck, distracted, or has a mental block during the problem-solving process.

[0042] Finally, the homework performance evaluation result can be comprehensively determined based on the calculated focus and / or fluency of answering questions. This means that the homework performance evaluation result can be a weighted total score of focus and fluency of answering questions, for example, focus of answering questions accounts for 60% of the weight and fluency of answering questions accounts for 40% of the weight; or it can be a combined rating of the two, which is not specifically limited in this embodiment of the invention. Through this process, the abstract learning attitude can be transformed into an objective evaluation index composed of focus and fluency of answering questions, thereby completing a precise profile of the user's homework performance dimensions.

[0043] In this embodiment of the invention, by introducing two core quantitative indicators—answer focus and answer fluency—the problem of traditional homework assessments being unable to quantify learning attitudes and relying solely on subjective feelings is solved. By calculating the percentage of distraction time and the percentage of writing pauses, the system can objectively capture the user's attention span and thought flow during the homework process, thereby effectively distinguishing whether the user's problem stems from insufficient ability or a poor attitude, and thus providing a scientific basis for subsequent targeted educational interventions.

[0044] Based on the above embodiments, the process evaluation results also include capability development evaluation results; the capability development evaluation results are determined based on the following steps: Based on the homework results data, determine the completeness of the user's problem-solving steps, and based on the completeness of the problem-solving steps, determine the completeness of the user's logical thinking. Based on the work process data and the work correction results, determine the number of self-corrected errors and the total number of errors corresponding to the user's work results data, and determine the comprehensiveness of the user's self-correction based on the ratio of the number of self-corrected errors to the total number of errors. The assessment results for ability development are determined based on the completeness of logical thinking and / or the comprehensiveness of self-correction.

[0045] Specifically, in addition to focusing on whether users are focused and fluent in answering questions, this embodiment of the invention also uses the results of ability development assessments to deeply measure the depth of users' thinking and self-reflection abilities, in order to solve the problem in traditional assignment assessments that cannot distinguish whether users are "guessing correctly" or truly understanding the material. Here, the ability development assessment results can be determined through the following steps: First, the completeness of logical thinking can be determined by analyzing the level of detail in the user's problem-solving steps based on the homework results data. Specifically, this can be achieved by using image recognition or text analysis techniques to perform structured parsing of the user's homework results data to identify the intermediate deduction processes before the answer is generated, thereby determining the completeness of the problem-solving steps. The completeness of the problem-solving steps reflects the thoroughness of the user's reasoning process when answering the question.

[0046] For example, in a math word problem, it's not enough to simply identify the final answer; it's also necessary to identify whether key steps such as formulas were presented, substitutions were performed, and unit conversions were made are present. If the completed work only contains the final answer and lacks intermediate steps, the completeness of the solution steps is considered low; conversely, if the steps are detailed and the logical chain is clear, the completeness of the solution steps is considered high. Based on this analysis, the completeness of the solution steps can be mapped to the completeness of logical thinking.

[0047] For example, the completeness of logical thinking can be directly calculated using the following formula: Logical thinking completeness score = Step completeness score × 10.

[0048] Logical thinking completeness can objectively reflect the logical rigor and depth of thinking when a user answers a question, thus distinguishing between guessing correctly by luck and answering correctly by ability.

[0049] Simultaneously, modification traces can be analyzed based on the work process data and the work grading results to determine the comprehensiveness of self-correction. Specifically, this involves tracing back the modification actions recorded in the work process data, such as erasing actions, correction marks, and undo operations, to identify the number of times the user independently discovered and corrected errors during the work period, and counting these as the number of self-corrected errors. Furthermore, the total number of errors generated by the user in this work can be determined based on the work grading results. This total number of errors can be understood as the sum of the number of self-corrected errors and the number of errors indicated in the work grading results, representing all deviations actually generated by the user during this work process.

[0050] Furthermore, the ratio of self-corrected errors to the total number of errors can be calculated. This ratio represents the comprehensiveness of the user's self-correction. For example, if a user makes 5 errors in this assignment but finds and corrects 4 of them, it indicates that they have a very strong ability to self-check and reflect; conversely, if a user makes a total of 5 errors but does not correct any of them, it indicates that they lack self-correction ability.

