Cooperative governance-based school vertical large model system construction method

Abstract

The invention relates to the technical field of school model system construction, in particular to a school vertical large model system construction method based on collaborative governance, and the method comprises the following steps: 1, obtaining a demand purpose which comprises demand analysis and target setting; 2, system architecture design, including module division, data stream design and technology type selection; and 3, data acquisition and integration, including data source determination, data cleaning and data storage. Multi-dimensional data sources (such as a school management system, a teaching platform and student archives) are accurately determined, and the cleaned data are stored in a standardized manner in combination with a distributed database technology, so that the fragmentation problem of large-scale heterogeneous data in the education field is solved, the data collection integrity and the cross-system query efficiency are remarkably improved, and the data collection efficiency is improved. And a high-consistency data basis is provided for subsequent analysis.

Description

Technical Field

[0001] The present invention relates to the technical field of school model system construction, and in particular to a method for constructing a school vertical large model system based on collaborative governance. Background Art

[0002] The school vertical large model system has a wide range of applications and plays an important role in the field of education. It provides personalized learning suggestions and tutoring plans based on students' learning situation and knowledge level. Through natural language processing technology, it automatically answers questions raised by students to reduce the workload of teachers. According to students' learning interests and needs, it recommends relevant teaching resources such as textbooks, videos, exercises, etc., tracks students' learning progress in real time, generates learning reports, and helps teachers and parents understand students' learning situation.

[0003] After searching, based on the authorization announcement number CN118917583A, it discloses a school management system based on a management mathematical model, which relates to the field of school management technology. The school management system based on the management mathematical model includes a central processing unit, which is provided with a demand acquisition unit for acquiring the school resource demand and an allocation unit for allocating school resources. The demand acquisition unit includes: a student total module, a course total module, a classroom total module, and a teacher total module; the signal output end of the student total module is connected to the signal receiving end of the course total module, and the signal output end of the course total module is connected to the signal receiving end of the classroom total module. Data-driven decision-making helps to improve management efficiency and the scientific nature of decision-making, reduce decision-making risks, and through analysis and optimization, a more scientific and practical planning scheme can be formulated, which helps the school better achieve management goals and improve the overall management level.

[0004] At present, existing technologies usually use a single database for centralized storage, which is difficult to be compatible with different structured data sources such as school management systems and teaching platforms, resulting in insufficient integrity of data collection, too long response time for cross-system queries, and inability to meet real-time analysis needs. Conventional filtering methods mostly rely on single keyword matching, and lack effective recognition mechanisms for repeated texts, extra-long or extra-short noise texts generated by crawlers, resulting in downstream analysis data sets containing a large amount of invalid data.

[0005] Specific issues include the following:

[0006] 1. Traditional education data is scattered across different systems (e.g., academic affairs, student registration), and contains a lot of noise (e.g., duplicate records, confusing formats).

[0007] 2. Traditional management relies on manual statistics and hierarchical reporting, which is inefficient. Summary of the Invention

[0008] In view of the defects in the prior art, the present invention provides a method for constructing a school vertical large model system based on collaborative governance. The method for constructing a school vertical large model system based on collaborative governance is as follows:

[0009] Step 1: Demand purpose, including demand analysis and goal setting;

[0010] Demand analysis: Conduct in-depth interviews with school administrators, teachers, students, and parents at all levels to understand their pain points and needs in daily management and teaching.

[0011] Goal setting: clarify the main functions and expected effects of the system, such as improving management efficiency, optimizing resource allocation, and improving teaching quality;

[0012] Step 2: System architecture design, including module division, data flow design and technology selection;

[0013] Module division: Divide the system into multiple functional modules, such as administrative management, teaching management, student management, and home-school interaction.

[0014] Data flow design: Determine the data flow and interaction mode between modules to ensure real-time sharing and collaborative processing of information.

[0015] Technology selection: Choose the appropriate technology stack, such as cloud computing platform, big data processing framework, and artificial intelligence algorithm;

[0016] Step 3: Data collection and integration, including data source determination, data cleaning, and data storage;

[0017] Data source determination: Determine the data sources required by the system, such as school management system, teaching platform, and student files.

[0018] Data cleaning and integration: Clean, deduplicate and standardize the collected data to ensure data accuracy and consistency.

