Subjective question intelligent marking and learning condition analysis system based on writing characteristics

Abstract

The application provides a subjective question intelligent marking and learning situation analysis system based on writing characteristics, relates to the technical field of education, and the system architecture comprises a writing characteristic acquisition and analysis subsystem, an intelligent marking subsystem and a learning situation analysis and prediction subsystem. With the aid of quantum technology, all-round innovation and improvement are realized. The model constructed by quantum sensors and calculation makes micro dynamic data accurate and observable, time sequence characteristics deep and analyzable, and the quantum algorithm is remarkably effective in constructing the font aesthetics and multi-factor correlation model on a macro level. Quantum natural language processing and neural network cooperation, semantic style fusion innovation, knowledge graph reasoning with the aid of quantum graph algorithm, etc. can accurately judge and scientifically guide various types of questions. Quantum hashing and federal learning help to integrate multi-source data, quantum random walk and autoregressive moving average model realize psychological correlation prediction, accurately understand learning situation, effectively intervene to help students' physical and mental and academic progress, and greatly enhance the efficiency and value of the system in the field of education.

Description

Technical Field

[0001] This invention relates to the field of educational technology, specifically to an intelligent grading and learning analysis system for subjective questions based on handwriting characteristics. Background Technology

[0002] According to the Chinese patent application CN110390032B, a method and system for grading handwritten essays is disclosed. The method includes: acquiring an image of the handwritten essay to be graded; using deep learning-based handwritten character recognition technology to generate the essay text from the image; performing cluster analysis on each paragraph of the essay text to identify problematic paragraphs; selecting the first sentence of the problematic paragraph as the benchmark sentence; searching for similar sentences in a sample sentence database, including: calculating the second edit distance between sentences in the first, second, and third priority sample sentence databases and the benchmark sentence; selecting M sentences with a second edit distance less than a second threshold as similar sentences; and generating grading results, including recommending similar sentences to the students corresponding to the essay texts to be graded. This invention enables large-scale quantitative processing of handwritten essays.

[0003] According to Chinese patent application CN109064814A, a test grading method and system are disclosed. The grading method includes: receiving test question information uploaded by teachers; generating an identification code based on the test question information; sending the identification code to a designated terminal, enabling the designated terminal to retrieve and display the corresponding test question based on the identification code; retrieving the test question answers uploaded by the designated terminal; comparing the test question answers with pre-stored standard answers to generate a comparison result; generating grading statistics with preset statistical dimensions based on the comparison result; and sending the grading statistics to the teachers. This invention pushes test questions to students via identification codes, is simple to operate, quickly collects student answers, automatically grades them, and obtains the analysis results needed by teachers, greatly reducing the workload of teachers.

[0004] According to a test paper grading system disclosed in Chinese Publication No. CN106355176A, the system comprises a transmission device module for separating and transmitting test papers, a photographing module for taking pictures of the test papers, an information sending module for transmitting the pictures to a computer, and an information receiving module. The transmission device module separates the test papers one by one, the photographing module takes pictures, and the resulting images are transmitted to the computer via the information sending and receiving modules. The computer uses an OCR technology recognition module to recognize the images and compares them with test questions in a large database module to analyze the correctness of the questions and score them using a scoring module. This test paper grading system does not require specific answer sheets, has no requirements on the paper type of the test papers, has a wide range of applications, and eliminates expensive components such as cursor answer sheet scanners. It can grade multiple-choice questions and simple fill-in-the-blank questions in daily exams or assignments, significantly reducing the grading workload for teachers.

[0005] The aforementioned patent documents and prior art have the following technical problems when used:

[0006] Problem 1: Traditional subjective question grading and learning analysis systems struggle to collect and analyze writing characteristics comprehensively, accurately, and in depth. They fail to fully explore the multifaceted information about students behind micro and macro writing characteristics, such as writing habits, thinking patterns, psychological states, and their connection with cultural background and subject preferences. This results in an insufficient and comprehensive understanding of students, making it difficult to provide precise and personalized educational services.

[0007] Problem 2: Existing intelligent grading technology is insufficient in comprehensively considering the semantic correctness of answers, writing style characteristics, and logical fit of subject knowledge system. It is difficult to achieve accurate evaluation and scientific guidance, and it cannot efficiently and accurately handle grading tasks that integrate semantic understanding and writing style. Furthermore, when faced with questions that require rigorous reasoning and open-ended questions, it lacks an effective evaluation and feedback mechanism, which limits the accurate assessment and promotion of students' knowledge mastery and thinking ability.

[0008] Thirdly, conventional learning analysis systems are inefficient and have limited accuracy when integrating multi-source data. They struggle to deeply explore the intrinsic connections and patterns between data, cannot accurately understand students' learning status, and cannot achieve timely and effective early warnings and accurate predictions. Furthermore, they face technical bottlenecks in promoting educational resource sharing through multi-source data joint learning while protecting data privacy, and cannot adequately meet the needs of personalized teaching and the coordinated development of students' physical and mental health and academic performance. Summary of the Invention

[0009] Technical problems to be solved

[0010] To address the shortcomings of existing technologies, this invention provides an intelligent grading and learning analysis system for subjective questions based on handwriting features, solving the following problems:

[0011] 1. Addressing the problem that traditional teaching and grading systems struggle to deeply and comprehensively analyze handwriting characteristics, thus hindering the provision of precise and personalized educational services;

[0012] 2. Regarding the shortcomings of intelligent grading technology in accurately judging multiple factors and responding to different question types;

[0013] 3. Addressing the issues of incomplete data analysis and difficulty in quickly and accurately predicting and issuing early warnings for anomalies in conventional learning analysis systems.

[0014] Technical solution

[0015] To achieve the above objectives, the present invention provides the following technical solution: an intelligent marking and learning analysis system for subjective questions based on handwriting features, wherein the system architecture includes a handwriting feature acquisition and analysis subsystem, an intelligent marking subsystem, and a learning analysis and prediction subsystem, wherein:

[0016] The writing feature acquisition and analysis subsystem is used to accurately collect multi-dimensional feature information during the student's writing process and deeply analyze the information such as the student's writing habits, thinking patterns, and psychological state behind these features. The writing feature acquisition and analysis subsystem includes a micro-feature acquisition and analysis module and a macro-feature quantification and correlation module. The micro-feature acquisition and analysis module uses a high-sensitivity quantum sensor to collect micro-dynamic data such as pen tip pressure changes, tilt angles, and accelerations. It quickly constructs a hand muscle movement model through quantum computing to analyze the stability of the force used by the student, the continuity of strokes, and emotional state during writing. It accurately locates the psychological fluctuations or knowledge uncertainties reflected by abnormal changes in pressure and angle. The macro-feature quantification and correlation module introduces a multi-dimensional quantitative index system and uses quantum machine learning algorithms (such as quantum support vector machines and quantum clustering algorithms) to quantify the aesthetic features of fonts (symmetry, balance, etc.) and analyze their correlation with psychological traits. It uses quantum Bayesian network algorithms to construct a correlation model between writing style and cultural background and subject preferences to explore the influence and hidden correlation of different cultures and subjects on the shaping of writing style.

[0017] The intelligent grading subsystem is used to comprehensively, accurately, and intelligently grade students' subjective answers. It considers multiple factors, including the semantic correctness of the answer, writing style characteristics, and logical consistency with the subject knowledge system, providing objective, fair, and constructive grading results and feedback. The intelligent grading subsystem includes a semantic style fusion grading module and an intelligent graph reasoning grading module. The semantic style fusion grading module uses quantum natural language processing algorithms (such as quantum latent semantic analysis) to understand the semantics of the answer, combines this with quantum neural network analysis of writing style characteristics, and scores the answer based on the relationship between the two. The system assigns appropriate weights to the font emphasis and semantic accuracy when describing knowledge points, and adjusts the score based on the degree of mastery of the knowledge points when the font is hesitant. The knowledge graph reasoning and grading module uses quantum graph algorithms to construct a subject knowledge graph. Based on the student's answer knowledge points, it locates the related logic in the graph, checks the coverage and derivation process of knowledge nodes. For example, in physics questions, it sorts out multiple knowledge chains and annotates loopholes. It uses quantum logic gate circuits to simulate the reasoning process, breaks down the reasoning steps of mathematical proofs, and checks their standardization and logic. It uses quantum search algorithms to evaluate the answer ideas for open-ended questions and provides score ranges and evaluation suggestions.

[0018] The learning analysis and prediction subsystem, through deep integration of multi-source data, mines the intrinsic connections and patterns between handwriting characteristics and learning behaviors, psychological states, and academic performance. This enables comprehensive insight, timely warnings, and accurate predictions of students' learning progress, providing strong support for personalized teaching and learning. The subsystem includes a multi-source data fusion and insight module and a psychological characteristic correlation and prediction module. The multi-source data fusion and insight module integrates handwriting characteristics and learning behavior data through a data integration platform, utilizes quantum encryption to ensure data security, and explores correlation patterns between the two through quantum data mining and machine learning algorithms, such as analyzing classroom notes. By linking homework and exam performance, the system identifies learning difficulties or strengths corresponding to different handwriting and behavioral characteristics. Leveraging quantum computing to accelerate complex correlation analysis, the psychological characteristic correlation prediction module constructs a handwriting psychological analysis model. Using a quantum random walk algorithm, it analyzes the correlation between handwriting pressure, speed, and other characteristics and psychological factors in a high-dimensional feature space, inferring changes in psychological pressure and interest. Combining historical performance data, it utilizes quantum time series analysis and machine learning prediction algorithms to construct a performance prediction model. Based on handwriting psychological state and performance trends, it predicts future performance trends, provides early warnings for different situations, and offers teaching suggestions to promote the coordinated development of students' physical and mental health and academic performance.