[0051] Finally, the ability development assessment result can be determined based on the completeness of logical thinking and / or the comprehensiveness of self-correction. For example, the completeness of logical thinking and the comprehensiveness of self-correction can be weighted and summed, such as logical thinking completeness accounting for 50% and self-correction comprehensiveness accounting for 50%. By performing a weighted sum according to this weighting, the final ability development assessment result can be obtained. Through the above process, abstract subject-specific abilities can be concretized into the completeness of logical thinking, which reflects the depth of the user's thinking, and the comprehensiveness of self-correction, which reflects the rigor of the user's thinking.

[0052] In this embodiment of the invention, by introducing the completeness of logical thinking and the comprehensiveness of self-correction, the homework assessment system is elevated from simply mastering knowledge points to developing thinking ability. Furthermore, by identifying the completeness of problem-solving steps and self-correction behavior, it is possible to effectively uncover the user's thinking potential and shortcomings in learning habits, such as a lack of checking habits and serious skipping of steps, thereby providing data-driven guidance and suggestions for cultivating the user's logical thinking and cognitive abilities.

[0053] Based on the above embodiments, the process evaluation results also include learning quality evaluation results; the learning quality evaluation results are determined based on the following steps: Based on the task process data, determine the time the user spends on the preset challenge and the number of times the user checks during the total task completion time. Based on the ratio of the time spent on the challenge to the standard time spent on the challenge, determine the task persistence. Based on the homework results data, the user's writing standardization is determined, and based on the number of checks and writing standardization, the homework rigor is determined; Based on the homework grading results, determine the reasonableness score of non-standard answers in the user's homework data, and determine the homework innovation level based on the reasonableness score; The learning quality assessment result is determined based on at least one of the following: persistence in completing assignments, rigor in completing assignments, and innovation in completing assignments.

[0054] Specifically, in addition to focusing on performance and ability in answering questions, this embodiment of the invention also considers the user's non-intellectual factors through learning quality assessment results, primarily evaluating the user's resilience in the face of difficulties, their diligence in completing tasks, and the activity of their thinking. Here, the learning quality assessment results can be determined through the following steps: First, user behavior patterns on challenging problems can be analyzed based on the data collected during the assignment process to determine the user's persistence level. Specifically, this involves identifying pre-set challenging problems in the assignment; these could be questions marked as difficult in a question bank or challenging questions determined based on the user's historical performance. In this embodiment of the invention, the time users spend on these pre-set challenging problem pages or areas can be accurately recorded during the assignment. This time spent reflects whether the user immediately gives up or tries to solve the problem when encountering difficulties.

[0055] Next, the duration of this stay will be compared with the standard duration of stay, such as the average time taken by peers to solve the problem or the preset reasonable thinking time, and the ratio of the two will be calculated to determine the task persistence.

[0056] For example, if the ratio is within a reasonable range, such as when a user persists in thinking for a relatively long time, then their persistence in completing the task is considered high; if the ratio is extremely low, such as when the user quits immediately, then their persistence in completing the task is considered low.

[0057] For example, the work persistence rate can be directly calculated using the following formula: Work persistence = (duration of stay / standard duration of stay) × 10.

[0058] It should be noted here that if the ratio of the dwell time to the standard dwell time is too high, such as exceeding 1 to reach 2 or even 3, it can be determined that the user has spent too much time on the challenge and points should be deducted.

[0059] In addition, when analyzing the data during the task completion process, the system also monitors and counts the number of checks performed by the user within the total task completion time. The number of checks can be determined by recognizing eye movement patterns, page-turning actions, and secondary touches on already answered areas.

[0060] Subsequently, the visual characteristics and number of checks of the assignment results data can be combined to determine the rigor of the assignment. Specifically, this can be done by first analyzing the assignment results data, such as analyzing the neatness of the handwriting and the cleanliness of the paper (e.g., whether there are large areas of scribbling, the alignment of formats, etc.), to determine the user's writing standardization; then, this writing standardization and the number of checks can be used to determine the rigor of the assignment.

[0061] For example, the rigor of an assignment can be calculated using the following formula: Assignment rigor = (Number of checks × Writing standardization) / Number of pages × 10.