[0019] Data storage: Use distributed database or data warehouse technology to achieve efficient storage and fast query of large-scale data;

[0020] Step 4: Model training and optimization, including feature engineering, model selection, model training, and model optimization;

[0021] Feature Engineering: Extract and construct effective features based on business needs, such as students' learning behavior and teachers' teaching quality.

[0022] Model selection: Choose an appropriate machine learning or deep learning model, such as decision tree, random forest, or neural network.

[0023] Model training: Use historical data to train the model and evaluate the model performance through methods such as cross-validation.

[0024] Model optimization: Based on the evaluation results, adjust the model parameters and structure to improve the model's prediction accuracy and generalization ability.

[0025] Step 5: System integration and testing, including module integration, system testing, and user training;

[0026] Module integration: Integrate various functional modules into a unified platform to achieve seamless connection and collaborative work between modules.

[0027] System testing: Conduct comprehensive functional testing, performance testing, and security testing to ensure system stability and reliability.

[0028] User training: Provide training to system users to help them master the system usage and operation procedures;

[0029] Step 6: Operation and maintenance and continuous improvement, including system monitoring, user feedback, and continuous improvement;

[0030] System monitoring: Establish a real-time monitoring mechanism to promptly detect and resolve problems in system operation.

[0031] User feedback: Collect user feedback to understand the system usage and shortcomings.

[0032] Continuous improvement: Based on user feedback and business needs, we continuously optimize system functions and performance to enhance user experience.

[0033] Step 7: Collaborative governance mechanism, including multi-level collaboration, role authority management, and decision support;

[0034] Multi-level collaboration: Establish collaborative governance mechanisms at the university, department, and class levels to ensure smooth information flow and resource sharing among all levels;

[0035] Role and authority management: Set corresponding permissions based on the roles and responsibilities of different users to ensure the security and controllability of the system;

[0036] Decision support: Utilize big data and artificial intelligence technologies to provide scientific decision support for school administrators and improve the scientific nature and effectiveness of management decisions.

[0037] Furthermore, the data collection and integration in the third step includes data cleaning to remove noise data, such as advertisements and irrelevant content, to ensure the accuracy and relevance of the data, and to annotate the data, such as knowledge point classification and difficulty grading, so that the model can better understand and generate content.

[0038] Furthermore, the steps of data cleaning are as follows:

[0039] A1. Data collection and preliminary inspection, including data collection and preliminary inspection;

[0040] Collect data: Collect raw data from various sources;

[0041] Preliminary inspection: Check the basic structure and content of the data to understand the size, type and distribution of the data;

[0042] A2. Handle missing values, identify missing values, and handle missing values;

[0043] Identify missing values: Use statistical methods or visualization tools to identify missing values;

[0044] Handling missing values: Choose appropriate methods to handle missing values, such as deletion, imputation, or predictive imputation;

[0045] A3. Processing outliers, including detecting and processing outliers;

[0046] Detect outliers: Use statistical methods (such as box plots, Z-score) or visualization tools (such as scatter plots) to detect outliers;

[0047] Handling outliers: Choose appropriate methods to handle outliers, such as deletion, replacement, or retention;

[0048] A4. Processing duplicate values, including detecting and processing duplicate values;

[0049] Detect duplicate values: Use hash tables or sorting methods to detect duplicate values;

[0050] Handling duplicate values: Choose an appropriate method to handle duplicate values, such as deleting or merging;

[0051] A5. Data formatting, including standardizing data formats and converting data types;

[0052] Unified data format: Ensure that all data uses a unified format, such as date format and numerical format;

[0053] Convert data types: convert data to the appropriate type, such as converting a string to a numeric value;

[0054] A6. Data consistency check, including checking logical consistency and cross-field consistency;

[0055] Check logical consistency: ensure that the data is logically reasonable, such as age should not be negative;

[0056] Check cross-field consistency: Ensure that data between different fields is consistent, such as gender and title should match;

[0057] A7. Data standardization, including standardized numerical values and standardized text;

[0058] Standardized numerical values: Standardize numerical data to have the same scale, such as converting numerical values to standard scores (Z - scores);

[0059] Standardized text: Standardize text data, such as converting all text to lowercase or removing special characters;

[0060] A8. Data verification, including verifying data sources and verifying data integrity;

[0061] Verifying data sources: Ensure that data sources are reliable and the data collection process complies with specifications;

[0062] Verifying data integrity: Ensure that the data is complete and no important information is missing;

[0063] A9. Data storage and backup, including storing cleaned data and backing up data;

[0064] A10. Documentation, including recording the cleaning process and generating reports;

[0065] Recording the cleaning process: Record every step of the data cleaning process in detail for future reference and replication;

[0066] Generating reports: Generate data cleaning reports to summarize the cleaning process and results.