[0019] Preferably, the system further includes a data visualization subsystem, which is used to present the data and results generated by the handwriting feature acquisition and analysis subsystem, the intelligent grading subsystem, and the learning analysis and prediction subsystem to teachers, students, and education administrators in an intuitive and visual manner. This subsystem adopts quantum rendering technology to improve the clarity and interactivity of the visualization effect, so that users can more easily understand and utilize the information output by the system. Through quantum three-dimensional visualization, it displays the distribution of students' handwriting features in different dimensions and their correlation with learning performance.

[0020] Preferably, the system has a self-optimization function. By periodically collecting user feedback data and new learning sample data, it uses quantum reinforcement learning algorithms to adjust parameters and optimize the structure of various models in the system, including the writing feature analysis model, the intelligent grading model, and the learning prediction model, in order to adapt to the ever-changing educational environment and student characteristics, and improve the performance and accuracy of the system. For example, based on students' questions about the grading results and the requirements of new teaching syllabi, it automatically optimizes the scoring rules and semantic understanding strategies in the intelligent grading model.

[0021] Preferably, when the macroscopic feature quantization and association module uses quantum support vector machine to quantify font aesthetic features, the kernel function adopts quantum Gaussian kernel function, and its parameters are adjusted by quantum optimization algorithm to improve the accuracy of font aesthetic feature classification. Furthermore, when using quantum Bayesian network algorithm to construct association model, the determination of prior probability distribution is based on quantum statistical analysis of large-scale sample data, enabling the model to more accurately reflect the relationship between writing style and multiple factors.

[0022] Preferably, the quantum neural network in the semantic style fusion grading module adopts a quantum convolutional neural network architecture. The kernel size and stride of its quantum convolutional layer are optimized according to the spatial distribution characteristics of writing style features. When combining quantum latent semantic analysis to understand the semantics of the answer, the dimension of the semantic vector is reduced by quantum principal component analysis algorithm to reduce computational complexity and retain key semantic information, thereby improving the efficiency and accuracy of comprehensive scoring.

[0023] Preferably, in the process of constructing the subject knowledge graph, the quantum graph algorithm uses a quantum random walk algorithm to traverse knowledge nodes and discover relationships, thereby enhancing the completeness and accuracy of the knowledge graph construction. Furthermore, when using quantum logic gate circuits to simulate the reasoning process, quantum error correction code technology is used to reduce the error rate in the quantum computing process, ensuring the reliability of the reasoning step checks.

[0024] Preferably, when fusing writing feature and learning behavior data, the multi-source data fusion insight module uses a quantum hash algorithm to preprocess the data to improve the speed and accuracy of data fusion. Furthermore, when using quantum data mining and machine learning algorithms to explore correlation patterns, a quantum federated learning algorithm is employed, enabling effective joint learning even when multi-source data is distributed across different storage locations. This protects data privacy while uncovering valuable correlation information. When constructing a writing psychology analysis model, the step size and iteration count of the quantum random walk algorithm in the psychological feature correlation prediction module are adaptively adjusted according to the scale and complexity of the writing feature data to ensure efficient and accurate analysis of the correlation between features and psychological factors in a high-dimensional feature space. Finally, when constructing a performance prediction model, the quantum time series analysis algorithm employs a quantum autoregressive moving average (QARIMA) model to better capture the time series characteristics and trend changes of performance data.

[0025] Beneficial effects

[0026] This invention provides an intelligent grading and learning analysis system for subjective questions based on handwriting features. It has the following beneficial effects:

[0027] 1. This invention employs a system that provides in-depth analysis and comprehensive quantification of writing characteristics. In terms of writing characteristics, quantum technology facilitates the ultra-precise capture and in-depth analysis of microscopic features. A high-sensitivity quantum sensor, combined with a hand muscle movement model constructed using quantum computing, can meticulously perceive dynamic data such as pen tip pressure, tilt angle, and acceleration. This allows for precise interpretation of the stability of force exerted by students during writing, the continuity of strokes, and subtle changes in emotions and knowledge acquisition. Simultaneously, it efficiently analyzes the temporal characteristics of strokes, uncovering the continuity of thought and the fluency of knowledge integration. On the macroscopic level, quantum machine learning algorithms, such as quantum support vector machines, quantum clustering algorithms, and quantum Bayesian network algorithms, respectively achieve precise quantification of font aesthetic features and in-depth construction of writing style and multi-factor correlation models. This comprehensively reveals the rich student information behind writing styles, laying a solid foundation for subsequent intelligent grading and learning analysis.

[0028] 2. This invention employs a system that provides precise evaluation and scientific guidance in intelligent grading. In intelligent grading, quantum technology facilitates an innovative grading model that integrates semantics and style. Quantum natural language processing algorithms and quantum neural networks work together to deeply analyze writing style characteristics while understanding the semantics of answers. Weights are intelligently allocated based on the font presentation and semantic accuracy of knowledge point descriptions, optimizing the comprehensive scoring mechanism and improving grading efficiency and accuracy. Knowledge graph reasoning grading utilizes quantum graph algorithms to construct a complete subject knowledge graph, accurately locating the logical connections between student answers within the knowledge system. Quantum logic gate circuits are used to simulate the reasoning process, combined with quantum error correction code technology to ensure high reliability of reasoning checks. For open-ended questions, quantum search algorithms are used to quickly explore multiple answer approaches, providing reasonable scoring ranges and constructive evaluation suggestions, effectively enhancing the scientific rigor and comprehensiveness of grading and promoting students' knowledge acquisition and thinking expansion.

[0029] 3. This invention employs a system capable of providing precise insights and effective interventions in learning analysis. In this area, quantum technology facilitates multi-source data fusion and psychological characteristic correlation prediction. The quantum hash algorithm preprocesses handwriting characteristics and learning behavior data, accelerating the fusion process and accurately uncovering correlation patterns between the two, quickly identifying students' learning difficulties and strengths. The quantum federated learning algorithm enables multi-source data joint learning while protecting data privacy, promoting educational resource sharing and collaborative optimization. The psychological characteristic correlation prediction module adaptively analyzes the correlation between handwriting characteristics and psychological factors using the quantum random walk algorithm, accurately inferring changes in students' psychological states. Combined with a performance prediction model constructed using the quantum autoregressive moving average model, it accurately predicts students' performance trends, thus providing a strong basis for personalized teaching and timely psychological intervention, promoting the synergistic progress of students' physical and mental health and academic advancement. Attached Figure Description

[0030] Figure 1 This is a system architecture diagram of the present invention;

[0031] Figure 2 This is a diagram illustrating the system operation steps of the present invention;

[0032] Figure 3 This is a line graph showing the changes in pressure, angle, and acceleration over time in the microscopic feature acquisition and analysis module of this invention;

[0033] Figure 4 This is a scatter plot of the font aesthetic feature classification results in the macro-feature quantification and correlation module of the present invention.

[0034] Figure 5 This invention provides a three-dimensional visualization of the macro-feature clustering results in the macro-feature quantification and correlation module.

[0035] Figure 6This is a distribution diagram of documents in the low-dimensional semantic space in the semantic style fusion review module of this invention;

[0036] Figure 7 This is a comparison image of the original handwriting style feature image and the convolved image in the semantic style fusion grading module of the present invention;

[0037] Figure 8 This is a probability amplitude distribution diagram of knowledge nodes after quantum random walk in the knowledge graph reasoning and review module of the present invention;

[0038] Figure 9 This is a graph showing the probability amplitude distribution of quantum states after NOT gate operations in the knowledge graph reasoning and review module of this invention.

[0039] Figure 10 This is a three-dimensional diagram showing the clustering results of writing features and learning behavior data in this invention. Detailed Implementation

[0040] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Specific Implementation Example 1:

[0042] like Figure 1-10 As shown, the intelligent marking and learning analysis system for subjective questions based on handwriting features has an architecture that includes a handwriting feature acquisition and analysis subsystem, an intelligent marking subsystem, and a learning analysis and prediction subsystem, wherein:

[0043] The writing feature acquisition and analysis subsystem is used to accurately collect multi-dimensional feature information in the student's writing process and deeply analyze the information such as the student's writing habits, thinking patterns and psychological state behind these features. The writing feature acquisition and analysis subsystem includes a micro feature acquisition and analysis module and a macro feature quantification and correlation module.