[0062] The rigor of the work can effectively distinguish between a careless and sloppy work attitude and a neat and meticulous work attitude, thus better reflecting the user's seriousness towards the work.

[0063] Furthermore, the level of innovation in assignments can be determined based on the grading results. Specifically, for subjective or open-ended questions, when a user's answer differs from the standard answer, the user's answer can be further evaluated to assess its logical coherence, uniqueness of viewpoint, and feasibility of method, thereby deriving a reasonableness score for the non-standard answer. Based on this reasonableness score, the user's level of innovation can be determined; that is, the reasonableness score can be directly considered as the user's level of innovation. For example, if a user uses a logically correct problem-solving method not found in the textbook, or presents a novel viewpoint in their essay, a higher reasonableness score will be given to encourage creative thinking and avoid solely relying on standard answers.

[0064] Finally, the learning quality assessment result can be determined based on at least one of the following calculated metrics: homework persistence, homework rigor, and homework innovation. This means that, based on actual teaching needs, the assessment result of a certain dimension can be selectively output, or a "quality radar chart" reflecting the user's comprehensive learning literacy can be generated through weighted calculations (e.g., homework persistence accounts for 30% weight, homework rigor accounts for 40% weight, and homework innovation accounts for 30% weight).

[0065] In this embodiment of the invention, by quantifying a user's persistence in tackling difficult problems, their rigor in writing and checking, and their innovation in answering questions, a comprehensive profile of the user's overall qualities is achieved. This not only helps to discover the user's strengths in non-intellectual factors, such as having many wrong answers but also having innovative ideas, but also helps to promptly identify potential habitual problems, such as giving up when encountering difficult problems or writing sloppily without checking, thereby truly implementing the assessment concept of quality education.

[0066] Based on the above embodiments, the process of evaluating work results may specifically include: First, based on the homework results data and homework correction results, the total number of questions in this homework assignment and the number of questions answered correctly by the user can be determined; then, the ratio between the two can be calculated to determine the knowledge point mastery rate.

[0067] For example, the knowledge point mastery rate can be calculated using the following formula: Knowledge point mastery rate = (number of correct questions / total number of questions) × 10.

[0068] The knowledge point mastery rate here directly reflects the user's familiarity with the knowledge points involved in this assignment.

[0069] At the same time, based on the homework correction results, the total number of errors in the user's current homework and the number of duplicate errors can be identified, and the error duplication rate can be calculated accordingly.

[0070] For example, the error repetition rate can be calculated using the following formula: Error repetition rate = (1 - number of repeated errors / total number of errors) × 10.

[0071] Error repetition rate is used to measure whether users have truly managed to avoid repeating the same mistake.

[0072] Finally, the evaluation results can be determined based on the knowledge point mastery rate and / or error repetition rate calculated above. For example, a weighted summation method can be used, such as assigning 70% weight to the knowledge point mastery rate and 30% weight to the error repetition rate, to calculate the evaluation results.

[0073] Furthermore, the final job evaluation result can be calculated based on the process evaluation result and the outcome evaluation result.

[0074] Specifically, after obtaining the three evaluation results (work performance, ability development, and learning quality) and the outcome evaluation results in this embodiment of the invention, a final fusion calculation can be performed; that is, a multi-dimensional weighted strategy is used to calculate the comprehensive score. Preferably, in this embodiment of the invention, the weight allocation strategy is as follows: Assignment performance evaluation results: weighted at 25%, focusing on assessing focus and consistency; The assessment results for competency development are weighted at 30%, focusing on the assessment of logical thinking and error correction abilities. Learning quality assessment results: weighted at 25%, focusing on assessing perseverance, rigor, and innovation; Results evaluation: weighted at 20%, focusing on the evaluation of work results.

[0075] Based on the above weights, the overall score can be calculated using the following formula: Overall score = Homework performance assessment result × 25% + Ability development assessment result × 30% + Learning quality assessment result × 25% + Outcome assessment result × 20%.

