Teacher material online examination and approval processing method and system and terminal equipment

Abstract

The invention discloses a teacher material online examination and approval processing method and system and terminal equipment, and belongs to the technical field of teacher material management.The method comprises the steps that an attachment document sequence is constructed by analyzing the logic relation between teacher materials of different sources, extracting a data unit from the attachment document sequence according to a preset standard document template; performing semantic matching and data source mapping on the data unit and the fillable fields in the standard document template to obtain a to-be-approved file; and calling a teaching information management system corresponding to the service label to perform rationality evaluation and abnormal point identification on the to-be-approved file to obtain a first approval result, and marking a to-be-rechecked point in the to-be-approved file to construct an approval feedback report of the to-be-approved file. Therefore, through implementation of the method and the device, the problems of relatively high approval cost and relatively low approval efficiency caused by incapability of performing more professional approval on authenticity and the like of teacher materials in the prior art can be solved.

Description

Technical Field

[0001] This application belongs to the field of teacher materials management technology, specifically involving an online approval and processing method, system and terminal equipment for teacher materials. Background Technology

[0002] Against the backdrop of educational informatization and the construction of digital campuses, the traditional offline approval model relying on paper documents is no longer adequate to meet the needs of modern schools for efficient, transparent, and collaborative management. Unlike general documents, teacher materials include different types of documents such as teaching syllabi, textbooks, and student records. The rigor of their content is crucial for teacher materials. Therefore, the approval of teacher materials, as a core link in ensuring teaching quality and maintaining educational equity, is of paramount importance in terms of accuracy and reliability, and the focus of the approval process differs from that of other materials.

[0003] Online approval systems allow users to handle approval matters anytime, anywhere via computers and other devices, eliminating the need for physical presence and significantly shortening the approval cycle. However, most current approval systems only check the format and compliance of materials, failing to assess the quality of the content. Therefore, existing online approval systems struggle to automatically determine if teacher materials align with current teaching objectives, necessitating further manual review by experienced teaching staff. This results in inefficient approval processes and high labor costs. Furthermore, integrating teacher materials from diverse sources into a single, compliant format further diminishes the convenience of online approval. Summary of the Invention

[0004] This application proposes an online approval and processing method, system, and terminal device for teacher materials, which can solve the problems of existing technologies that cannot conduct more professional approval of teacher materials and require a lot of time to integrate teacher documents from different sources, resulting in high approval costs and low approval efficiency.

[0005] The first aspect of this application provides a method for online approval and processing of teacher materials, the method comprising: A sequence of attachment documents is constructed by analyzing the logical relationships between teacher materials from different sources, and data units are extracted from the attachment document sequence according to a preset standard document template; wherein, the standard document template is determined by the business tags of the teacher materials; Semantic matching and data source mapping are performed between the data units and the fillable fields in the standard document template to obtain the document to be approved. According to the preset teaching requirements, the teaching information management system corresponding to the business tag is invoked to conduct a rationality assessment and anomaly identification on the document to be approved, and a first approval result is obtained; Based on the first approval result, mark the points to be reviewed in the document to be approved and add corresponding unreasonable evidence at the points to be reviewed, and construct the approval feedback report of the document to be approved.

[0006] The above solution first integrates the logical relationships between teacher materials from different sources, determining the connections and order between the materials. Then, based on these logical relationships, the heterogeneous teacher materials are sorted and integrated into a sequence of attachment documents. This sequence guides the extraction of content required for a standard document template, automatically integrating heterogeneous data into documents that meet approval requirements, effectively improving approval efficiency and reducing document complexity. To enhance the quality of data integration, data units and fillable fields are matched to find semantically similar content for fillable fields. Furthermore, data source mapping ensures that all filled fields have data traceability capabilities, allowing approvers to directly view the original data. The solution uses business tags to call the teaching information management system to check whether teacher materials are suitable for teaching and can achieve the established teaching objectives. Compared to general material approval, this provides more professional approval methods and knowledge, eliminating the need for experienced teaching staff to conduct detailed reviews of teacher materials. It can quickly and accurately identify anomalies that contradict teaching requirements or deviate significantly from them, marking these anomalies for further detailed review by teaching staff, greatly reducing manual workload. Furthermore, it provides relevant justifications for any unreasonable aspects of the review process, making it easier for users to understand and ensuring that the approval results are more reliable.

[0007] In one possible implementation of the first aspect, a sequence of attachment documents is constructed by analyzing the logical relationships between teacher materials from different sources, and data units are extracted from the sequence of attachment documents according to a preset standard document template, specifically as follows: To conform to the thematic order of the standard document template, the teacher materials are dimensionally aligned with the standard document template to obtain the logical relationship; Based on the logical relationships and the preset approval document directory, all the teacher materials are sorted and renamed to obtain the attachment document sequence; Based on the standard document template, the data unit is obtained by extracting data and performing structured transformation on the attached document sequence.

[0008] The above solution uses the thematic order of a standard document template as a benchmark to analyze the logical relationships between heterogeneous teacher materials. By standardizing and reordering the naming of teacher materials, they are integrated into a standard sequence of attachment documents that conforms to the order of the standard document template, providing support for subsequent data population. Because the attachment document sequence conforms to the dimensional and thematic order of attachment documents, data content can be quickly extracted from the attachment document sequence and converted into structured data units that can be used for data population.