[0044] The micro-feature acquisition and analysis module utilizes a high-sensitivity quantum sensor to collect micro-dynamic data such as pen tip pressure changes, tilt angles, and acceleration. Through quantum computing, it rapidly constructs a hand muscle movement model to analyze students' writing stability, stroke continuity, and emotional state. It precisely locates psychological fluctuations or knowledge uncertainties reflected in abnormal pressure and angle changes. The hand muscle movement model constructed through the high-sensitivity quantum sensor and quantum computing can capture micro-dynamic data such as pen tip pressure, tilt angle, and acceleration with extremely high precision. This allows for an unprecedented level of detail in analyzing students' writing stability and stroke continuity; for example, it can perceive pressure changes down to the micro-Newton level, thereby accurately locating students' writing process. Subtle emotional fluctuations or moments of uncertainty about specific knowledge points—this kind of in-depth insight at the micro level helps teachers fully understand students' psychological state and instantaneous reactions to knowledge acquisition while writing, providing a highly valuable reference for personalized teaching. Micro-feature analysis based on quantum computing can process massive amounts of temporal data, quickly and accurately analyzing temporal features such as the connection order between strokes, the duration and frequency of pauses, and the curve of writing speed changes. Compared with traditional analysis methods, it not only significantly improves speed but also uncovers deeper information on the coherence of thought and the fluency of knowledge integration. For example, when analyzing a student's writing of a complex argumentative essay, it can clearly identify the points where the student's thinking stalls when switching between different paragraphs or knowledge points, providing a precise basis for targeted tutoring.

[0045] The macro-feature quantification and correlation module introduces a multi-dimensional quantitative indicator system and uses quantum machine learning algorithms (such as quantum support vector machines and quantum clustering algorithms) to quantify font aesthetic features (symmetry, balance, etc.) and analyze their correlation with psychological traits. It utilizes quantum Bayesian network algorithms to construct a correlation model between handwriting style and cultural background and subject preferences, uncovering the influence and hidden correlations of different cultures and disciplines on handwriting style. When using quantum support vector machines to quantify font aesthetic features, the module employs a quantum Gaussian kernel function, with its parameters adjusted through quantum optimization algorithms to improve the accuracy of font aesthetic feature classification. Furthermore, when constructing the correlation model using quantum Bayesian network algorithms, the prior probability distribution is determined based on quantum statistical analysis of large-scale sample data, enabling the model to more accurately reflect the relationship between handwriting style and multiple factors. The module also utilizes quantum machine learning algorithms (such as quantum support vector machines and quantum clustering algorithms) to... The quantification of font aesthetic features (symmetry, balance, etc.) and the use of quantum Gaussian kernel functions and their optimized parameters significantly improve the accuracy of font aesthetic feature classification. This helps to more accurately characterize students' handwriting style. For example, it can more accurately distinguish handwriting samples with similar traditional features but slightly different aesthetic styles, providing a more reliable data foundation for studying the relationship between handwriting style and students' personality traits. By using quantum Bayesian network algorithms to construct a correlation model between handwriting style and cultural background and subject preferences, and determining the prior probability distribution based on quantum statistical analysis of large-scale sample data, it is possible to deeply explore the influence and hidden correlation of different cultures and subjects on the shaping of handwriting style. For example, it can discover the potential connection between the regularity of strokes in students' handwriting in the context of Eastern culture and the inheritance of Chinese character culture, as well as the mechanism by which science learning promotes the simplicity and symbolic use of students' handwriting. This provides strong support for cross-cultural and interdisciplinary educational research and teaching strategy development.

[0046] The intelligent grading subsystem is used to comprehensively, accurately, and intelligently grade students' subjective answers. It takes into account various factors such as the semantic correctness of the answers, writing style characteristics, and logical fit with the subject knowledge system, and provides objective, fair, and constructive grading results and feedback. The intelligent grading subsystem includes a semantic style fusion grading module and an intelligent graph reasoning grading module.

[0047] The semantic style fusion grading module employs quantum natural language processing algorithms (such as quantum latent semantic analysis) to understand the semantics of answers, combined with quantum neural network analysis of handwriting style features. A comprehensive score is awarded based on the relationship between these two methods, assigning appropriate weights based on the emphasis of font and semantic accuracy in the expression of knowledge points. Hesitation in font indicates a lack of understanding of the knowledge points, and the score is adjusted accordingly. The quantum neural network in the semantic style fusion grading module uses a quantum convolutional neural network architecture. The kernel size and stride of its quantum convolutional layers are optimized based on the spatial distribution characteristics of handwriting style features. When combining quantum latent semantic analysis to understand the semantics of answers, the dimension of the semantic vector is reduced using quantum principal component analysis to reduce computational complexity while retaining key semantic information, thereby improving the efficiency and accuracy of the comprehensive scoring. The combination of quantum natural language processing algorithms (such as quantum latent semantic analysis) and quantum neural networks achieves an organic fusion of semantic understanding and handwriting style feature analysis. During the grading process, the module can determine the appropriate handwriting style based on the emphasis of font and semantic accuracy in the expression of knowledge points. Accuracy is given appropriate weight. For example, core knowledge points highlighted in large font and bold are given higher weight if the semantic expression is accurate and clear, and vice versa. This integrated grading method not only considers the correctness of the answer content, but also takes into account the student's emphasis on knowledge and confidence, making the grading results more objective, fair and insightful. When analyzing handwriting style features, the quantum convolutional neural network architecture can more accurately capture the spatial distribution characteristics of handwriting style features by optimizing the kernel size and stride of the quantum convolutional layer, such as the correlation between the regional distribution of font size changes and the importance of knowledge points. At the same time, the quantum principal component analysis algorithm performs dimensionality reduction on semantic vectors, which reduces computational complexity while retaining key semantic information, greatly improving the efficiency and accuracy of comprehensive scoring. For example, when grading a subjective question containing multiple paragraphs, it can quickly and accurately assess the quality of each paragraph and synthesize the overall score, while providing students with detailed feedback and suggestions on semantics and handwriting style.

[0048] The knowledge graph reasoning and grading module utilizes quantum graph algorithms to construct subject-specific knowledge graphs. Based on students' answers, it locates logical connections within the graph, checks the coverage and derivation process of knowledge nodes (e.g., in physics problems, it organizes multiple knowledge chains and annotates gaps), uses quantum logic gates to simulate the reasoning process, breaks down mathematical proofs, and checks their standardization and logic. It also uses quantum search algorithms to evaluate the answer approach to open-ended questions, providing score ranges and evaluation suggestions. In constructing the subject-specific knowledge graph, the quantum graph algorithm employs quantum random walks to traverse knowledge nodes and discover relationships, enhancing the completeness and accuracy of the knowledge graph construction. Furthermore, it utilizes quantum logic gates to simulate reasoning... During the process, quantum error correction code technology is used to reduce the error rate in quantum computing and ensure the reliability of the reasoning steps. The quantum graph algorithm uses quantum random walk algorithm to construct subject knowledge graphs, which can more comprehensively traverse and discover knowledge nodes and their relationships, significantly enhancing the integrity and accuracy of knowledge graph construction. During grading, the related logic can be more accurately located in the graph based on the knowledge points of the student's answer. For example, in the knowledge graph of physics, for complex questions involving multiple physical concepts and principles, the complete knowledge chain can be quickly sorted out, and the student's answer can be accurately checked to see if it covers all key nodes and their derivation process. The gaps in the student's knowledge system can be discovered in a timely manner and detailed annotations can be given, improving the depth and quality of grading.

[0049] By simulating the reasoning process using quantum logic gates and employing quantum error-correcting codes to reduce the error rate in quantum computing, the reliability of the reasoning steps is ensured. In grading mathematical proofs and other problems requiring rigorous logical reasoning, this method can accurately judge the standardization and logic of each step of the reasoning, effectively avoiding grading biases caused by calculation errors or logical misjudgments. At the same time, the quantum search algorithm evaluates the answer ideas for open-ended questions, quickly exploring multiple possible answer directions, providing students with reasonable score ranges and inspiring evaluation suggestions, encouraging students' innovative thinking and diverse expressions, and improving the flexibility and scientific nature of grading.

[0050] Learning Situation Analysis and Prediction Subsystem: By deeply integrating multi-source data, it explores the intrinsic connections and patterns between writing characteristics and learning behaviors, psychological states, and academic performance, achieving comprehensive insight, timely warnings, and accurate predictions of students' learning situation, providing a strong basis for personalized teaching and learning support. The Learning Situation Analysis and Prediction Subsystem includes a multi-source data fusion insight module and a psychological characteristic correlation prediction module.