[0076] Finally, to provide intuitive feedback, the overall score can be mapped to a specific grade as the final assignment evaluation result. For example: If the overall score is 9-10, the grade is determined to be excellent; If the overall score is 7-8 points, the grade is determined to be good; If the overall score is 5-6 points, the level is determined to be "improvement". If the overall score is below 5, the level is determined to be "needs to be improved".

[0077] In this embodiment of the invention, the weight of the traditional "score-only" approach is reduced by quantitative evaluation, while the weight of processes, abilities, qualities, etc., that reflect the user's comprehensive qualities is increased. This can guide users to pay more attention to the development of learning habits and the improvement of thinking abilities, rather than just focusing on test scores, thereby truly achieving a scientific evaluation oriented towards quality education.

[0078] Based on the above embodiments, the focus level for answering questions is determined based on the proportion of distraction time in the total task time, and then the process further includes: Based on the data from the work process, determine the user's behavioral performance and emotional state; If a user's behavior reflects plagiarism, their emotional state reflects excessive anxiety, or their focus on answering questions is consistently below the focus threshold, warning data will be generated based on the task process data, process evaluation results, and task evaluation results. The warning data will be sent to the linked guardian's terminal and / or teacher's terminal.

[0079] Specifically, after the completion of the assignment process evaluation, in order to prevent the solidification of bad learning habits or the emergence of psychological problems in a timely manner, this embodiment of the invention introduces a real-time monitoring and early warning mechanism for abnormal behavior.

[0080] In detail, this could involve first determining the user's behavior and emotional state based on the data from the assignment process. Specifically, in terms of behavior, visual analysis algorithms, such as motion recognition algorithms, are used to analyze the user's movement trajectories. For example, monitoring the user's head and eye movement patterns can identify abnormal actions such as frequent head turning (possibly looking at classmates' answers) or frequent operation of external electronic devices (possibly searching for answers), thus determining the user's behavioral performance. In terms of emotion, facial expression recognition technology is used to capture the user's micro-expression features, such as frowning or biting a pen, reflecting anxiety, thus determining the user's emotional state.

[0081] Next, it is necessary to determine whether to activate the early warning mechanism based on preset trigger conditions. The specific judgment logic includes at least one of the following three situations: Plagiarism detection: When the behavior reflects that the user has plagiarized, for example, if the frequency of the action of "turning the head to the left and staying there" exceeds 5 times per minute and the focus of the gaze is not in the current work area; Excessive anxiety determination: When the emotional state reflects that the user is excessively anxious, for example, when painful / anxious expressions are detected to appear continuously for more than 3 minutes, or when high-frequency sighing sounds are detected. Extreme Distraction Detection: When the focus level during answering questions remains below the preset focus threshold, for example, when the focus level is below 5 points for 5 minutes.

[0082] If any of the above situations occur, the early warning mechanism will be triggered immediately. That is, the system will comprehensively retrieve work process data, process evaluation results, and work evaluation results, and generate early warning data accordingly. This early warning data is not just a simple alarm signal; it also includes slices of records of abnormal user behavior, such as screenshots or short videos, data analysis reports, such as "focus level was only 10% in the past 20 minutes," improvement suggestions, and supplementary practice exercises.

[0083] Finally, the generated warning data can be sent in real-time or near real-time to the linked guardian's terminal, such as the parent's mobile application, and / or the teacher's terminal, such as the teacher's management backend, thus realizing a closed loop of home-school communication. After receiving the warning, parents can intervene in a timely manner, such as reminding the user to take a break or asking if they have encountered any difficulties. After receiving the warning, teachers can understand the user's actual homework status, such as "the user is struggling with this question, perhaps because the knowledge point was not explained clearly," thereby avoiding blaming the user simply for homework errors.

[0084] In this embodiment of the invention, by constructing an active intervention mechanism that monitors, judges, and sends early warnings, plagiarism, anxiety, and long-term distraction can be identified in a timely manner, and the early warning data can be promptly delivered to parents and teachers. This can effectively prevent the development of bad study habits and promptly address users' mental health issues.

[0085] Based on the above embodiments, the early warning data includes early warning data and early warning suggestion data; Warning data includes user images corresponding to plagiarism and / or user images showing excessive anxiety; The early warning and recommendation data includes at least one of the following: work process analysis report, work process improvement suggestions, and supplementary practice exercises.