[0009] In one possible implementation of the first aspect, semantic matching and data source mapping are performed between the data unit and the fillable fields in the standard document template to obtain the document to be approved, specifically: Based on a preset vocabulary knowledge graph, the contextual semantics of each data unit are matched with the text description of the fillable field by reasoning the meaning of the words, and the data unit that is successfully matched is taken as the unit to be filled. Based on the matching results, a dynamic index is constructed between the cell to be filled and the fillable field by mapping the data source information of the cell to be filled with the position information of the fillable field; wherein, the data source information records the teacher material to which the cell to be filled belongs and the position of the cell to be filled in the attachment document sequence; Based on the dynamic index, the cells to be filled are automatically filled into the standard document template to generate a document awaiting approval.

[0010] The above solution uses semantic matching to find matching data units for fillable fields in the template, automatically integrating teacher materials from different sources into a document ready for electronic approval, eliminating the need for manual document creation. Furthermore, data source mapping enables the filled fields to have traceability capabilities, allowing approvers to see which attachment a field originated from during the review process, improving the accuracy and efficiency of the approval process.

[0011] In one possible implementation of the first aspect, based on a preset lexical knowledge graph, the contextual semantics of each data unit are matched with the text description of the fillable field by reasoning about the meaning of the words, specifically: Based on the contextual semantics and the text description, the positions of the data unit and the fillable field in the vocabulary knowledge graph are determined respectively, and the concept node of the data unit and the meaning node of the fillable field are obtained. Calculate the path distance between the concept node and the meaning node, and obtain the matching degree between the data unit and the fillable field by comparing the magnitude of the path distance.

[0012] The above solution addresses the polysemy between data units and fillable fields by introducing a lexical knowledge graph, finding the most suitable data unit for each fillable field and improving the accuracy of semantic matching. Because the graph itself provides traceable paths and node relationships, the semantic matching results are highly interpretable and easy for users to understand.

[0013] In one possible implementation of the first aspect, based on preset teaching requirements, the teaching information management system corresponding to the business tag is invoked to perform a rationality assessment and anomaly identification on the document to be approved, thereby obtaining a first approval result, specifically as follows: Based on the text titles in the document to be approved, the document to be approved is segmented into several modules to be approved. Extract the review criteria related to the business tags from the teaching requirements; Based on the teaching information management system, anomalies that do not meet the review criteria in the module to be approved are detected by a preset large language model, and a second approval result is obtained.

[0014] The above scheme divides the documents to be approved into multiple modules based on their content, facilitating subsequent content analysis. Then, it analyzes whether each module meets relevant review standards to determine the rationality of the content and identify any anomalies that do not conform to teaching objectives, thus achieving a professional review of the teachers' materials.

[0015] In one possible implementation of the first aspect, based on the teaching information management system, anomalies that do not meet the review criteria in the module to be approved are detected by a preset large language model to obtain a second approval result, specifically as follows: Based on the aforementioned review criteria, the modules to be approved are clustered using the large language model. The teaching information management system is invoked based on the clustering results to obtain the standard text corresponding to the clustering results; Based on the priority of the review criteria and the publication time of the standard text, a weighted calculation is performed on the similarity between the module to be approved and the standard text to obtain a reasonableness score for the module to be approved. Modules to be approved with reasonableness scores lower than a first threshold are marked as anomalies. The later the publication time, the higher the corresponding weight coefficient.

[0016] The aforementioned scheme uses a large language model to identify the linguistic features of the document to be approved and assess its similarity to the standard text. This measures the degree of conformity between the document to be approved and the review standards, thereby obtaining the initial approval result that characterizes whether the document can achieve its teaching objectives. Furthermore, considering the focus of the review standards during the detection process and the version of the standard text, the similarity is weighted to ensure that the approval process focuses on content requiring key review and content that is more in line with the latest standard text, thus improving the accuracy of the assessment.

[0017] In one possible implementation of the first aspect, based on the first approval result, points requiring review in the document to be approved are marked, and corresponding evidence of unreasonableness is added at the points requiring review, thereby constructing an approval feedback report for the document to be approved, specifically as follows: Based on the first approval result, mark the points to be reviewed in the documents pending approval; Obtain the call information of the teaching information management system, perform data trimming and keyword extraction on the call information, and obtain the unreasonable proof of the point to be reviewed; Based on the text type of the point to be reviewed, the unreasonable proof is formatted and then displayed hierarchically at the corresponding point to be reviewed, thus obtaining the approval feedback report.

[0018] The above solution adds supporting evidence to the points to be reviewed, explaining why they are unreasonable or why there are abnormalities. This provides auxiliary prompts for the subsequent approval process of teaching staff, speeds up the approval process and reduces approval costs. It also makes it easier for users who upload teacher materials to correct any unreasonable aspects of the documents.

[0019] In one possible implementation of the first aspect, the approval feedback report also includes: Based on the results of the first approval, the points pending review are ranked according to their severity. If the sorting results contain more than the second threshold of low-quality content, then, in conjunction with the teaching requirements, the rejection reason for the approval feedback report will be generated.

[0020] The above scheme will automatically generate rejection reasons for approval feedback reports with serious problems, providing a reasonable explanation for the rejection of teachers' materials.

[0021] The second aspect of this application provides an online approval and processing system for teacher materials, the system comprising: a material classification module, an authenticity detection module, a rationality detection module, and a feedback report generation module; The data extraction module is used to construct a sequence of attachment documents by analyzing the logical relationships between teacher materials from different sources, and to extract data units from the sequence of attachment documents according to a preset standard document template; wherein, the standard document template is determined by the business tags of the teacher materials; The data population module is used to perform semantic matching and data source mapping between the data unit and the fillable fields in the standard document template to obtain the document to be approved; The data pre-approval module is used to call the teaching information management system corresponding to the business tag to conduct a rationality assessment and anomaly identification of the document to be approved according to the preset teaching requirements, and obtain the first approval result; The approval report generation module is used to mark the points to be reviewed in the document to be approved based on the first approval result, add corresponding unreasonable evidence at the points to be reviewed, and construct the approval feedback report of the document to be approved.