[0051] The multi-source data fusion and insight module integrates handwriting characteristics and learning behavior data through a data integration platform. It utilizes quantum encryption to ensure data security and employs quantum data mining and machine learning algorithms to explore correlation patterns between the two, such as analyzing the relationship between classroom notes and homework / exam performance. This reveals learning difficulties or strengths corresponding to different handwriting and behavioral characteristics. Quantum computing accelerates complex correlation analysis. When integrating handwriting characteristics and learning behavior data, the module uses quantum hashing algorithms for preprocessing to improve the speed and accuracy of data fusion. Furthermore, when using quantum data mining and machine learning algorithms to explore correlation patterns, it employs quantum federated learning algorithms. This enables effective joint learning even when multi-source data is distributed across different storage locations, protecting data privacy while uncovering valuable correlation information. The quantum hashing algorithm's preprocessing of handwriting characteristics and learning behavior data significantly improves the speed and accuracy of data fusion, particularly in analyzing classroom notes and homework / exam performance. When correlated with homework and exam performance, quantum hashing algorithms can more quickly identify learning difficulties or strengths corresponding to different handwriting and behavioral characteristics. For example, by using quantum hashing algorithms to quickly filter out students who frequently make corrections and take long to complete assignments, further analysis of their learning behavior data can accurately pinpoint potential difficulties in knowledge comprehension or application, providing teachers with timely intervention guidelines and improving the relevance and timeliness of teaching. Quantum federated learning algorithms can effectively perform joint learning even when multi-source data is distributed across different storage locations, uncovering valuable correlations while protecting data privacy. This allows different schools and educational institutions to share learning analysis experiences and results without disclosing student privacy data, promoting the optimal allocation of educational resources and the realization of educational equity. For instance, schools in different regions can jointly participate in learning analysis projects based on quantum federated learning, sharing model parameter update information to jointly improve their understanding of student learning and the development of teaching strategies.

[0052] The psychological feature correlation prediction module constructs a writing psychology analysis model and uses the quantum random walk algorithm to analyze the correlation between writing pressure, speed, and other features and psychological factors in a high-dimensional feature space. It infers changes in psychological pressure, interest, and other states. Combining historical performance data, it uses quantum time series analysis and machine learning prediction algorithms to construct a performance prediction model. Based on the writing psychology state and performance trends, it predicts future performance trends, provides early warnings for different situations, and offers teaching suggestions to promote the coordinated development of students' physical and mental health and academic performance. When constructing the writing psychology analysis model, the step size and iteration number of the quantum random walk algorithm are adaptively adjusted according to the scale and complexity of the writing feature data to ensure efficient and accurate analysis of the correlation between features and psychological factors in a high-dimensional feature space. Furthermore, when constructing the performance prediction model, the quantum time series analysis algorithm adopts the quantum autoregressive moving average model (QARIMA) to better capture the time series characteristics and trend changes of performance data. In the psychological feature correlation prediction module, the step size and iteration number of the quantum random walk algorithm are adjusted according to the scale and complexity of the writing feature data. The complexity is adaptively adjusted to ensure efficient and accurate analysis of the correlation between writing pressure, speed, and psychological factors in a high-dimensional feature space. This allows for precise inference of students' psychological stress levels, fluctuations in learning interest, changes in self-efficacy, and emotional state. For example, when facing exam pressure, analyzing sudden changes in students' writing speed and fluctuations in stress data can accurately determine changes in students' psychological state, providing timely and accurate information support for schools to conduct psychological counseling and intervention. The performance prediction model uses the Quantum Autoregressive Moving Average (QARIMA) model, which can better capture the time series characteristics and trend changes of performance data. Combining writing psychological state with historical performance trends, it can more accurately predict students' future performance trends. For example, for a student whose writing psychological state is gradually stabilizing and whose historical performance is on the rise, it can accurately predict the extent of their performance improvement in future exams and provide teachers with personalized teaching resource recommendations and learning task arrangement suggestions, promoting the coordinated development of students' physical and mental health and academic performance, and achieving precise educational management and personalized learning guidance.

[0053] The system also includes a data visualization subsystem, which presents the data and results generated by the handwriting feature collection and analysis subsystem, the intelligent grading subsystem, and the learning analysis and prediction subsystem to teachers, students, and education administrators in an intuitive and visual way. This subsystem uses quantum rendering technology to improve the clarity and interactivity of the visualization effect, so that users can more easily understand and utilize the information output by the system. Through quantum 3D visualization, it shows the distribution of students' handwriting features in different dimensions and their correlation with academic performance.

[0054] The system has a self-optimization function. By regularly collecting user feedback data and new learning sample data, it uses quantum reinforcement learning algorithms to adjust parameters and optimize the structure of various models in the system, including the handwriting feature analysis model, intelligent grading model, and learning prediction model, in order to adapt to the ever-changing educational environment and student characteristics, and improve the system's performance and accuracy. For example, based on students' questions about the grading results and the requirements of new teaching syllabi, it automatically optimizes the scoring rules and semantic understanding strategies in the intelligent grading model.

[0055] This system architecture leverages quantum technology to achieve comprehensive innovation and improvement. At the level of handwriting characteristics, models constructed by quantum sensors and computing enable precise observation of microscopic dynamic data and in-depth analysis of temporal features. On a macroscopic level, quantum algorithms have achieved remarkable results in the construction of font aesthetics and multi-factor correlation models, laying the foundation for the overall system. In the intelligent grading process, quantum natural language processing and neural networks collaborate, and semantic style is innovatively integrated. Knowledge graph reasoning, aided by quantum graph algorithms, ensures rigorous reasoning and can accurately evaluate and scientifically guide various question types. In terms of learning analysis, quantum hashing and federated learning facilitate the fusion of multi-source data, while quantum random walk and autoregressive moving average models enable psychological correlation prediction, accurately insight into learning, and effective intervention to help students' physical and mental well-being and academic progress in synergy, greatly enhancing the system's effectiveness and value in the field of education. Specific Implementation Example 2:

[0057] like Figure 1-10 As shown, based on the content of the above specific embodiments, the following content is further disclosed:

[0058] The specific details of each algorithm in the above system architecture are as follows:

[0059] Data acquisition using high-sensitivity quantum sensors and the construction of hand muscle movement models are related.

[0060] Implementation steps:

[0061] First, a high-sensitivity quantum sensor is used to collect microscopic dynamic data such as pen tip pressure changes, tilt angle, and acceleration in real time. The acquisition of this data is the basis for subsequent analysis. For example, during the writing process, the sensor will record the current pen tip pressure value, tilt angle value, and acceleration value at certain time intervals (such as every millisecond).

[0062] Next, quantum computing methods are used to construct a hand muscle movement model based on the collected dynamic data. Quantum computing, with its unique qubit characteristics, can process more complex state information. Compared with traditional computing, it can perform operations on multiple states simultaneously, thereby accelerating the model construction speed. For example, qubits are used to characterize the data state combinations corresponding to different hand muscle movements, and then these state information are integrated to construct a model that reflects the hand muscle movement.

[0063] Operating Logic: The movement of hand muscles during writing directly affects the state of the pen tip. These microscopic dynamic data are the external manifestations of hand muscle movement. Quantum computing utilizes the superposition and entanglement properties of quantum states to comprehensively analyze a large amount of data collected at different times, uncovering the intrinsic connections within them. The model constructed based on this can accurately reflect how hand muscles move to complete the writing action. For example, which muscles exert force will cause changes in pen tip pressure, and what tilt angle changes correspond to different directions of hand movements. It can accurately locate the psychological fluctuations or knowledge uncertainties reflected by abnormal changes in pressure and angle. For example, if a student's pen tip pressure suddenly increases and the stroke continuity becomes abnormal when writing content related to a certain knowledge point, the model, combined with subsequent analysis, can infer that it may be due to a lack of solid understanding of this knowledge point or changes in psychological state such as nervousness. This helps teachers gain a deeper understanding of students' mastery of knowledge.

[0064] Quantum support vector machines are used for font aesthetic feature quantization in macroscopic feature quantization and association modules.

[0065] The basic mathematical formula for a standard support vector machine for binary classification is as follows:

[0066]

[0067] Where w is the weight vector (representing the direction of the normal vector of the hyperplane), b is the bias term, and x... i It is the input sample feature vector (in this case, a feature vector related to font aesthetics, such as a vector composed of feature values ​​related to font symmetry, balance, etc.), y i Here, represents the corresponding sample category label (e.g., a font with high aesthetic appeal can be set to +1, and one with low aesthetic appeal to -1, etc.), n is the number of samples, and represents the inner product operation of vectors, ||w||. 2 This represents the square of the L2 norm of the weight vector (that is, the sum of the squares of the elements of the vector, the purpose of which is to control the margin of the hyperplane to maximize the classification effect).

[0068] Implementation steps:

[0069] The first step is to collect a large amount of font sample data with annotations (font aesthetic category labels), and extract the font aesthetic feature vector and its category label y for each sample. i ;

[0070] The second step is to substitute these data into the objective function and constraints mentioned above, and solve for the values ​​of w and b by using optimization algorithms (such as commonly used sequence minimum optimization algorithms, which may be combined with quantum optimization algorithms to accelerate the process in quantum scenarios). The weight vector and bias terms are continuously adjusted so that the hyperplane can better separate fonts with different aesthetic levels.

[0071] The third step is to actually quantify and classify the aesthetic features of the new font, and substitute its corresponding feature vector into the classification function f(x) = w·x + b (the details of the sign function for category judgment are omitted here for simplification) determined by the trained w and b to determine which category of the font is aesthetically pleasing.

[0072] Improving the accuracy of font aesthetic feature classification can more accurately distinguish fonts with different levels of beauty and styles, thereby allowing for a deeper exploration of the relationship between font aesthetic features and students' psychological traits. For example, it can be found that students with beautiful handwriting may pay more attention to details during the learning process, providing a strong basis for subsequent learning analysis.