[0086] Specifically, in practical applications, simple alarm prompts, such as "Your child / student is distracted," may trigger anxiety in parents or conflict between teachers and students. To address this issue, the early warning data generated in this embodiment of the invention is a composite data package containing objective evidence and solutions, specifically including early warning data and early warning suggestion data.

[0087] In detail, to ensure the objectivity and traceability of early warnings, warning data containing intuitive visual evidence can be generated. Specifically, when plagiarism and / or excessive anxiety are detected, a simple text message won't be sent. Instead, the video stream will be automatically reviewed to accurately capture the moment the abnormal behavior occurred. For example, if a user is detected frequently turning their head to peek at others, the image of the user corresponding to this plagiarism behavior will be captured; if a user is detected frowning for a long time with a pained expression, the image of the user under this excessive anxiety will be captured. These images constitute the main body of the warning data, allowing parents or teachers to see for themselves and intuitively confirm the situation at the time, thereby avoiding ineffective communication due to misjudgment.

[0088] Accordingly, to achieve closed-loop feedback in education and truly help users solve problems, in-depth analysis of assignment data can generate instructive early warning and suggestion data. This early warning and suggestion data can include at least one of the following three types of content: Homework Process Analysis Report: This is a visual data chart. It summarizes various indicators during the homework process, such as changes in focus and fluctuations in answer consistency, into a report. For example, the report can visually show "focus was all green for the first 20 minutes, then dropped sharply starting at the 25th minute, accompanied by anxiety," helping parents understand the trajectory of the problem.

[0089] Suggestions for improving the homework process: These are specific action guidelines given for abnormal behavior. For example, for excessive anxiety, the suggestion might be "The user may have encountered difficulties. Parents are advised to offer encouragement or pause the homework for a 5-minute break." For plagiarism tendencies, the suggestion might be "Please check for distractions in the surroundings and complete the work independently."

[0090] Supplementary practice exercises are designed to address knowledge gaps. For example, if a user's anxiety or pause is due to a weak grasp of a specific knowledge point, several appropriately challenging and targeted supplementary practice exercises can be automatically generated through knowledge graph connections. These exercises are pushed to parents and teachers as part of the early warning and suggestion data, helping users rebuild confidence and master the knowledge point through tiered practice with reduced difficulty, thereby eliminating the source of anxiety.

[0091] In this embodiment of the invention, the early warning data provides objective factual basis, reducing communication costs and misunderstandings; while the early warning suggestion data provides scientific solutions, enabling parents and teachers to not only know what happened after receiving the warning, but also why and what to do, thereby truly realizing effective intervention and personalized assistance in the user's homework process.

[0092] The work supervision device provided by the present invention is described below. The work supervision device described below and the work supervision method described above can be referred to in correspondence.

[0093] Figure 2 This is a schematic diagram of the work supervision device provided by the present invention, as shown below. Figure 2 As shown, the device includes: The acquisition unit 210 is used to acquire the user's work process data and work result data, and determine the work grading result corresponding to the work result data; Evaluation unit 220 is used to evaluate the work process and the work results based on the work process data, the work result data and their corresponding work correction results; The determining unit 230 is used to determine the user's job evaluation result based on the process evaluation result and the result evaluation result obtained from the job process evaluation and job result evaluation.

[0094] The homework supervision device provided by this invention integrates homework process data, homework result data, and homework correction results for evaluation, solving the one-sided problem of traditional solutions that only focus on the final right or wrong. Furthermore, by simultaneously evaluating the homework process and homework result, the final homework evaluation result can take into account both learning attitude and learning effect, and can more comprehensively and objectively reflect the user's true learning status. This provides teachers and parents with more valuable educational reference, thereby helping to achieve accurate learning diagnosis and personalized tutoring.

[0095] Based on the above embodiments, the process evaluation results include job performance evaluation results; Evaluation unit 220 is used for: Based on the work process data, the user's total work time and the distraction time corresponding to distraction behaviors within the total work time are determined, and the answering focus is determined based on the proportion of the distraction time in the total work time; Based on the work process data, the user's total writing time and the pause time corresponding to the pause action within the total writing time are determined, and the answer continuity is determined based on the proportion of the pause time in the total writing time. The assessment result of the assignment performance is determined based on the aforementioned focus and / or consistency of answering questions.