[0022] A third aspect of this application provides a terminal device, the device comprising: a terminal device including a processor and a memory, the memory storing a computer program, wherein the processor executes the computer program to implement the steps of the online approval and processing method for teacher materials as described in any one of the embodiments of this application. Attached Figure Description

[0023] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0024] Figure 1 This is a schematic diagram illustrating the specific process of an online approval and processing method for teacher materials provided in one embodiment of this application; Figure 2 This is a structural diagram of an online approval and processing system for teacher materials provided in one embodiment of this application; Figure 3 This application provides a structural diagram of a terminal device according to one embodiment. Detailed Implementation

[0025] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0026] It should be understood that the step numbers used in the text are for ease of description only and are not intended to limit the order in which the steps are performed.

[0027] First Embodiment Teacher materials mainly include textbooks, lesson plans, and exercise materials used to achieve various teaching objectives. Compared with general materials, teacher materials place greater emphasis on their rigor and their ability to serve teaching goals. Therefore, teacher materials have a higher professional threshold, typically requiring detailed review by teaching-related personnel. This review process demands substantial professional knowledge, making manual review time-consuming and inefficient. This application proposes an online method for reviewing teacher materials. It intelligently detects the rationality of teacher materials and integrates heterogeneous materials from different sources into a single document that meets approval standards. This effectively reduces the workload of manual review and provides sufficient and reasonable explanations for the approval results.

[0028] like Figure 1 As shown, to address the problems in existing technologies where teacher materials cannot be professionally reviewed and require significant time to integrate teacher documents from different sources, resulting in high review costs and low efficiency, the first embodiment of this application provides a detailed flowchart of an online teacher material review process. This online teacher material review process includes steps S1 to S4, detailed below: Step S1: Construct a sequence of attachment documents by analyzing the logical relationships between teacher materials from different sources, and extract data units from the sequence of attachment documents according to a preset standard document template.

[0029] In this embodiment of the application, for more complex approval processes, clients need to upload multiple different teacher materials, which may vary in origin and format. For example, for teaching qualification certification, users are required to upload their teacher qualification certificate, a copy of their ID card, teaching experience, and a personal qualification evaluation form. Before uploading, users need to integrate these documents together to facilitate subsequent electronic review.

[0030] However, integrating teacher materials from different sources and formats into a single document awaiting approval requires a significant amount of manual time and effort for format conversion, file renaming, and data concatenation, which undoubtedly reduces the efficiency of electronic approval. Therefore, to address these challenges, this application's embodiment constructs logical relationships between teacher materials from different sources, directly extracting data units that might be needed from a pre-defined standard document template. By automatically constructing a document awaiting approval that conforms to approval business standards, it reduces the amount of manual work required from users.

[0031] First, based on the business tags of the uploaded teacher materials, the corresponding standard document template is determined. These business tags are determined by the teacher material upload interface and describe the approval process corresponding to the teacher materials.

[0032] Using the thematic order of the standard document template as a benchmark, the logical relationships between multiple teacher materials are clarified. Specifically, character recognition technology is used to parse each teacher material, determine its function and included information, and align the teacher materials with the standard document template based on this information. This identifies which part of the standard document template needs to reference content from the teacher materials, as well as the referencing order among the teacher materials, thus obtaining the corresponding logical relationships. Based on these logical relationships and a pre-defined approval document directory, all teacher materials are renamed to improve their standardization, and then the teacher materials are sorted as attachments to generate a corresponding sequence of attachment documents.

[0033] Based on the established standard document template, structured, unstructured, and semi-structured data are extracted from the attached document sequence. The extracted data is then formatted and transformed to obtain structured data units that are filled into the template. Optionally, for parsing teachers' materials, if the material is an image file, text information is extracted using optical character recognition technology; if the material is a layout file, its internal text stream and objects are parsed; if the material is a structured document, its tag or key-value pair information is read directly. Unstructured files are generally images or scanned documents, while semi-structured files are generally Word or PDF data with inconsistent formats.

[0034] The approval document directory specifies the file format and name of the teacher materials required for uploading for various approval processes.

[0035] Step S2: Perform semantic matching and data source mapping between the data unit and the fillable fields in the standard document template to obtain the document to be approved.

[0036] This application's embodiments introduce a lexical knowledge graph to solve the problem of polysemy, determine the data unit that best matches the fillable field, and achieve accurate data filling.

[0037] The vocabulary knowledge graph gathers the meanings of words that appear in the currently approved teacher materials. It can treat the meaning of a word as a node, and analyze the relationship between nodes by exploring the "edges" around the node, thereby unifying the meanings of multiple words and achieving semantic matching between different words.

[0038] First, obtain the contextual semantics of the data unit, and then obtain the textual description of the fillable field based on the information function and the wording of the field.

[0039] For example, the information provided for "Obtained ______ (certificate name) on _____ year __ month __ day" is a text description for a populateable field used to obtain the user's XXX qualification certificate acquisition time.

[0040] Based on the contextual semantics and the word meanings provided by the text description, the positions of data units and fillable fields in the lexical knowledge graph are determined, and then the concept nodes corresponding to the data units and the meaning nodes corresponding to the fillable fields are determined.