[0073] Quantum clustering algorithms are used for correlation analysis in macroscopic feature quantization and association modules. This is explained by combining general clustering algorithm ideas with quantum properties.

[0074] Implementation steps:

[0075] First, determine the objects to be clustered, that is, collect relevant handwriting feature data. For example, take the overall macro-feature data of students' handwriting (which may include multiple dimensions such as font size distribution, character spacing and other comprehensive feature vector representations) as the data set to be clustered.

[0076] Then, initialize the cluster centers. Some data points can be randomly selected as the initial cluster centers (in quantum clustering algorithms, the selection or representation of these initial points may be done in a more efficient way, such as using quantum states for storage).

[0077] Next, the distance from each data point to each cluster center is calculated (in the quantum scenario, the distance calculation may be completed more quickly by utilizing the properties of quantum state superposition and interference, such as reflecting the distance relationship through the evolution of quantum states), and the data points are assigned to the category of the nearest cluster center based on the distance.

[0078] Then, recalculate the cluster centers based on the newly divided data points (for example, by taking the mean of the feature vectors of data points of the same category to update the cluster centers; quantum computing can accelerate this mean calculation process). Repeat the above steps until the cluster centers no longer change significantly or the set number of iterations is reached, or other stopping conditions are met.

[0079] Operating Logic: Based on the similarity between data points (reflected by metrics such as distance), similar handwriting features are clustered into one class. By continuously adjusting the cluster centers and re-dividing the data points, the similarity of data points within the same class is maximized, while the similarity between data points in different classes is minimized. The quantum clustering algorithm utilizes the unique computational and storage characteristics of quantum states. For example, a qubit can represent multiple states simultaneously. Multiple cases can be processed in parallel during distance calculation and cluster center updates, improving clustering efficiency and effectiveness. It can discover clustering situations of different handwriting styles and features from a macroscopic perspective. For example, it can group students with similar macroscopic font features into one class, and then analyze the commonalities that these students may have in terms of psychological traits, learning habits, etc. It helps to explore the correlation between handwriting features and other factors, providing a basis for data division for subsequent construction of correlation models.

[0080] Quantum Bayesian network algorithms are used to construct writing style and multi-factor association models for macroscopic feature quantification and correlation modules.

[0081] In a Bayesian network, for a node X, which can represent variables such as writing style, given its parent node set Pa(X), its conditional probability distribution formula is as follows:

[0082]

[0083] in,

[0084] P(X|Pa(X)) represents the probability of node X occurring given the values ​​of the set of parent nodes Pa(X). It reflects the conditional dependency between nodes. For example, given a student's cultural background (parent node), the probability that their writing style (node ​​X) exhibits a certain characteristic is given.

[0085] P(X, Pa(X)) represents the joint probability, which is the probability that node X and its parent node set Pa(X) occur simultaneously. It covers the probability description of all possible combinations between the two.

[0086] P(Pa(X)) is the probability of the parent node set itself, which is the basis for calculating conditional probability and is determined through statistics or known prior information;

[0087] The joint probability distribution of the entire Bayesian network can be obtained by multiplying the conditional probability distributions of each node (assuming the network structure is a directed acyclic graph and the dependencies between nodes conform to the Bayesian network structure), that is:

[0088]

[0089] Here X1, X2, ..., X n It refers to all node variables in the network (such as writing style, cultural background, subject preference, etc., P(X1, X2, ..., X...). n The joint probability distribution of the entire Bayesian network represents the probability of all node variables taking various combinations of values ​​simultaneously. It is calculated by multiplying the conditional probability distributions of each node based on its parent node, reflecting the complex probabilistic relationships between variables in the entire network structure.

[0090] The network structure is built based on the logical relationships between various factors, and the dependencies between these factors are described in probabilistic form. By learning and adjusting the conditional probability distribution through prior probabilities and sample data, probabilistic reasoning can be performed with known partial information to derive the probabilities of other related factors. The quantum Bayesian network algorithm utilizes the special properties of quantum states to accelerate the processes of probability calculation, statistical analysis, and network training optimization, and more accurately reflects the complex relationship between handwriting style and multiple factors such as cultural background and subject preferences. It can construct a model of the relationship between handwriting style and cultural background and subject preferences, and explore the influence and hidden correlation of different cultures and subjects on the shaping of handwriting style. For example, it can discover that students from certain cultural backgrounds tend to have a certain font style in their handwriting, or that students with a certain subject learning preference have specific norms in their handwriting style. This helps to more comprehensively understand the influencing factors behind students' handwriting and provides a reference for personalized teaching.

[0091] Quantum latent semantic analysis is used in the semantic style fusion grading module to understand the semantics of answers.

[0092] Traditional latent semantic analysis mainly involves singular value decomposition. Suppose we have a document-term matrix A. m ×n, where m is the number of documents and n is the number of terms, its singular value decomposition formula is:

[0093] A=U∑V T

[0094] Among them, U m×n It is a left singular vector matrix, ∑ m×n It is a diagonal matrix, and the elements on the diagonal are singular values ​​σ. i (i=1,2,…,min(m,n)) and σ1≥σ2≥…≥σ min(m，n) ≥0, V n×nIt is a right singular vector matrix, where T denotes the matrix transpose operation, and V T (or ) is the transpose of the right singular vector matrix, used to construct a complete singular value decomposition (or a similar decomposition under the concept of quantization) equation relation to accurately extract semantic information from the matrix;

[0095] Quantum latent semantic analysis involves the representation of quantum states and related quantum computation operations. However, there is currently no completely unified and definitive quantization formula. Data processing and analysis can be performed within a quantum computing framework, analogous to the classical SVD principle. For example, qubits can be used to represent the characteristic states of documents or terms. Here, we define the quantized matrix representation as Q. A It has a similar decomposition form:

[0096]

[0097] Q U Q ∑ Q V These are the corresponding matrix representations after quantization, whose elements are quantum state-related representations. Their dimensions are similar to those in the classical case, but the properties of their internal elements are different, based on quantum information carriers such as qubits.

[0098] Operational Logic: By performing singular value decomposition (or quantization decomposition) on the document-term matrix, high-dimensional sparse text data is transformed into vector representations in a low-dimensional semantic space. This allows for the mining of the underlying semantic structure of the text. In this low-dimensional space, documents or terms with similar semantics are closer together in the vector space, enabling semantic analysis using vector operations and metrics. Quantum latent semantic analysis leverages the characteristics of quantum computing, such as the parallel processing capability of qubits, to simultaneously process the quantum state representations of multiple terms or documents, accelerating the decomposition process and the extraction and analysis of semantic information. Compared to classical methods, this approach offers significant advantages in large-scale text processing. This data processing and complex semantic mining approach has potential efficiency advantages. It can understand the semantics of answers and mine deeper semantic information, rather than being limited to surface-level word matching. For example, it can discover that students use different expressions but the same semantic knowledge points in their answers, avoiding misjudging the correctness of answers due to differences in expression. Through the efficiency of quantum latent semantic analysis (when processing large-scale data), it can quickly process the subjective answers of many students, improve the speed and accuracy of grading, and better adapt to the needs of semantic analysis of subjective answers of different subjects and types, providing a more reliable semantic basis for comprehensive scoring.

[0099] The quantum neural network in the semantic style fusion grading module is used to analyze handwriting style features.

[0100] The quantum convolutional layer operation in a quantum convolutional neural network can be compared to the convolutional layer formula of a classical convolutional neural network, but it involves the operation of quantum states. Assuming the input quantum state represents an image (analogous to a two-dimensional representation of handwriting style features, such as the spatial distribution features of strokes), it is |ψ in The quantum convolution kernel is |κ>, and the quantum convolution operation can be represented as:

[0101]

[0102] Where s represents the displacement (translation) parameter of the convolution kernel, C(s) is the coefficient related to the displacement, and |k s > represents the state of the quantum convolution kernel after displacement;

[0103] In quantum convolutional neural networks, there are also operations such as pooling layers and fully connected layers. The computation of a fully connected layer is similar to the matrix multiplication form of a classical neural network, but the elements are quantum states and related quantum computation operations. For example, for a quantum fully connected layer, suppose the input quantum state vector is... The weight matrix is ​​|W>, the bias quantum state is |b>, and the output quantum state is... It can be represented as:

[0104]

[0105] Here, addition and multiplication are operations based on quantum states, which are fundamentally different from classical numerical operations. For example, the superposition and entanglement of quantum states will participate in these operations.

[0106] Operating Logic: The quantum neural network extracts local features of handwriting style characteristics through quantum convolutional layers, scans the input data in quantum state space using quantum convolutional kernels, aggregates and reduces the dimensionality of features through pooling layers, and integrates and maps the extracted features through fully connected layers, ultimately outputting a quantum state representation of handwriting style characteristics. Throughout the process, based on the parallel processing capability of qubits and the superposition and entanglement characteristics of quantum states, the quantum neural network can simultaneously handle multiple feature dimensions and complex feature relationships. Compared with classical neural networks, it has stronger expressive power and analytical efficiency when dealing with complex and somewhat ambiguous features such as handwriting style. By continuously adjusting parameters such as the quantum weight matrix during the training process, the network can learn the relationship between different handwriting style features and the grading results, thereby accurately analyzing handwriting style features and providing valuable information for comprehensive scoring.