[0096] Based on the above embodiments, the process evaluation results also include capacity development evaluation results; Evaluation unit 220 is used for: Based on the homework results data, the completeness of the user's problem-solving steps is determined, and based on the completeness of the problem-solving steps, the completeness of the user's logical thinking is determined; Based on the work process data and the work correction results, determine the number of self-corrected errors and the total number of errors corresponding to the user's work results data, and determine the comprehensiveness of the user's self-correction based on the ratio of the number of self-corrected errors to the total number of errors. The assessment results of the ability development are determined based on the completeness of logical thinking and / or the comprehensiveness of autonomous error correction.

[0097] Based on the above embodiments, the process evaluation results also include learning quality evaluation results; Evaluation unit 220 is used for: Based on the work process data, the user's dwell time on the preset challenge and the number of times the user checks within the total work time are determined, and the work persistence is determined based on the ratio of the dwell time to the standard dwell time. Based on the homework results data, the user's writing standardization is determined, and based on the number of checks and the writing standardization, the homework rigor is determined. Based on the homework grading results, determine the reasonableness score of non-standard answers in the user's homework data, and determine the homework innovation level based on the reasonableness score; The learning quality assessment result is determined based on at least one of the following: the persistence in completing the assignment, the rigor of completing the assignment, and the innovation of completing the assignment.

[0098] Based on the above embodiments, the device further includes an early warning unit, used for: Based on the work process data, the user's behavioral performance and emotional state are determined; If the behavior reflects that the user has plagiarized, the emotional state reflects that the user has excessive anxiety, or the focus of answering questions is consistently below the focus threshold, then warning data is generated based on the task process data, the process evaluation results, and the task evaluation results. The warning data will be sent to the bound guardian's terminal and / or teacher's terminal.

[0099] Based on the above embodiments, the early warning data includes early warning data and early warning suggestion data; The warning data includes user images corresponding to the plagiarism behavior and / or user images when experiencing excessive anxiety. The early warning and suggestion data includes at least one of the following: operation process analysis report, operation process improvement suggestions, and supplementary practice exercises.

[0100] The present invention also provides a work supervision device. Figure 3 This is a schematic diagram of the work supervision equipment provided by the present invention, as shown below. Figure 3As shown, the device includes a camera 310, a touch screen 320, and a processor 330; The processor 330 is used to acquire the work process data and work result data collected by the camera 310 and the touch screen 320 when the user works on the touch screen 320, and to determine the work grading result corresponding to the work result data; based on the work process data, the work result data and their corresponding work grading results, to perform work process evaluation and work result evaluation; based on the process evaluation result and result evaluation result obtained from the work process evaluation and work result evaluation, to determine the user's work evaluation result.

[0101] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4 As shown, the electronic device may include a processor 410, a communications interface 420, a memory 430, and a communication bus 440, wherein the processor 410, communications interface 420, and memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions in the memory 430 to execute a job supervision method, which includes: acquiring user job process data and job result data, and determining the job grading result corresponding to the job result data; performing job process evaluation and job result evaluation based on the job process data, the job result data, and their corresponding job grading results; and determining the user's job evaluation result based on the process evaluation result and result evaluation result obtained from the job process evaluation and job result evaluation.

[0102] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0103] On the other hand, the present invention also provides a computer program product, the computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, wherein when the program instructions are executed by a computer, the computer is able to execute the job supervision method provided by the above methods, the method comprising: acquiring user job process data and job result data, and determining the job grading result corresponding to the job result data; performing job process evaluation and job result evaluation based on the job process data, the job result data and their corresponding job grading result; and determining the user's job evaluation result based on the process evaluation result and result evaluation result obtained from the job process evaluation and job result evaluation.

[0104] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the job supervision method provided by the above methods. The method includes: acquiring user job process data and job result data, and determining the job grading result corresponding to the job result data; performing job process evaluation and job result evaluation based on the job process data, the job result data and their corresponding job grading results; and determining the user's job evaluation result based on the process evaluation result and result evaluation result obtained from the job process evaluation and job result evaluation.