[0067] Furthermore, the methods for removing noise data in cleaning data include the following:

[0068] B1. Text cleaning, including removing special characters, removing HTML tags, and removing stop words;

[0069] Removing special characters: Delete unnecessary special characters, such as punctuation marks and emojis;

[0070] Removing HTML tags: If the data comes from a web page, HTML tags need to be removed;

[0071] Removing stop words: Remove common meaningless words, such as "de", "le", "zai", etc.;

[0072] B2. Content filtering, including keyword filtering, length filtering, and duplicate filtering;

[0073] Keyword filtering: Filter out content irrelevant to the theme based on keywords; <www. <www.

[0074] Length filtering: Remove text that is too short or too long, which may be noise data;

[0075] Duplicate filtering: removes duplicate text, which may be the result of crawler errors or human copying;

[0076] B3, grammatical and semantic analysis, including grammatical checking and semantic analysis;

[0077] Grammar check: Use grammar checker to filter out texts with serious grammatical errors;

[0078] Semantic analysis: Use natural language processing techniques, such as word embedding and sentence embedding, to filter out text with unclear semantics or irrelevant to the topic.

[0079] B4. Statistical methods, including frequency analysis and anomaly detection;

[0080] Frequency analysis: Count the frequency of words or phrases and remove texts with abnormally high or low frequencies;

[0081] Anomaly detection: Use statistical methods such as Z-score and IQR to detect and remove abnormal data.

[0082] Furthermore, the multi-level collaboration in the seventh step is implemented through cross-level collaboration, specifically:

[0083] Information sharing mechanism: Establish an information sharing mechanism to ensure smooth communication between all levels. For example, information can be quickly transmitted through the campus network, WeChat groups, etc.

[0084] Resource sharing platform: Establish a resource sharing platform to integrate various campus resources, such as teaching materials, scientific research results, etc., to achieve efficient resource sharing;

[0085] Joint training and exchange: Joint training and exchange activities are organized regularly, inviting managers and teachers at all levels to participate, to improve the overall management level and teaching quality.

[0086] Furthermore, the multi-level collaboration in the seventh step is also supported by technical operations, specifically:

[0087] Information management system: Introduce advanced information management systems, such as the academic affairs management system and the student management system, to improve management efficiency and information transparency;

[0088] Data analysis and decision support: Utilize big data analysis technology to analyze and mine data at all levels to provide support for decision making;

[0089] Online collaboration tools: Promote the use of online collaboration tools, such as Google Docs and Microsoft Teams, to facilitate collaborative work among all levels.

[0090] The beneficial effects of the present invention are as follows: 1. By accurately determining multi-dimensional data sources (such as school management systems, teaching platforms, and student archives) and combining distributed database technology to standardize the storage of cleaned data, the fragmentation problem of large-scale heterogeneous data in the education field is solved, the integrity of data collection and the efficiency of cross-system queries are significantly improved, and a highly consistent data foundation is provided for subsequent analysis;

[0091] 2. As described in 1, a three-level collaborative mechanism of keyword filtering, text length filtering, and duplicate content filtering is used to effectively eliminate irrelevant information, noise data, and redundant text. Compared with traditional single-dimensional filtering methods, the recognition rate of invalid data is improved, the purity and usability of the data set are significantly improved, and the consumption of downstream computing resources is reduced;

[0092] 3. As described in 2, through the dual statistical verification of frequency analysis and anomaly detection, abnormal high-frequency or low-frequency words and statistical outliers are accurately identified and eliminated, so that the data distribution meets the preset quality threshold, solving the efficiency bottleneck problem existing in traditional manual quality inspection, achieving quantitative guarantee of data reliability, and providing statistical confidence support for model training. BRIEF DESCRIPTION OF THE DRAWINGS

[0093] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly describes the drawings required for the specific embodiments or the description of the prior art. Similar elements or parts are generally identified by similar reference numerals throughout the drawings. Elements or parts in the drawings are not necessarily drawn to scale.