[0041] The degree of matching between the data unit and the fillable field is measured by calculating the closeness of the association between concept nodes and meaning nodes in the knowledge graph. Specifically, based on the lexical knowledge graph, the path distance from the concept node to the meaning node is calculated. Then, the magnitude of the path distance is compared to determine the semantic similarity between each concept node and the meaning node. The closer the distance, the stronger the association, and thus the degree of matching between the data unit and the fillable field is determined.

[0042] Because lexical knowledge graphs provide traceable node paths and relationships, they are highly interpretable and can accurately eliminate polysemy based on the relationships between words, providing strong data support for subsequent data filling.

[0043] If the matching degree meets the set threshold, the corresponding data unit can be considered to be filled into the fillable field, and the fillable field is used as the unit to be filled.

[0044] Obtain the data source information for the cell to be populated, including the teacher material to which the cell belongs and its name, and the cell's position in the attached document sequence. Simultaneously, obtain the position information of the populated fields, i.e., the position of the populated fields in the standard document template.

[0045] Based on the cells to be filled provided by the matching results, the data source information is mapped to the location information, and this mapping relationship is used as a dynamic index between the cells to be filled and the fillable fields. The dynamic index is inserted into the cells to be filled, and then the cells to be filled are automatically filled into the standard document template to generate an integrated document awaiting approval. In the document awaiting approval, each filled field has a data traceability function, which can trace back to its corresponding original data cell and source file through the dynamic index. Approver can directly locate and view the data source and original form of the filled data by clicking on the index, improving the convenience of subsequent approval and review.

[0046] Step S3: Based on the preset teaching requirements, the teaching information management system corresponding to the business tag is invoked to conduct a rationality assessment and anomaly identification on the document to be approved, and a first approval result is obtained.

[0047] Determining whether a teacher's application materials conform to teaching objectives / standards is also one of the approval criteria. Existing technologies generally involve experienced teachers approving the materials to determine their reasonableness, educational value, and suitability for the teaching environment. However, this approval method is highly inefficient and has a high barrier to entry. To address these shortcomings, this application's embodiment first performs intelligent detection on the teacher's materials, identifying anomalies that do not conform to teaching objectives. This provides data support for subsequent manual review of these anomalies, effectively reducing the workload and improving the efficiency of the materials approval process.

[0048] Firstly, to improve approval efficiency, documents awaiting approval are divided into multiple approval modules based on different text titles, ensuring that the content of each approval module is built around a single text title, making it easier to conduct a rationale analysis.

[0049] The system extracts review criteria related to business tags from the pre-defined teaching requirements. To meet these criteria, a pre-trained large language model is used to cluster the modules to be approved, resulting in text vectors for each cluster center. The system then calls the teaching information management system corresponding to the business tags to obtain the standard text for each cluster center and calculates the similarity between each text vector to be evaluated and the standard text to assess the reasonableness of the modules. Finally, based on the reasonableness assessment results, anomalies in the teacher's materials that do not conform to the teaching objectives are marked, resulting in the first approval result.

[0050] Specifically, based on the large language model and a predefined reference set, the modules to be approved are clustered using the reference set as cluster centers to obtain the text vectors to be evaluated. The reference set contains a series of generally accepted texts that conform to the review criteria. Then, based on the clustering results and business tags, the teaching information management system is invoked to obtain standard texts related to the clustering results. These standard texts are data that has been published on the teaching information management system, conforms to the review criteria, and meets the document type required for the approval process.

[0051] For example, the uploaded lesson plans are clustered according to different content themes, and then the smart teaching platform is called to obtain lesson plan data that is related to the content of the lesson plans and the approval process and meets the review standards.

[0052] Based on the degree to which the module to be approved meets the aforementioned review criteria, the similarity between the module and the standard text is calculated. During the calculation, the priority of the review criteria is taken into account, and the similarity is weighted and adjusted to obtain a reasonableness score for each module to be approved. A higher priority review criterion indicates a higher similarity between the two.

[0053] As an improvement to the above scheme, the publication date of the standard text is also considered when calculating the reasonableness score. The more similar the text is to the standard text published later, the higher the weighting coefficient used in the calculation, and the higher the corresponding reasonableness score. In other words, the more similar the text is to the latest version, the more it conforms to the review criteria.

[0054] Finally, modules whose rationality score is lower than the first threshold are marked as anomalies, indicating that the modules to be approved have significant discrepancies with the actual review standards and require manual re-approval to determine whether the documents to be approved meet the teaching objectives.

[0055] Based on the above approval results, the first approval result that can measure the reasonableness of the materials is obtained.

[0056] Step S4: Based on the first approval result, mark the points to be reviewed in the document to be approved and add corresponding unreasonable evidence at the points to be reviewed, and construct the approval feedback report of the document to be approved.

[0057] Based on the initial approval results, key review points in the pending documents are marked. These review points are essentially content that failed authenticity verification or does not conform to the teaching objectives. Simultaneously, records of calls to the teaching information management system during previous approval processes are retrieved to obtain call information. This call information is then processed through data trimming and keyword extraction to extract and refine information relevant to the review points, revealing any unreasonable claims.

[0058] Based on the text type of the point to be reviewed, the unreasonable evidence is formatted to obtain information that matches the format of the point to be reviewed. Then, the formatted unreasonable evidence is inserted into the corresponding point to be reviewed for display. During manual review, this clearly demonstrates why the content needs review and why it was judged as unreasonable, and a corresponding approval feedback report is generated.