[0107] Beneficial effects: It can analyze writing style features more precisely. For example, it can accurately identify style characteristics such as the stroke coherence and force stability when students are writing. And through the powerful feature learning ability of the quantum neural network, it can discover some information related to learning attitudes, knowledge mastery levels, etc. hidden in the writing style. For example, a relatively chaotic writing style may imply that the student is in a hurry when answering questions or not confident enough in knowledge mastery. By combining quantum latent semantic analysis to understand the semantics of answers and being able to comprehensively score based on the connection between the two, it can improve the comprehensiveness and accuracy of grading, better evaluate the quality of students' subjective question answers, provide more targeted feedback for students, and provide more accurate teaching evaluation basis for teachers.

[0108] The quantum graph algorithm in the knowledge graph reasoning and grading module is used to construct a subject knowledge graph:

[0109] Mathematical formula (taking the quantum random walk algorithm as an example): The mathematical description of quantum random walk is relatively complex. Briefly speaking, on a graph G=(V, E) (where V is the vertex set and E is the edge set), the state of quantum random walk can be represented by a quantum state |ψ t > at time t:

[0110] |ψ t+1 > = U|ψ t >

[0111] where U is the unitary evolution operator of quantum random walk, which is related to the structure of the graph (the connection relationship between vertices) and the evolution rule of the quantum state. Its specific form can be expressed as (taking a simple discrete-time quantum random walk as an example):

[0112]

[0113] Here A uv is the element of the adjacency matrix of graph G, indicating whether there is an edge between vertex u and vertex v (if there is an edge, then A uv = 1; if there is no edge, then A uv = 0), |v><u| is the projection operator of the quantum state, and S is the local unitary transformation operator acting on the quantum state (such as a combination of basic quantum gate operations like Pauli matrices, used to achieve the transfer and state change of the quantum state between vertices).

[0114] Operational Logic: Based on the graph structure and the evolution rules of quantum states, quantum random walk starts from an initial state. At each step, it performs quantum state transitions and state changes between vertices of the knowledge graph according to the unitary evolution operator. Through multiple iterations, quantum states can simultaneously explore multiple paths and node relationships in a quantum parallel processing manner. By utilizing the superposition and entanglement properties of quantum states, it can mine deep structural information and hidden connections in the knowledge graph. Compared with traditional random walk algorithms, quantum random walk can traverse the knowledge graph more efficiently and discover more complex relationship patterns because it is not limited by the serial processing in classical computing and can consider multiple possibilities at the same time, thereby improving the completeness and accuracy of knowledge graph construction.

[0115] Beneficial effects: It enhances the completeness and accuracy of knowledge graph construction. In the construction of subject-specific knowledge graphs, it can more comprehensively discover various relationships between knowledge points. For example, in physics, it can more accurately sort out the multifaceted knowledge chains from basic concepts to complex physical phenomena, and discover some knowledge point connections that are easily overlooked under traditional methods. This helps to more accurately check the coverage and derivation process of knowledge nodes in students' answers when grading students' subjective questions, and to discover gaps in students' knowledge and logical errors. This lays the foundation for providing more accurate annotations and feedback, thereby improving the quality and effectiveness of teaching assessment and promoting students' in-depth understanding and mastery of the subject knowledge system.

[0116] Quantum logic gates (used to simulate the reasoning process) in the knowledge graph reasoning and grading module.

[0117] Mathematical formula (using simple quantum logic gate operations as an example): For example, the matrix representation of a single-qubit NOT gate operation is as follows:

[0118] If the input quantum state is |ψ>=α|0>+β|1> (where α and β are complex numbers, satisfying |α| 2 +|β| 2 =1), the output quantum state |ψ after NOT gate operation ′ >For:

[0119]

[0120] For multi-qubit logic gate operations, such as the CNOT gate (controlled NOT gate), if there are two qubits |ψ1|ψ2>, where |ψ1>=α1|0>+β1|1> and |ψ2>=α2|0>+β2|1>, the matrix representation of the CNOT gate is:

[0121]

[0122] When |ψ1> is the control bit and |ψ2> is the target bit, the output state after the CNOT gate operation is:

[0123]

[0124] Operating Logic: Quantum logic gates map the reasoning process to the combination operations of quantum logic gates. Utilizing the superposition and entanglement properties of quantum states and the operational rules of logic gates, they process the encoded student answer information. At the quantum bit level, multiple possible reasoning branches and states are processed simultaneously using quantum parallel computing. This simulates various logical judgments and state transitions during the reasoning process. Compared to traditional Boolean logic-based reasoning simulations, quantum logic gates can handle more complex quantum state information, more comprehensively consider the uncertainties and multiple possibilities in the reasoning process, and thus more accurately evaluate the reasoning process of the student's answer.

[0125] Beneficial effects: It can check the standardization and logic of students' reasoning steps in their answers. For example, in mathematical proof problems, it can accurately identify problems such as jumps in students' reasoning process and incorrect application of theorems. Through the efficient simulation of quantum logic gate circuits, it can quickly process the reasoning evaluation of a large number of students' answers, improve the efficiency of grading, and due to its good ability to handle complex reasoning situations, it can provide students with more detailed and accurate feedback on reasoning errors, help students better understand and improve their reasoning abilities, and promote the improvement of students' logical thinking ability and in-depth mastery of subject knowledge.

[0126] The quantum data mining and machine learning algorithms in the multi-source data fusion and insight module, taking quantum clustering as an example, are similar in concept to the quantum clustering algorithm in the intelligent grading subsystem mentioned earlier, but with different application scenarios:

[0127] Let the set of data points consisting of writing features and learning behavior data to be clustered be X = {x1, x2, ..., x...} n}, where x i This represents the i-th data point, which could be a vector composed of a student's writing pressure feature value, class note integrity feature value, etc.

[0128] Quantum clustering algorithms may involve operations such as representing data points in quantum states and calculating distances based on quantum states, for example, calculating the quantum distance between data points:

[0129]

[0130] Where |ψ(x) i )> and |ψ(x j The data points x are respectively. i and x jThe corresponding quantum state representation, f, is a quantum state-based function used to calculate quantum distance, which may involve operations such as the inner product and norm of quantum states, as well as comparison operations of qubits.

[0131] Operating Logic: Based on the quantum distance metric between data points, similar writing features and learning behavior data are clustered into one class. By continuously adjusting the cluster centers and re-dividing the data points, the similarity of data points within the same class in the quantum state space is maximized, while the similarity of data points between different classes is minimized. The quantum clustering algorithm utilizes the unique computational and storage characteristics of quantum states, such as the fact that qubits can represent multiple states simultaneously. Multiple cases can be processed in parallel during distance calculation and cluster center updates, improving clustering efficiency and effectiveness. In learning analysis, this clustering method can be used to mine different learning patterns of students from multi-source data, such as discovering characteristic combinations of different types of students, such as those with learning difficulties or high-achieving students, providing data support for personalized teaching.

[0132] Beneficial effects: It can analyze the correlation between classroom notes and homework / exam performance, and discover the learning difficulties or strengths corresponding to different writing and behavioral characteristics. For example, if it is found that a certain group of students writes slowly and has a high error rate in their homework, cluster analysis can group them into one category. Further analysis may reveal that this may be due to a lack of in-depth understanding of the knowledge, which leads to long thinking time when writing and easy errors in homework. Teachers can then provide specialized knowledge tutoring and learning method guidance to these students. The high efficiency of quantum clustering algorithms (utilizing characteristics such as quantum parallel computing) can quickly process large-scale multi-source student data, providing teachers with timely learning analysis results so that teachers can adjust their teaching strategies in a timely manner, improve teaching effectiveness, and promote students' learning progress.

[0133] The quantum random walk algorithm in the psychological feature association prediction module is used to analyze the association between features such as writing pressure and speed and psychological factors in a high-dimensional feature space.

[0134] Similar to the quantum random walk algorithm in the knowledge graph reasoning and grading module above, but with different application scenarios and analysis objects, in a high-dimensional feature space graph G = (V, E) (where V is the vertex set and E is the edge set) constructed from writing characteristics (such as pressure, speed, etc.) and psychological factors, the state of the quantum random walk can be represented by a quantum state over time, and its evolution equation can be roughly expressed as:

[0135] |ψ t+1 >=U|ψ t >

[0136] Among them, U is the unitary evolution operator of the quantum random walk, which is related to the structure of the high-dimensional feature space graph (the connection relationship between vertices) and the evolution rule of the quantum state. Its specific form can be expressed as (taking the simple discrete-time quantum random walk as an example):

[0137]

[0138] Here, A uv is the element of the adjacency matrix of the high-dimensional feature space graph, indicating whether there is an edge connecting vertex u and vertex v (if there is an edge, then A uv = 1; if there is no edge, then A uv = 0), |v><u| is the projection operator of the quantum state, and S is the local unitary transformation operator acting on the quantum state, such as the combination of basic quantum gate operations like Pauli matrices, etc., which is used to realize the transfer and state change of the quantum state between vertices. In this application scenario, vertices u and v can represent a certain combined state of writing eigenvalues or psychological factor eigenvalues. For example, vertex u may represent the state of medium writing pressure and low psychological pressure, and vertex v may represent the state of high writing pressure and high psychological pressure, etc. The edge represents a certain correlation or transfer possibility between these states.