[0094] Figure 1 Schematic diagram of the construction method of the present invention;

[0095] Figure 2 Schematic diagram of data analysis of the present invention;

[0096] Figure 3 This is a schematic diagram of the principle of removing noise data according to the present invention;

[0097] Figure 4 Schematic diagram of the three-stage filtration system of the present invention. DETAILED DESCRIPTION

[0098] The following embodiments of the technical solution of the present invention will be described in detail with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention and are therefore only examples and are not intended to limit the scope of protection of the present invention.

[0099] It should be noted that, unless otherwise specified, the technical or scientific terms used in this application should have the common meanings understood by those skilled in the art to which the present invention belongs.

[0100] like Figure 1-Figure 4 As shown, a method for constructing a school vertical large model system based on collaborative governance is as follows:

[0101] Step 1: Demand purpose, including demand analysis and goal setting;

[0102] Demand analysis: Conduct in-depth interviews with school administrators, teachers, students, and parents at all levels to understand their pain points and needs in daily management and teaching.

[0103] Goal setting: clarify the main functions and expected effects of the system, such as improving management efficiency, optimizing resource allocation, and improving teaching quality;

[0104] Step 2: System architecture design, including module division, data flow design and technology selection;

[0105] Module division: Divide the system into multiple functional modules, such as administrative management, teaching management, student management, and home-school interaction.

[0106] Data flow design: Determine the data flow and interaction mode between modules to ensure real-time sharing and collaborative processing of information.

[0107] Technology selection: Choose the appropriate technology stack, such as cloud computing platform, big data processing framework, and artificial intelligence algorithm;

[0108] Step 3: Data collection and integration, including data source determination, data cleaning, and data storage;

[0109] Data source determination: Determine the data sources required by the system, such as school management system, teaching platform, and student files;

[0110] Data cleaning and integration: Clean, remove duplicates and standardize the collected data to ensure data accuracy and consistency;

[0111] Data storage: Use distributed database or data warehouse technology to achieve efficient storage and fast query of large-scale data;

[0112] Step 4: Model training and optimization, including feature engineering, model selection, model training, and model optimization;

[0113] Feature Engineering: Extract and construct effective features based on business needs, such as students' learning behavior and teachers' teaching quality;

[0114] Model selection: Choose an appropriate machine learning or deep learning model, such as decision tree, random forest, or neural network;

[0115] Model training: Use historical data to train the model and evaluate the model's performance through methods such as cross-validation;

[0116] Model optimization: Based on the evaluation results, adjust the model parameters and structure to improve the model's prediction accuracy and generalization ability.

[0117] Step 5: System integration and testing, including module integration, system testing, and user training;

[0118] Module integration: Integrate various functional modules into a unified platform to achieve seamless connection and collaborative work between modules;

[0119] System testing: Conduct comprehensive functional testing, performance testing, and security testing to ensure system stability and reliability;

[0120] User training: Provide training to system users to help them master the system usage and operation procedures;

[0121] Step 6: Operation and maintenance and continuous improvement, including system monitoring, user feedback, and continuous improvement;

[0122] System monitoring: Establish a real-time monitoring mechanism to promptly detect and resolve problems in system operation.

[0123] User feedback: Collect user feedback to understand the system usage and shortcomings.

[0124] Continuous improvement: Based on user feedback and business needs, we continuously optimize system functions and performance to enhance user experience.

[0125] Step 7: Collaborative governance mechanism, including multi-level collaboration, role authority management, and decision support;

[0126] Multi-level collaboration: Establish collaborative governance mechanisms at the university, department, and class levels to ensure smooth information flow and resource sharing among all levels;

[0127] Role and authority management: Set corresponding permissions based on the roles and responsibilities of different users to ensure the security and controllability of the system;

[0128] Decision support: Utilize big data and artificial intelligence technologies to provide scientific decision support for school administrators and improve the scientific nature and effectiveness of management decisions.

[0129] The third step involves data collection and integration, in which data cleaning is performed to remove noise data, such as advertisements and irrelevant content, to ensure the accuracy and relevance of the data, and data annotation is performed to label the data, such as knowledge point classification and difficulty grading, so that the model can better understand and generate content.