[0059] Furthermore, this application embodiment also provides a tiered display of unreasonable proofs, including strong, medium, and weak prompts. Strong prompts are used to mark key errors and usually appear as red warning boxes; medium prompts are used as warnings and usually appear as yellow prompt boxes to draw the approver's attention; weak prompts appear with light-colored backgrounds or small floating icons.

[0060] Additionally, if images or text cannot fully demonstrate the unreasonable proof, the data source corresponding to the unreasonable proof can be inserted as an address index at the point to be reviewed.

[0061] To improve the quality of the approval feedback report, the points to be reviewed can be ranked according to their severity based on the feedback information from the first approval result, resulting in a review sequence. If more than a set threshold of low-quality content with high severity exists in the review sequence, the document to be approved is considered to have significant flaws and needs to be rejected and revised. The rejection reason in the approval feedback report is then generated, taking into account the teaching requirements related to the business tags. This rejection reason provides a detailed explanation of the errors and inconsistencies in the document to be approved and the teacher's materials, along with relevant supporting documentation, guiding the uploader to correct the teacher's materials.

[0062] Implementing the embodiments of this application has the following beneficial effects: This application first integrates the logical relationships between teacher materials from different sources, determining the connections and order between the materials. Then, based on these logical relationships, the heterogeneous teacher materials are sorted and integrated into a sequence of attachment documents. This sequence guides the extraction of content required for a standard document template, automatically integrating heterogeneous data into a document ready for approval that meets the approval requirements, effectively improving approval efficiency. To enhance the quality of data integration, data units and fillable fields are matched to find semantically similar content for fillable fields. Furthermore, data source mapping ensures that all filled fields have data traceability capabilities, allowing approvers to directly view the original data. The application calls the teaching information management system based on business tags to check whether teacher materials are suitable for teaching and can achieve the established teaching objectives. Compared to general material approval, this provides more professional approval methods and knowledge, eliminating the need for experienced teaching staff to conduct detailed reviews of teacher materials. It can quickly and accurately identify anomalies that contradict teaching requirements or deviate significantly from them, marking these anomalies for further detailed review by teaching staff, greatly reducing the workload. Furthermore, it provides relevant justifications for any unreasonable aspects of the review process, making it easier for users to understand and ensuring that the approval results are more reliable.

[0063] Second Embodiment Furthermore, in order to implement the online teacher material approval and processing system corresponding to the above method embodiments, and to achieve the corresponding functions and technical effects, Figure 2 A structural diagram of an online approval and processing system for teacher materials is provided. For ease of explanation, only the parts relevant to this embodiment are shown. The online approval and processing system for teacher materials provided in this embodiment includes: The data extraction module 201 is used to construct a sequence of attachment documents by analyzing the logical relationships between teacher materials from different sources, and to extract data units from the sequence of attachment documents according to a preset standard document template.

[0064] In this embodiment of the application, for more complex approval processes, clients need to upload multiple different teacher materials, which may vary in origin and format. For example, for teaching qualification certification, users are required to upload their teacher qualification certificate, a copy of their ID card, teaching experience, and a personal qualification evaluation form. Before uploading, users need to integrate these documents together to facilitate subsequent electronic review.

[0065] However, integrating teacher materials from different sources and formats into a single document awaiting approval requires a significant amount of manual time and effort for format conversion, file renaming, and data concatenation, which undoubtedly reduces the efficiency of electronic approval. Therefore, to address these challenges, this application's embodiment constructs logical relationships between teacher materials from different sources, directly extracting data units that might be needed from a pre-defined standard document template. By automatically constructing a document awaiting approval that conforms to approval business standards, it reduces the amount of manual work required from users.

[0066] First, based on the business tags of the uploaded teacher materials, the corresponding standard document template is determined. These business tags are determined by the teacher material upload interface and describe the approval process corresponding to the teacher materials.

[0067] Using the thematic order of the standard document template as a benchmark, the logical relationships between multiple teacher materials are clarified. Specifically, character recognition technology is used to parse each teacher material, determine its function and included information, and align the teacher materials with the standard document template based on this information. This identifies which part of the standard document template needs to reference content from the teacher materials, as well as the referencing order among the teacher materials, thus obtaining the corresponding logical relationships. Based on these logical relationships and a pre-defined approval document directory, all teacher materials are renamed to improve their standardization, and then the teacher materials are sorted as attachments to generate a corresponding sequence of attachment documents.

[0068] Based on the established standard document template, structured, unstructured, and semi-structured data are extracted from the attached document sequence. The extracted data is then formatted and transformed to obtain structured data units that are filled into the template. Optionally, for parsing teachers' materials, if the material is an image file, text information is extracted using optical character recognition technology; if the material is a layout file, its internal text stream and objects are parsed; if the material is a structured document, its tag or key-value pair information is read directly. Unstructured files are generally images or scanned documents, while semi-structured files are generally Word or PDF data with inconsistent formats.

[0069] The approval document directory specifies the file formats and names required for uploading for various approval processes related to teacher materials.

[0070] The data population module 202 is used to perform semantic matching and data source mapping between the data unit and the fillable fields in the standard document template to obtain the document to be approved.

[0071] In this embodiment of the application, based on a preset vocabulary knowledge graph, the contextual semantics of each data unit are matched with the text description of the fillable field by reasoning the meaning of the vocabulary, and the data unit that is successfully matched is taken as the unit to be filled. Based on the matching results, a dynamic index is constructed between the cell to be filled and the fillable field by mapping the data source information of the cell to be filled with the position information of the fillable field; wherein, the data source information records the teacher material to which the cell to be filled belongs and the position of the cell to be filled in the attachment document sequence; Based on the dynamic index, the cells to be filled are automatically filled into the standard document template to generate a document awaiting approval.