[0139] Operation logic: Based on the structure of the high-dimensional feature space graph and the evolution rule of the quantum state, the quantum random walk starts from the initial state and, at each step, performs quantum state transfer and state change between the vertices of the combined state of writing features and psychological factors according to the unitary evolution operator. Through multiple iterations, the quantum state can simultaneously explore multiple paths and state relationships in a quantum parallel processing manner, and utilize the superposition and entanglement characteristics of the quantum state to挖掘出深层次的关联信息 in the high-dimensional feature space. Compared with traditional analysis methods, the quantum random walk can traverse the high-dimensional feature space more efficiently, discover more complex correlation patterns between writing features and psychological factors, because it is not restricted by the serial processing in classical computing and can consider multiple possibilities simultaneously, thus more accurately inferring the change of students' psychological states and providing a more accurate basis for psychological intervention in the teaching process.

[0140] Beneficial effects: It can infer the changes in states such as psychological pressure and interest. For example, if it is found that during the quantum random walk process, when the writing speed gradually slows down and the stroke coherence deteriorates, the quantum state tends to the vertex of high psychological pressure and low learning interest, then it can be inferred that the student may have relatively high psychological pressure and a decline in learning interest in this writing state. In this way, teachers can timely discover the psychological changes of students and adopt corresponding teaching strategy adjustments. Specific Embodiment Three:

[0142] As Figure 1-10 shown, based on the content in the above specific embodiment, the following content is further disclosed:

[0143] The operating steps of the above system architecture when in use are as follows;

[0144] Sp1: The writing feature acquisition and analysis subsystem operates first. The micro-feature acquisition and analysis module uses a high-sensitivity quantum sensor to accurately capture micro-dynamic data such as pen tip pressure, tilt angle, and acceleration. It constructs a hand muscle movement model through quantum computing to analyze the stability of writing force, stroke continuity, and emotional and psychological state, and locates anomalies. The macro-feature quantification and correlation module introduces a quantitative index system, uses quantum machine learning algorithms to quantify font aesthetic features and correlate them with psychological traits, and uses quantum Bayesian network algorithms to construct a correlation model between writing style and culture and subject preferences. At the same time, it optimizes algorithm parameters and prior probability determination, deeply explores hidden correlations, and lays the foundation for subsequent analysis.

[0145] SP2: The intelligent grading subsystem is then activated. The semantic style fusion grading module uses quantum natural language processing algorithms to understand the semantics of the answers, combines quantum neural networks to analyze writing style characteristics, and scores comprehensively based on the correlation between the two. It assigns weights based on the font presentation and semantic accuracy of knowledge points, judges the knowledge mastery behind the font hesitation, and adjusts the score accordingly. The knowledge graph reasoning grading module uses quantum graph algorithms to construct a subject knowledge graph, locates logical connections in the graph based on the student's answers, checks the coverage and derivation process of knowledge nodes, simulates reasoning steps, uses quantum search algorithms to evaluate open-ended questions, provides score ranges and evaluation suggestions, and optimizes the quantum neural network architecture and semantic vector processing to improve grading efficiency and accuracy.

[0146] SP3: The learning analysis and prediction subsystem begins operation. The multi-source data fusion and insight module builds a data integration platform, integrating handwriting characteristics and learning behavior data. Quantum encryption is used to ensure data security, quantum hashing algorithms are used to preprocess the data, and quantum federated learning algorithms are used to explore correlation patterns. The system analyzes the learning status corresponding to handwriting and behavioral characteristics, and quantum computing is used to accelerate the analysis. The psychological characteristic correlation prediction module constructs a handwriting psychology analysis model, uses quantum random walk algorithms to analyze the correlation between handwriting pressure, speed, and psychological factors, infers changes in psychological state, and combines historical performance data to construct a performance prediction model using a quantum autoregressive moving average model. Based on handwriting psychology and performance trends, the system predicts future performance, provides early warnings and teaching suggestions, and adaptively adjusts algorithm parameters.

[0147] Sp4: The data visualization subsystem plays a crucial role. It employs quantum rendering technology to integrate and process the data and results generated by the handwriting feature acquisition and analysis subsystem, the intelligent grading subsystem, and the learning analysis and prediction subsystem. Through quantum 3D visualization, it clearly displays the distribution of students' handwriting features in different dimensions and intuitively presents the correlation between them and academic performance. This greatly improves the clarity and interactivity of the visualization effect, enabling teachers, students, and education administrators to easily understand the complex information output by the system and thus better conduct teaching, learning, and management activities based on this information.

[0148] SP5: The system's self-optimization function is activated, regularly collecting user feedback data and new learning sample data. It utilizes quantum reinforcement learning algorithms to deeply optimize key components such as the handwriting feature analysis model, intelligent grading model, and learning prediction model. Based on student feedback on grading results and new teaching syllabus requirements, it automatically adjusts the scoring rules and semantic understanding strategies in the intelligent grading model. Simultaneously, it adaptively modifies the parameters and structure of other models, enabling the system to closely adapt to the ever-changing educational environment and the characteristics of the student population. This continuously improves the overall performance and accuracy of the system, ensuring it always operates efficiently and provides strong support for education and teaching. Specific Implementation Example 3:

[0150] like Figure 1-10 As shown, based on the content of the above specific embodiments, the following content is further disclosed:

[0151] The specific process for achieving comprehensive and accurate intelligent grading of answers using the various modules of the intelligent grading subsystem is shown below:

[0152] Semantic style fusion grading module

[0153] In terms of semantic understanding: The Quantum Latent Semantic Analysis (QLSA) algorithm is used to deeply mine the semantic information in the answer text. QLSA can process large-scale text data and, through the parallelism advantage of quantum computing, can quickly identify the core concepts, themes and semantic relationships between them in the answer. For example, in the answer to a history subjective question, it can accurately capture the historical events, figures, periods mentioned by the student and the causal relationships and influences between them, thereby judging whether the answer completely and accurately responds to the requirements of the question at the semantic level.

[0154] In terms of handwriting style analysis: Quantum neural networks are used to encode and learn students' handwriting style characteristics. The unique architecture of quantum neural networks and the powerful computing capabilities of quantum computing enable them to process complex handwriting style characteristic data, such as font size, shape, stroke thickness, slant, and writing speed variation curves. It can learn the potential patterns between different handwriting style characteristics and students' knowledge mastery and thinking state. For example, when students use larger fonts and thicker strokes when writing important knowledge points, quantum neural networks can combine semantic information to determine that this may be an emphasis on the knowledge point by the student, and if the semantics are correct, a certain positive evaluation weight will be given; conversely, if the font is small and the strokes are hesitant, with many corrections, it may indicate that the student lacks confidence in the knowledge or has a shallow understanding, and even if the semantics are roughly correct, the score will be adjusted accordingly.

[0155] In terms of comprehensive evaluation: By organically integrating semantic understanding results with handwriting style analysis results, a comprehensive review is achieved. Semantics and style are not viewed in isolation, but rather, based on pre-set integration rules and model weights, the impact of both on the quality of the answer is considered comprehensively. For example, for a literary appreciation question, in addition to judging whether the student's understanding of the work's theme, characters, and artistic techniques is accurate (semantic level), the review also analyzes whether their handwriting style reflects an emotional resonance with the work. For example, whether the fluency and rhythm of the font match the emotional expression of the work. If the semantics are accurate and the handwriting style effectively conveys emotion, a high score is given; if the semantics are flawed and the handwriting style also indicates a possible lack of understanding of the work (e.g., illegible handwriting, chaotic structure), a lower score is given. Detailed feedback suggestions are generated, pointing out semantic errors and problems reflected in the handwriting style, achieving precise and intelligent review.

[0156] Knowledge Graph Reasoning and Grading Module

[0157] In terms of knowledge graph construction and localization: Quantum graph algorithms are used to construct subject-specific knowledge graphs. This algorithm can efficiently handle the organization and association of large-scale knowledge data. During the construction process, various concepts, principles, and knowledge points in the subject are used as nodes, and the logical relationships between them (such as causal relationships, inclusion relationships, deductive relationships, etc.) are used as edges, forming a vast and orderly knowledge network. When grading student answers, the system can quickly locate the relevant concept networks and logical links in the knowledge graph based on the knowledge points mentioned in the answers. For example, when grading a physics circuit analysis question, the system can quickly find the nodes and their connections related to circuit components (resistors, capacitors, inductors, etc.) and circuit laws (Ohm's law, Kirchhoff's laws, etc.) in the knowledge graph, thereby gaining a comprehensive understanding of the knowledge scope and logical structure that the student's answer should cover.