[0130] The steps for data cleaning are as follows:

[0131] A1. Data collection and preliminary inspection, including data collection and preliminary inspection;

[0132] Collect data: Collect raw data from various sources;

[0133] Preliminary inspection: Check the basic structure and content of the data to understand the size, type and distribution of the data;

[0134] A2. Handle missing values, identify missing values, and handle missing values;

[0135] Identify missing values: Use statistical methods or visualization tools to identify missing values;

[0136] Handling missing values: Choose appropriate methods to handle missing values, such as deletion, imputation, or predictive imputation;

[0137] A3. Processing outliers, including detecting and processing outliers;

[0138] Detect outliers: Use statistical methods (such as box plots, Z-score) or visualization tools (such as scatter plots) to detect outliers;

[0139] Handling outliers: Choose appropriate methods to handle outliers, such as deletion, replacement, or retention;

[0140] A4. Processing duplicate values, including detecting and processing duplicate values;

[0141] Detect duplicate values: Use hash tables or sorting methods to detect duplicate values;

[0142] Handling duplicate values: Choose an appropriate method to handle duplicate values, such as deleting or merging;

[0143] A5. Data formatting, including standardizing data formats and converting data types;

[0144] Unified data format: Ensure that all data uses a unified format, such as date format and numerical format;

[0145] Convert data types: convert data to the appropriate type, such as converting a string to a numeric value;

[0146] A6. Data consistency check, including checking logical consistency and cross-field consistency;

[0147] Check logical consistency: ensure that the data is logically reasonable, such as age should not be negative;

[0148] Check cross-field consistency: Ensure that data between different fields is consistent, such as gender and title should match;

[0149] A7. Data standardization, including standardized numerical values and standardized text;

[0150] Standardized numerical values: Standardize numerical data to have the same scale, such as converting numerical values to standard scores (Z - scores);

[0151] Standardized text: Standardize text data, such as converting all text to lowercase or removing special characters;

[0152] A8. Data verification, including verifying data sources and verifying data integrity;

[0153] Verifying data sources: Ensure that data sources are reliable and the data collection process complies with specifications;

[0154] Verifying data integrity: Ensure that the data is complete and no important information is missing;

[0155] A9. Data storage and backup, including storing cleaned data and backing up data;

[0156] Storing cleaned data: Store the cleaned data in an appropriate database or file system;

[0157] Backing up data: Regularly back up data to prevent data loss;

[0158] A10. Documentation, including recording the cleaning process and generating reports;

[0159] Recording the cleaning process: Detail every step of the data cleaning operation for future reference and reproduction; Generating reports: Generate data cleaning reports to summarize the cleaning process and results.

[0160] The methods for removing noisy data in data cleaning include the following:

[0161] B1. Text cleaning, including removing special characters, removing HTML tags, and removing stop words;

[0162] Removing special characters: Delete unnecessary special characters, such as punctuation marks and emojis;

[0163] Removing HTML tags: If the data comes from a web page, HTML tags need to be removed;

[0164] Removing stop words: Remove common meaningless words, such as "de", "le", "zai", etc.;

[0165] B2. Content filtering, including keyword filtering, length filtering, and duplicate filtering;

[0166] Keyword filtering: Filter out content irrelevant to the theme based on keywords;

[0167] Length filtering: remove text that is too short or too long, which may be noise data;

[0168] Duplicate filtering: remove duplicate text, which may be the result of crawler errors or human copying; keyword filtering:

[0169] Goal: Keep text that contains at least one keyword and filter out content that is irrelevant to the topic;

[0170]

[0171] Where: K = {k1, k2, k3...k n} is a predefined keyword set;

[0172] Text is the word segmentation result of the text to be filtered;

[0173] Length filtering:

[0174] Goal: Remove text that is too short or too long (measured in number of characters or words);

[0175]

[0176] Where: |Text| is the length of the text (number of characters or words)

[0177] L min and L max are the minimum and maximum retention length thresholds;

[0178] Repeat filtering:

[0179] Goal: Remove completely duplicate or highly similar texts;

[0180] First: accurate deduplication, judging exact duplication by hash value:

[0181]

[0182] Where: H(Text) is the hash value of the text (such as MD5, SHA-1);

[0183] D is the stored hash value set;

[0184] Second: Approximate deduplication, using a similarity threshold to determine duplication (such as Jaccard):

[0185]

[0186] Where: θ T ∈[0,1] is the similarity threshold;

[0187] D is the existing text collection;

[0188] B3, grammatical and semantic analysis, including grammatical checking and semantic analysis;

[0189] Grammar check: Use grammar checker to filter out texts with serious grammatical errors;

[0190] Semantic analysis: Use natural language processing techniques, such as word embedding and sentence embedding, to filter out text with unclear semantics or irrelevant to the topic;

[0191] B4. Statistical methods, including frequency analysis and anomaly detection;

[0192] Frequency analysis: Count the frequency of words or phrases and remove texts with abnormally high or low frequencies;

[0193] Anomaly detection: Use statistical methods such as Z-score and IQR to detect and remove abnormal data.