[0072] The data pre-approval module 203 is used to call the teaching information management system corresponding to the business tag to perform a rationality assessment and anomaly identification on the document to be approved according to the preset teaching requirements, and obtain the first approval result.

[0073] In this embodiment of the application, the document to be approved is segmented into several modules to be approved based on the text title in the document to be approved; Extract the review criteria related to the business tags from the teaching requirements; Based on the teaching information management system, anomalies that do not meet the review criteria in the module to be approved are detected by a preset large language model, and a first approval result is obtained.

[0074] The feedback report generation module 204 is used to mark the points to be reviewed in the document to be approved based on the first approval result, add corresponding unreasonable evidence at the points to be reviewed, and construct the approval feedback report of the document to be approved.

[0075] In this embodiment of the application, the points to be reviewed in the document to be approved are marked according to the first approval result; Obtain the call information of the teaching information management system, perform data trimming and keyword extraction on the call information, and obtain the unreasonable proof of the point to be reviewed; Based on the text type of the point to be reviewed, the unreasonable proof is formatted and then displayed hierarchically at the corresponding point to be reviewed, thus obtaining the approval feedback report.

[0076] In some embodiments, the data filling module 202 specifically comprises: This application's embodiments introduce a lexical knowledge graph to solve the problem of polysemy, determine the data unit that best matches the fillable field, and achieve accurate data filling.

[0077] The vocabulary knowledge graph gathers the meanings of words that appear in the currently approved teacher materials. It can treat the meaning of a word as a node, and analyze the relationship between nodes by exploring the "edges" around the node, thereby unifying the meanings of multiple words and achieving semantic matching between different words.

[0078] First, obtain the contextual semantics of the data unit, and then obtain the textual description of the fillable field based on the information function and the wording of the field.

[0079] For example, the information provided for "Obtained ______ (certificate name) on _____ year __ month __ day" is a text description for a populateable field used to obtain the user's XXX qualification certificate acquisition time.

[0080] Based on the contextual semantics and the word meanings provided by the text description, the positions of data units and fillable fields in the lexical knowledge graph are determined, and then the concept nodes corresponding to the data units and the meaning nodes corresponding to the fillable fields are determined.

[0081] The degree of matching between the data unit and the fillable field is measured by calculating the closeness of the association between concept nodes and meaning nodes in the knowledge graph. Specifically, based on the lexical knowledge graph, the path distance from the concept node to the meaning node is calculated. Then, the magnitude of the path distance is compared to determine the semantic similarity between each concept node and the meaning node. The closer the distance, the stronger the association, and thus the degree of matching between the data unit and the fillable field is determined.

[0082] Because lexical knowledge graphs provide traceable node paths and relationships, they are highly interpretable and can accurately eliminate polysemy based on the relationships between words, providing strong data support for subsequent data filling.

[0083] If the matching degree meets the set threshold, the corresponding data unit can be considered to be filled into the fillable field, and the fillable field is used as the unit to be filled.

[0084] Obtain the data source information for the cell to be populated, including the teacher material to which the cell belongs and its name, and the cell's position in the attached document sequence. Simultaneously, obtain the position information of the populated fields, i.e., the position of the populated fields in the standard document template.

[0085] Based on the cells to be filled provided by the matching results, the data source information is mapped to the location information, and this mapping relationship is used as a dynamic index between the cells to be filled and the fillable fields. The dynamic index is inserted into the cells to be filled, and then the cells to be filled are automatically filled into the standard document template to generate an integrated document awaiting approval. In the document awaiting approval, each filled field has a data traceability function, which can trace back to its corresponding original data cell and source file through the dynamic index. Approver can directly locate and view the data source and original form of the filled data by clicking on the index, improving the convenience of subsequent approval and review.

[0086] In some embodiments, the data pre-approval module 203 specifically comprises: Determining whether a teacher's application materials conform to teaching objectives / standards is also one of the approval criteria. Existing technologies generally involve experienced teachers approving the materials to determine their reasonableness, educational value, and suitability for the teaching environment. However, this approval method is highly inefficient and has a high barrier to entry. To address these shortcomings, this application's embodiment first performs intelligent detection on the teacher's materials, identifying anomalies that do not conform to teaching objectives. This provides data support for subsequent manual review of these anomalies, effectively reducing the workload and improving the efficiency of the materials approval process.

[0087] Firstly, to improve approval efficiency, documents awaiting approval are divided into multiple approval modules based on different text titles, ensuring that the content of each approval module is built around a single text title, making it easier to conduct a rationale analysis.

[0088] The system extracts review criteria related to business tags from the pre-defined teaching requirements. To meet these criteria, a pre-trained large language model is used to cluster the modules to be approved, resulting in text vectors for each cluster center. The system then calls the teaching information management system corresponding to the business tags to obtain the standard text for each cluster center and calculates the similarity between each text vector to be evaluated and the standard text to assess the reasonableness of the modules. Finally, based on the reasonableness assessment results, anomalies in the teacher's materials that do not conform to the teaching objectives are marked, resulting in the first approval result.

[0089] Specifically, based on the large language model and a predefined reference set, the modules to be approved are clustered using the reference set as cluster centers to obtain the text vectors to be evaluated. The reference set contains a series of generally accepted texts that conform to the review criteria. Then, based on the clustering results and business tags, the teaching information management system is invoked to obtain standard texts related to the clustering results. These standard texts are data that has been published on the teaching information management system, conforms to the review criteria, and meets the document type required for the approval process.