[0158] In terms of reasoning process analysis: the system uses quantum logic gate circuits to simulate the reasoning process. This is based on the principle of quantum computing to accurately simulate and verify the logical reasoning steps. For mathematical proof problems and logical reasoning problems, the system transforms the reasoning steps in the student's answer into a sequence of operations of quantum logic gate circuits according to the reasoning rules and theorem system of the discipline. For example, in a geometry proof problem, each step of the derivation (such as drawing a new conclusion based on known conditions, or using theorems to prove it) corresponds to a specific combination of operations in quantum logic gate circuits. The system judges the correctness and rigor of the reasoning process by checking whether these operations conform to the rules of mathematical logic and whether there are problems such as logical jumps, circular arguments, or incorrect application of theorems. For scientific experiment analysis problems, the system also uses quantum logic gate circuits to check the rationality of the experimental design, the accuracy of data processing, and the scientific nature of the conclusion derivation in the student's answer, based on the experimental principles and the logical requirements of data analysis.

[0159] In handling open-ended questions: Quantum search algorithms are used to comprehensively explore and evaluate multiple possible solutions to open-ended questions. Open-ended questions often do not have a single, definitive answer. Quantum search algorithms utilize the superposition and parallelism characteristics of quantum computing to quickly traverse a large number of possible answer directions and approaches. For example, in a subjective question about analyzing social phenomena, quantum search algorithms can simultaneously consider multiple possibilities such as different analytical angles, theoretical frameworks, and examples. Based on pre-set evaluation criteria (such as the depth, breadth, innovativeness, and rationality of the analysis), each possible answer path is evaluated, and a reasonable score range is derived. Furthermore, it can provide students with insightful evaluation suggestions, such as pointing out some previously unconsidered analytical angles or directions for further exploration, thereby achieving intelligent grading of open-ended questions.

[0160] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising a reference structure" does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0161] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A subjective question intelligent grading and learning analysis system based on handwriting characteristics, characterized in that: The system includes a handwriting feature acquisition and analysis subsystem, an intelligent grading subsystem, and a learning analysis and prediction subsystem, wherein: The writing feature acquisition and analysis subsystem is used to accurately collect multi-dimensional feature information during the student's writing process and to deeply analyze the student's writing habits, thinking patterns and psychological state information behind these features. The writing feature acquisition and analysis subsystem includes a micro feature acquisition and analysis module and a macro feature quantification and correlation module. The intelligent grading subsystem is used to comprehensively, accurately, and intelligently grade students' subjective answers. It takes into account multiple factors such as the semantic correctness of the answers, writing style characteristics, and logical fit with the subject knowledge system, and provides objective, fair, and constructive grading results and feedback. The intelligent grading subsystem includes a semantic style fusion grading module and a knowledge graph reasoning grading module. The learning situation analysis and prediction subsystem: By deeply integrating multi-source data, it explores the intrinsic connections and patterns between writing characteristics and learning behavior, psychological state and academic performance, so as to achieve a comprehensive understanding of students' learning situation, timely warning and accurate prediction, and provide a strong basis for personalized teaching and learning support. The learning situation analysis and prediction subsystem includes a multi-source data fusion insight module and a psychological feature association prediction module. The micro-feature acquisition and analysis module uses a high-sensitivity quantum sensor to collect micro-dynamic data on pen tip pressure changes, tilt angles, and acceleration. Through quantum computing, it quickly constructs a hand muscle movement model to analyze the stability of force exertion, stroke continuity, and emotional state during student writing. It accurately locates psychological fluctuations or knowledge uncertainties reflected by abnormal changes in pressure and angle. The macro-feature quantification and correlation module introduces a multi-dimensional quantitative indicator system and uses quantum machine learning algorithms to quantify font aesthetic features and analyze their correlation with psychological traits. It uses quantum Bayesian network algorithms to construct a correlation model between writing style and cultural background and subject preferences, and explores the influence and hidden correlations of different cultures and subjects on the shaping of writing style. The semantic style fusion grading module uses quantum natural language processing algorithms to understand the semantics of answers, combines quantum neural networks to analyze writing style characteristics, and scores comprehensively based on the relationship between the two. It assigns reasonable weights based on the font emphasis and semantic accuracy when expressing knowledge points, and judges the degree of knowledge point mastery when the font is hesitant and adjusts the score accordingly. The knowledge graph reasoning grading module uses quantum graph algorithms to construct a subject knowledge graph, locates the related logic of knowledge points in the graph based on the student's answer, checks the coverage and derivation process of knowledge nodes, uses quantum logic gate circuits to simulate the reasoning process, breaks down the reasoning steps of mathematical proof problems, checks their standardization and logic, uses quantum search algorithms to evaluate the answer ideas of open-ended questions, and provides score ranges and evaluation suggestions. The multi-source data fusion and insight module integrates handwriting characteristics and learning behavior data through a data integration platform. It utilizes quantum encryption to ensure data security and employs quantum data mining and machine learning algorithms to explore the correlation patterns between the two, identifying learning difficulties or strengths corresponding to different handwriting and behavioral characteristics. Quantum computing accelerates complex correlation analysis. The psychological characteristic correlation prediction module constructs a handwriting psychological analysis model and uses a quantum random walk algorithm to analyze the correlation between handwriting pressure, speed, and psychological factors in a high-dimensional feature space. It infers changes in psychological pressure and interest, and combines historical performance data with quantum time series analysis and machine learning prediction algorithms to construct a performance prediction model. Based on handwriting psychological state and performance trends, it predicts future performance trends, provides early warnings for different situations, and offers teaching suggestions to promote the coordinated development of students' physical and mental health and academic performance.

2. The intelligent marking and learning analysis system for subjective questions based on handwriting features according to claim 1, characterized in that: The system also includes a data visualization subsystem, which presents the data and results generated by the writing feature acquisition and analysis subsystem, the intelligent grading subsystem, and the learning analysis and prediction subsystem to teachers, students, and education administrators in an intuitive and visual way. This subsystem uses quantum rendering technology to improve the clarity and interactivity of the visualization effect, so that users can more easily understand and utilize the information output by the system.

3. The intelligent marking and learning analysis system for subjective questions based on handwriting features according to claim 1, characterized in that: The system has a self-optimization function. By regularly collecting user feedback data and new learning sample data, it uses quantum reinforcement learning algorithms to adjust the parameters and optimize the structure of various models in the system, including the handwriting feature analysis model, the intelligent grading model, and the learning prediction model, in order to adapt to the ever-changing educational environment and the characteristics of the student group, and improve the performance and accuracy of the system.

4. The intelligent marking and learning analysis system for subjective questions based on handwriting features according to claim 1, characterized in that: When using quantum support vector machines to quantify font aesthetic features, the macroscopic feature quantization and association module employs a quantum Gaussian kernel function, the parameters of which are adjusted through a quantum optimization algorithm to improve the accuracy of font aesthetic feature classification. Furthermore, when constructing an association model using a quantum Bayesian network algorithm, the prior probability distribution is determined based on quantum statistical analysis of large-scale sample data, enabling the model to more accurately reflect the relationship between writing style and multiple factors.

5. The intelligent marking and learning analysis system for subjective questions based on handwriting features according to claim 1, characterized in that: The quantum neural network in the semantic style fusion grading module adopts a quantum convolutional neural network architecture. The kernel size and stride of its quantum convolutional layer are optimized according to the spatial distribution characteristics of writing style features. When combining quantum latent semantic analysis to understand the semantics of the answer, the dimension of the semantic vector is reduced by quantum principal component analysis algorithm to reduce computational complexity and retain key semantic information, thereby improving the efficiency and accuracy of comprehensive scoring.

6. The intelligent marking and learning analysis system for subjective questions based on handwriting features according to claim 1, characterized in that: In the process of constructing a subject knowledge graph, the knowledge graph reasoning and review module uses a quantum random walk algorithm to traverse knowledge nodes and discover relationships, thereby enhancing the completeness and accuracy of the knowledge graph construction. Furthermore, when using quantum logic gate circuits to simulate the reasoning process, quantum error correction code technology is used to reduce the error rate in the quantum computing process, ensuring the reliability of the reasoning step checks.

7. The intelligent marking and learning analysis system for subjective questions based on handwriting features according to claim 1, characterized in that: The multi-source data fusion insight module employs a quantum hash algorithm for data preprocessing when fusing writing features and learning behavior data to improve the speed and accuracy of data fusion. Furthermore, when exploring correlation patterns using quantum data mining and machine learning algorithms, it utilizes a quantum federated learning algorithm, enabling effective joint learning even when multi-source data is distributed across different storage locations. This protects data privacy while uncovering valuable correlation information. In the psychological feature correlation prediction module, when constructing a writing psychology analysis model, the step size and iteration count of the quantum random walk algorithm are adaptively adjusted based on the scale and complexity of the writing feature data to ensure efficient and accurate analysis of the correlation between features and psychological factors in a high-dimensional feature space. Finally, when constructing a performance prediction model, the quantum time series analysis algorithm employs a quantum autoregressive moving average (QARIMA) model to better capture the time series characteristics and trend changes of performance data.

Images