[0194] The multi-level collaboration in step 7 is implemented through cross-level collaboration, specifically:

[0195] Information sharing mechanism: Establish an information sharing mechanism to ensure smooth communication between all levels. For example, information can be quickly transmitted through the campus network, WeChat groups, etc.

[0196] Resource sharing platform: Establish a resource sharing platform to integrate various campus resources, such as teaching materials, scientific research results, etc., to achieve efficient resource sharing;

[0197] Joint training and exchange: Joint training and exchange activities are organized regularly, inviting managers and teachers at all levels to participate, to improve the overall management level and teaching quality.

[0198] An information sharing mechanism whose efficiency is based on:

[0199]

[0200] Where: E shared : The actual amount of information shared (such as the number of documents, number of items, etc.);

[0201] E total : The total amount of information to be shared;

[0202] T avg : Average time for information transmission (e.g., the immediacy of information transmission via the campus network or WeChat groups);

[0203] T delag : Delay time of information transmission;

[0204] Measures the efficiency of the information sharing mechanism. Higher values indicate more timely information transmission and wider coverage.

[0205] Resource sharing platform, whose utilization is based on:

[0206]

[0207] in: The actual usage of resources in category (i) (e.g., number of downloads of teaching materials, number of accesses to scientific research results);

[0208] The total stock of resources in category (i);

[0209] C collab : Number of cross-level collaborations (e.g., number of joint training or exchange activities);

[0210] Evaluate the resource integration and utilization efficiency of the resource sharing platform. The greater the number of collaborations, the more significant the logarithmic growth of resource utilization.

[0211] The multi-level collaboration in step 7 is also supported by technical support, specifically:

[0212] Information management system: Introduce advanced information management systems, such as the academic affairs management system and the student management system, to improve management efficiency and information transparency;

[0213] Data analysis and decision support: Utilize big data analysis technology to analyze and mine data at all levels to provide support for decision making;

[0214] Online collaboration tools: Promote the use of online collaboration tools, such as Google Docs and Microsoft Teams, to facilitate collaborative work among all levels.

[0215] Multi-level collaborative work, its effectiveness is based on:

[0216]

[0217] Where: T base : Task completion time when no online collaboration tools are used;

[0218] T tool : Task completion time after using online collaboration tools (such as Google Docs and Teams);

[0219] A data : Decision accuracy supported by data analysis;

[0220] A total : total number of decisions;

[0221] Quantitative technology supports the improvement of collaborative efficiency. The product of time reduction and improved decision-making accuracy reflects the overall efficiency gain.

[0222] Based on the above, the innovations include:

[0223] 1. Deep integration of collaborative governance mechanisms and vertical fields;

[0224] Through multi-level collaboration (school, department, and class) and role-based authority management, we deeply integrate educational management processes with vertical big model technologies, achieving a two-way empowerment of "governance-technology" in educational scenarios. For example, through information management systems and data analysis technologies, we standardize and intelligentize processes such as teaching management and resource allocation, solving the problems of hierarchical fragmentation and information silos in traditional education management.

[0225] 2. Standardized data cleaning process and enhanced vertical field features;

[0226] We propose standardized data cleaning processes for educational scenarios (such as knowledge point classification and difficulty grading), combined with denoising techniques such as grammatical / semantic analysis and statistical methods, to improve the professionalism and usability of educational data. For example, by removing advertisements and irrelevant content and annotating knowledge points and difficulty levels, we enable models to more accurately adapt to teaching needs (such as personalized learning recommendations).

[0227] 3. Multi-level collaborative implementation mechanism;

[0228] Through information sharing mechanisms, resource sharing platforms, and joint training, we achieve cross-level resource integration and capacity building. For example, by leveraging online collaboration tools (such as WeChat for Business and DingTalk) and data analysis technologies, we can connect school-level management with classroom execution, resolving the disconnect between decision-making and execution in traditional education management.

[0229] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some or all of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and description of the present invention.