[0090] For example, the uploaded lesson plans are clustered according to different content themes, and then the smart teaching platform is called to obtain lesson plan data that is related to the content of the lesson plans and the approval process and meets the review standards.

[0091] Based on the degree to which the module to be approved meets the aforementioned review criteria, the similarity between the module and the standard text is calculated. During the calculation, the priority of the review criteria is taken into account, and the similarity is weighted and adjusted to obtain a reasonableness score for each module to be approved. A higher priority review criterion indicates a higher similarity between the two.

[0092] As an improvement to the above scheme, the publication date of the standard text is also considered when calculating the reasonableness score. The more similar the text is to the standard text published later, the higher the weighting coefficient used in the calculation, and the higher the corresponding reasonableness score. In other words, the more similar the text is to the latest version, the more it conforms to the review criteria.

[0093] Finally, modules whose rationality score is lower than the first threshold are marked as anomalies, indicating that the modules to be approved have significant discrepancies with the actual review standards and require manual re-approval to determine whether the documents to be approved meet the teaching objectives.

[0094] Based on the above approval results, the first approval result that can measure the reasonableness of the materials is obtained.

[0095] In some embodiments, the approval report generation module 204 specifically comprises: Based on the initial approval results, key review points in the pending documents are marked. These review points are essentially content that failed authenticity verification or does not conform to the teaching objectives. Simultaneously, records of calls to the teaching information management system during previous approval processes are retrieved to obtain call information. This call information is then processed through data trimming and keyword extraction to extract and refine information relevant to the review points, revealing any unreasonable claims.

[0096] Based on the text type of the point to be reviewed, the unreasonable evidence is formatted to obtain information that matches the format of the point to be reviewed. Then, the formatted unreasonable evidence is inserted into the corresponding point to be reviewed for display. During manual review, this clearly demonstrates why the content needs review and why it was judged as unreasonable, and a corresponding approval feedback report is generated.

[0097] Furthermore, this application embodiment also provides a tiered display of unreasonable proofs, including strong, medium, and weak prompts. Strong prompts are used to mark key errors and usually appear as red warning boxes; medium prompts are used as warnings and usually appear as yellow prompt boxes to draw the approver's attention; weak prompts appear with light-colored backgrounds or small floating icons.

[0098] Additionally, if images or text cannot fully demonstrate the unreasonable proof, the data source corresponding to the unreasonable proof can be inserted as an address index at the point to be reviewed.

[0099] To improve the quality of the approval feedback report, the points to be reviewed can be ranked according to their severity based on the feedback information from the first approval result, resulting in a review sequence. If more than a set threshold of low-quality content with high severity exists in the review sequence, the document to be approved is considered to have significant flaws and needs to be rejected and revised. The rejection reason in the approval feedback report is then generated, taking into account the teaching requirements related to the business tags. This rejection reason provides a detailed explanation of the errors and inconsistencies in the document to be approved and the teacher's materials, along with relevant supporting documentation, guiding the uploader to correct the teacher's materials.

[0100] Implementing the embodiments of this application has the following beneficial effects: This application first integrates the logical relationships between teacher materials from different sources, determining the connections and order between the materials. Then, based on these logical relationships, the heterogeneous teacher materials are sorted and integrated into a sequence of attachment documents. This sequence guides the extraction of content required for a standard document template, automatically integrating heterogeneous data into a document ready for approval that meets the approval requirements, effectively improving approval efficiency. To enhance the quality of data integration, data units and fillable fields are matched to find semantically similar content for fillable fields. Furthermore, data source mapping ensures that all filled fields have data traceability capabilities, allowing approvers to directly view the original data. The application calls the teaching information management system based on business tags to check whether teacher materials are suitable for teaching and can achieve the established teaching objectives. Compared to general material approval, this provides more professional approval methods and knowledge, eliminating the need for experienced teaching staff to conduct detailed reviews of teacher materials. It can quickly and accurately identify anomalies that contradict teaching requirements or deviate significantly from them, marking these anomalies for further detailed review by teaching staff, greatly reducing the workload. Furthermore, it provides relevant justifications for any unreasonable aspects of the review process, making it easier for users to understand and ensuring that the approval results are more reliable.

[0101] Furthermore, Figure 3 This is a structural diagram of a terminal device provided in one embodiment of this application. Figure 3 As shown, the terminal device 3 of this embodiment includes: at least one processor 30 (in... Figure 3 (Only one is shown in the image) and a memory 31 and a computer program 32 stored in the memory 31 and executable on the at least one processor, wherein when the processor 30 executes the computer program 32, it can implement the steps of the online approval processing method for teacher materials as described in any one of the embodiments of this application.

[0102] The terminal device 3 may be a computing device such as a desktop computer, a cloud server, or a laptop computer, and the computing device may include, but is not limited to, a processor 30 and a memory 31. Figure 3 This is merely an example of terminal device 3 and does not constitute a limitation on terminal device 3. It may include more or fewer components than those shown in the figure.

[0103] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of this application. It should be understood that the above descriptions are merely specific embodiments of this application and are not intended to limit the scope of protection of this application. In particular, it should be noted that any modifications, equivalent substitutions, or improvements made by those skilled in the art within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for online approval and processing of teacher materials, characterized in that, include: A sequence of attachment documents is constructed by analyzing the logical relationships between teacher materials from different sources, and data units are extracted from the attachment document sequence according to a preset standard document template; wherein, the standard document template is determined by the business tags of the teacher materials; Semantic matching and data source mapping are performed between the data units and the fillable fields in the standard document template to obtain the document to be approved. According to the preset teaching requirements, the teaching information management system corresponding to the business tag is invoked to conduct a rationality assessment and anomaly identification on the document to be approved, and a first approval result is obtained; Based on the first approval result, mark the points to be reviewed in the document to be approved and add corresponding unreasonable evidence at the points to be reviewed, and construct the approval feedback report of the document to be approved.