[0105] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0106] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0107] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring work, characterized in that, include: Obtain the user's work process data and work result data, and determine the work grading result corresponding to the work result data; Based on the work process data, the work result data and their corresponding work correction results, conduct work process evaluation and work result evaluation; Based on the process evaluation results and outcome evaluation results obtained from the work process evaluation and work outcome evaluation, the user's work evaluation result is determined.

2. The work supervision method according to claim 1, characterized in that, The process evaluation results include job performance evaluation results; these job performance evaluation results are determined based on the following steps: Based on the work process data, the user's total work time and the distraction time corresponding to distraction behaviors within the total work time are determined, and the answering focus is determined based on the proportion of the distraction time in the total work time; Based on the work process data, the user's total writing time and the pause time corresponding to the pause action within the total writing time are determined, and the answer continuity is determined based on the proportion of the pause time in the total writing time. The assessment result of the assignment performance is determined based on the aforementioned focus and / or consistency of answering questions.

3. The work supervision method according to claim 2, characterized in that, The process evaluation results also include capability development evaluation results; these capability development evaluation results are determined based on the following steps: Based on the homework results data, the completeness of the user's problem-solving steps is determined, and based on the completeness of the problem-solving steps, the completeness of the user's logical thinking is determined; Based on the work process data and the work correction results, determine the number of self-corrected errors and the total number of errors corresponding to the user's work results data, and determine the comprehensiveness of the user's self-correction based on the ratio of the number of self-corrected errors to the total number of errors. The assessment results of the ability development are determined based on the completeness of logical thinking and / or the comprehensiveness of autonomous error correction.

4. The work supervision method according to claim 3, characterized in that, The process evaluation results also include learning quality evaluation results; these learning quality evaluation results are determined based on the following steps: Based on the work process data, the user's dwell time on the preset challenge and the number of times the user checks within the total work time are determined, and the work persistence is determined based on the ratio of the dwell time to the standard dwell time. Based on the homework results data, the user's writing standardization is determined, and based on the number of checks and the writing standardization, the homework rigor is determined. Based on the homework grading results, determine the reasonableness score of non-standard answers in the user's homework data, and determine the homework innovation level based on the reasonableness score; The learning quality assessment result is determined based on at least one of the following: the persistence in completing the assignment, the rigor of completing the assignment, and the innovation of completing the assignment.

5. The work supervision method according to any one of claims 2 to 4, characterized in that, The determination of focus level based on the proportion of distraction time in the total task time also includes: Based on the work process data, the user's behavioral performance and emotional state are determined; If the behavior reflects that the user has plagiarized, the emotional state reflects that the user has excessive anxiety, or the focus of answering questions is consistently below the focus threshold, then warning data is generated based on the task process data, the process evaluation results, and the task evaluation results. The warning data will be sent to the bound guardian's terminal and / or teacher's terminal.

6. The work supervision method according to claim 5, characterized in that, The early warning data includes early warning data and early warning suggestion data; The warning data includes user images corresponding to the plagiarism behavior and / or user images when experiencing excessive anxiety. The early warning and suggestion data includes at least one of the following: operation process analysis report, operation process improvement suggestions, and supplementary practice exercises.

7. A work monitoring device, characterized in that, include: The acquisition unit is used to acquire the user's work process data and work result data, and determine the work grading result corresponding to the work result data; The evaluation unit is used to evaluate the work process and the work results based on the work process data, the work result data and their corresponding work correction results. The determining unit is used to determine the user's job evaluation result based on the process evaluation result and the result evaluation result obtained from the job process evaluation and job result evaluation.

8. A work monitoring device, characterized in that, Includes camera, touchscreen, and processor; The processor is used to acquire the work process data and work result data collected by the camera and the touch screen when the user works on the touch screen, and to determine the work correction result corresponding to the work result data; Based on the work process data, the work result data and their corresponding work correction results, conduct work process evaluation and work result evaluation; Based on the process evaluation results and outcome evaluation results obtained from the work process evaluation and work outcome evaluation, the user's work evaluation result is determined.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the job supervision method as described in any one of claims 1 to 6.

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

Images