Claims

1. A method for constructing a large vertical model system for a school based on collaborative governance, characterized by: A method for constructing a large vertical model system of a school based on collaborative governance is as follows: Step 1: Demand purpose, including demand analysis and goal setting; Step 2: System architecture design, including module division, data flow design and technology selection; Step 3: Data collection and integration, including data source determination, data cleaning, and data storage; Step 4: Model training and optimization, including feature engineering, model selection, model training, and model optimization; Step 5: System integration and testing, including module integration, system testing, and user training; Step 6: Operation and maintenance and continuous improvement, including system monitoring, user feedback, and continuous improvement; Step 7: Collaborative governance mechanism, including multi-level collaboration, role authority management and decision support.

2. The method for constructing a school vertical large model system based on collaborative governance according to claim 1 is characterized by: The data collection and integration in the third step are specifically as follows: Data source determination: Determine the data source required by the system; Data cleaning and integration: Clean, remove duplicates and standardize the collected data to ensure data accuracy and consistency; Data storage: Use distributed database or data warehouse technology to achieve efficient storage and fast query of large-scale data.

3. The method for constructing a large vertical model system of a school based on collaborative governance according to claim 2 is characterized by: The data collection and integration in the third step, data cleaning to remove noise data, ensure the accuracy and relevance of the data, and annotate the data to facilitate the model to better understand and generate content.

4. The method for constructing a large vertical model system of a school based on collaborative governance according to claim 3 is characterized by: The steps of data cleaning are as follows: A1. Data collection and preliminary inspection, including data collection and preliminary inspection; A2. Handle missing values, identify missing values, and handle missing values; A3. Processing outliers, including detecting and processing outliers; A4. Processing duplicate values, including detecting and processing duplicate values; A5. Data formatting, including standardizing data formats and converting data types; A6. Data consistency check, including checking logical consistency and cross-field consistency; A7. Data standardization, including standardized values and standardized text; A8. Data verification, including verification of data source and data integrity; A9. Data storage and backup, including storage of cleaned data and backup data; A10. Documentation, including recording the cleaning process and generating reports.

5. The method for constructing a large vertical model system of a school based on collaborative governance according to claim 4 is characterized by: The methods for removing noise data in cleaning data include the following: B1. Text cleaning, including removing special characters, HTML tags, and stop words; B2. Content filtering, including keyword filtering, length filtering, and duplicate filtering; B3, grammatical and semantic analysis, including grammatical checking and semantic analysis; B4. Statistical methods, including frequency analysis and anomaly detection.

6. The method for constructing a large vertical model system of a school based on collaborative governance according to claim 5 is characterized by: The content filtering in B2 is specifically as follows: Keyword filtering: filter out content irrelevant to the topic based on keywords; Length filtering: remove text that is too short or too long; Duplicate Filtering: Remove duplicate text.

7. The method for constructing a large vertical model system of a school based on collaborative governance according to claim 6 is characterized by: The statistical method in B4 is specifically: Frequency analysis: Count the frequency of words or phrases and remove texts with abnormally high or low frequencies; Anomaly detection: using statistical methods.

8. The method for constructing a large vertical model system of a school based on collaborative governance according to claim 7 is characterized by: The collaborative governance mechanism in the seventh step is specifically: Multi-level collaboration: Establish collaborative governance mechanisms at the university, department, and class levels to ensure smooth information flow and resource sharing among all levels; Role and authority management: Set corresponding permissions based on the roles and responsibilities of different users to ensure the security and controllability of the system; Decision support: Utilize big data and artificial intelligence technologies to provide scientific decision support for school administrators and improve the scientific nature and effectiveness of management decisions.

9. The method for constructing a large vertical model system of a school based on collaborative governance according to claim 8 is characterized by: The multi-level collaboration in the seventh step is implemented through cross-level collaboration, specifically: Information sharing mechanism: Establish an information sharing mechanism to ensure smooth information flow between all levels; Resource sharing platform: Establish a resource sharing platform to integrate various resources within the school and realize efficient resource sharing; Joint training and exchange: Joint training and exchange activities are organized regularly, inviting managers and teachers at all levels to participate, to improve the overall management level and teaching quality.

10. The method for constructing a large vertical model system of a school based on collaborative governance according to claim 8 is characterized by: The multi-level collaboration in the seventh step is also supported by technical support operations, specifically: Information management system: Introducing advanced information management system; Data analysis and decision support: Utilize big data analysis technology to analyze and mine data at all levels to provide support for decision making; Online collaboration tools: Promote the use of online collaboration tools to facilitate collaborative work among all levels.

Images