2. The online approval and processing method for teacher materials according to claim 1, characterized in that, The process involves constructing a sequence of attachment documents by analyzing the logical relationships between teacher materials from different sources, and extracting data units from this sequence according to a preset standard document template. Specifically: To conform to the thematic order of the standard document template, the teacher materials are dimensionally aligned with the standard document template to obtain the logical relationship; Based on the logical relationships and the preset approval document directory, all the teacher materials are sorted and renamed to obtain the attachment document sequence; Based on the standard document template, the data unit is obtained by extracting data and performing structured transformation on the attached document sequence.

3. The online approval and processing method for teacher materials according to claim 1, characterized in that, The step of semantically matching and data source mapping between the data unit and the fillable fields in the standard document template to obtain the document to be approved is as follows: Based on a preset vocabulary knowledge graph, the contextual semantics of each data unit are matched with the text description of the fillable field by reasoning the meaning of the words, and the data unit that is successfully matched is taken as the unit to be filled. Based on the matching results, a dynamic index is constructed between the cell to be filled and the fillable field by mapping the data source information of the cell to be filled with the position information of the fillable field; wherein, the data source information records the teacher material to which the cell to be filled belongs and the position of the cell to be filled in the attachment document sequence; Based on the dynamic index, the cells to be filled are automatically filled into the standard document template to generate a document awaiting approval.

4. The online approval and processing method for teacher materials according to claim 3, characterized in that, The method based on a preset vocabulary knowledge graph matches the contextual semantics of each data unit with the text description of the fillable field by inferring the meaning of the words. Specifically: Based on the contextual semantics and the text description, the positions of the data unit and the fillable field in the vocabulary knowledge graph are determined respectively, and the concept node of the data unit and the meaning node of the fillable field are obtained. Calculate the path distance between the concept node and the meaning node, and obtain the matching degree between the data unit and the fillable field by comparing the magnitude of the path distance.

5. The online approval and processing method for teacher materials according to claim 1, characterized in that, According to preset teaching requirements, the teaching information management system corresponding to the business tag is invoked to perform a rationality assessment and anomaly identification on the document to be approved, and a first approval result is obtained, specifically as follows: Based on the text titles in the document to be approved, the document to be approved is segmented into several modules to be approved. Extract the review criteria related to the business tags from the teaching requirements; Based on the teaching information management system, anomalies that do not meet the review criteria in the module to be approved are detected by a preset large language model, and a first approval result is obtained.

6. The online approval and processing method for teacher materials according to claim 5, characterized in that, Based on the teaching information management system, the system detects anomalies in the module to be approved that do not meet the review criteria using a preset large language model, and obtains a first approval result, specifically as follows: Based on the aforementioned review criteria, the modules to be approved are clustered using the large language model. The teaching information management system is invoked based on the clustering results to obtain the standard text corresponding to the clustering results; Based on the priority of the review criteria and the publication time of the standard text, a weighted calculation is performed on the similarity between the module to be approved and the standard text to obtain a reasonableness score for the module to be approved. Modules to be approved with reasonableness scores lower than a first threshold are marked as anomalies. The later the publication time, the higher the corresponding weight coefficient.

7. The online approval and processing method for teacher materials according to claim 1, characterized in that, Based on the first approval result, the steps involve marking the points requiring review in the document to be approved and adding corresponding evidence of unreasonableness at these points to construct an approval feedback report for the document to be approved. Specifically: Based on the first approval result, mark the points to be reviewed in the documents pending approval; Obtain the call information of the teaching information management system, perform data trimming and keyword extraction on the call information, and obtain the unreasonable proof of the point to be reviewed; Based on the text type of the point to be reviewed, the unreasonable proof is formatted and then displayed hierarchically at the corresponding point to be reviewed, thus obtaining the approval feedback report.

8. The online approval and processing method for teacher materials according to any one of claims 1 to 7, characterized in that, The approval feedback report also includes: Based on the results of the first approval, the points pending review are ranked according to their severity. If the sorting results contain low-quality content exceeding the second threshold, then the rejection reason for the approval feedback report is generated in conjunction with the teaching requirements.

9. An online approval and processing system for teacher materials, characterized in that, include: Data extraction module, data population module, data pre-approval module, and approval report generation module; The data extraction module is used to construct a sequence of attachment documents by analyzing the logical relationships between teacher materials from different sources, and to extract data units from the sequence of attachment documents according to a preset standard document template; wherein, the standard document template is determined by the business tags of the teacher materials; The data population module is used to perform semantic matching and data source mapping between the data unit and the fillable fields in the standard document template to obtain the document to be approved; The data pre-approval module is used to call the teaching information management system corresponding to the business tag to conduct a rationality assessment and anomaly identification of the document to be approved according to the preset teaching requirements, and obtain the first approval result; The approval report generation module is used to mark the points to be reviewed in the document to be approved based on the first approval result, add corresponding unreasonable evidence at the points to be reviewed, and construct the approval feedback report of the document to be approved.

10. A terminal device, characterized in that, It includes a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement the steps of the online approval and processing method for teacher materials according to any one of claims 1 to 8.

Images