One-stop integrated management system for intelligent educational administration data

Abstract

The invention provides a one-stop integrated management system for intelligent educational administration data, and the system comprises a data architecture which outputs a data source which can be automatically optimized after constructing and optimizing an educational administration data model; the algorithm architecture is internally provided with an algorithm model and an algorithm engine, and the algorithm model generates an evolution sandbox which can be processed by the algorithm engine by utilizing a quantization evaluation technology after the algorithm model passes through the distillation data source; relates to the technical field of educational administration. According to the method, advanced technologies such as quantization evaluation, digital twinborn cognitive synchronization and federal multi-agent game are creatively introduced into the field of education management, education decision ecology with self-evolution ability is constructed, breakthrough innovation is formed in three levels of data management, algorithm architecture and system fusion, and the method has the advantages of being high in practicability and easy to popularize. The problems of decision black box, resource mismatching, algorithm rigidity and the like of a traditional system are solved, the educational resource configuration efficiency is effectively improved, and the educational administration system has the high-quality characteristic of sustainable evolution.

Description

Technical Field

[0001] The present invention relates to the technical field of educational administration management, and specifically to a one-stop integrated management system for intelligent educational administration data. Background Art

[0002] Currently, most intelligent educational administration systems are implemented through the technical path of combining traditional database architectures and rule engines. They mainly rely on Internet of Things sensing devices to collect basic teaching data and then use statistical methods for teaching quality evaluation. They generally integrate teaching information through data warehouses and form mature educational administration management systems.

[0003] However, the existing technologies have the following key defects:

[0004] 1. There is a lack of a credible governance mechanism for multi-source heterogeneous teaching data. Traditional centralized storage easily leads to data islands and it is difficult to trace the data evolution process, unable to meet the dynamic traceability requirements of teaching process data;

[0005] 2. The degree of intelligence of the algorithm system is limited. The existing evaluation models have insufficient accuracy in quantitatively analyzing the causal relationship of teaching effects, and static rule engines are difficult to adapt to the dynamic changes of educational policies;

[0006] 3. The resource allocation decision lacks global coordination. Traditional digital twin systems only achieve a simple mapping of the physical space and do not construct a cognitive optimization mechanism for virtual-real interaction, resulting in low matching efficiency of teachers - courses - classrooms;

[0007] 4. The decision interpretability and self-evolution ability are lacking. Existing systems rely on the output results of black-box models, which can neither form a visual decision logic chain nor have the autonomous evolution ability of the algorithm architecture. Summary of the Invention

[0008] The purpose of the present invention is to provide a one-stop integrated management system for intelligent educational administration data to solve the problems raised in the background art above and overcome the existing technical defects.

[0009] To solve the above technical problems, the technical solution adopted by the present invention is: A one-stop integrated management system for intelligent educational administration data, including a data architecture, which outputs an automatically optimizable data source after constructing and optimizing an educational administration data model; an algorithm architecture, which is provided with an algorithm model and an algorithm engine, and the algorithm model generates an evolution sandbox that can be processed by the algorithm engine by distilling the data source using quantization evaluation technology; and a fusion architecture, which includes a fusion twin and a fusion decision unit, and the fusion twin provides data training for the fusion decision unit and assists the fusion decision unit in generating a smart educational administration super-brain system.

[0010] As a further aspect of the present invention: A one-stop comprehensive management system for intelligent educational administration data. The data architecture is provided with a data acquisition unit, a data modeling unit, a data verification unit, and a data optimization unit. The data architecture collects structured and unstructured data streams of the teaching scenario through the data acquisition unit. The data architecture establishes a knowledge graph that can semantically associate teaching entity relationships through the data modeling unit. The data architecture verifies and extracts effective data features in the knowledge graph through the data verification unit. The data verification unit optimizes the effective data features through the data optimization unit and cyclically purifies and upgrades them.

[0011] As a further aspect of the present invention: A one-stop comprehensive management system for intelligent educational administration data. The data architecture further includes a data lake sub-architecture, and the data lake sub-architecture is provided with a timestamp mark that can verify effective data features.

[0012] As a further aspect of the present invention: A one-stop comprehensive management system for intelligent educational administration data. The algorithm model includes a hybrid intelligent framework and a multi-modal time series prediction and causal reasoning algorithm. The algorithm model constructs a quantization evaluation model of teaching quality through the hybrid intelligent framework and the multi-modal time series prediction and causal reasoning algorithm.

[0013] As a further aspect of the present invention: A one-stop comprehensive management system for intelligent educational administration data further includes a decision rule tree. The algorithm engine generates a decision rule tree that can run the algorithm model through a neural network. After the decision rule tree and the evolutionary sandbox are paired, self-optimization construction is carried out through genetic programming technology.

[0014] As a further aspect of the present invention: A one-stop comprehensive management system for intelligent educational administration data. The fusion twin generates an autonomous game network of multi-data after synchronizing, collaborating, and intelligentizing data. The fusion decision unit generates an entropy weight TOPSIS model for the fusion decision unit in combination with the Nash equilibrium matching algorithm. The entropy weight TOPSIS model is implanted into the intelligent educational administration superbrain system for use.

[0015] A one-stop comprehensive management system for intelligent educational administration data and its usage method, including

[0016] Step S1: Construct an intelligent data base with global perception to achieve trustworthy acquisition and dynamic governance of multi-modal data in the teaching scenario;

[0017] Step S11: Real-time collect students' biometric characteristics, classroom behavior heat maps, and teaching resource interaction trajectories through a distributed Internet of Things terminal network;

[0018] Step S12: Build a semantically associated educational knowledge graph using dynamic ontology modeling technology, and combine it with a blockchain-anchored data lake to ensure data traceability;

[0019] Step S13: Deploy a data entropy increase suppression algorithm to achieve self-extraction of data features and dynamic cleaning of outliers through edge computing nodes;

[0020] Step S2: Establish a cognitive enhancement algorithm system to generate interpretable teaching decision rules through a hybrid intelligent framework;

[0021] Step S21: Construct a quantized evaluation model that fuses multi-modal time series prediction and causal reasoning to quantitatively analyze the dynamic association of teaching effects;

[0022] Step S22: Build a teaching strategy game engine based on deep reinforcement learning to dynamically adjust the parameters of course difficulty and teaching rhythm;

[0023] Step S23: Compress complex algorithms into lightweight decision trees through knowledge distillation technology, and use genetic programming to achieve self-optimization of the algorithm architecture;

[0024] Step S3: Integrate and form a teaching affairs super brain system with global decision entropy reduction, and output an optimal resource allocation plan based on multi-objective dynamic game;

[0025] Step S31: Create a campus-level educational digital twin, and achieve cognitive synchronization between virtual and real spaces through a federated learning framework;

[0026] Step S32: Construct a multi-agent collaborative game network, and use the Nash equilibrium algorithm to optimize the dynamic matching of teachers, classrooms, and courses;

[0027] Step S33: Inject the experience knowledge of educational experts to form a computable cognitive module, and generate a set of decision-making plans in combination with the entropy weight TOPSIS mechanism.

[0028] Compared with the prior art, the beneficial effects of the present invention include:

[0029] This application creatively introduces frontier technologies such as quantized evaluation, cognitive synchronization of digital twins, and federated multi-agent games into the field of education management, constructs an educational decision-making ecosystem with self-evolution ability, forms breakthrough innovations in three levels of data governance, algorithm architecture, and system integration, solves problems such as decision-making black boxes, resource misallocation, and algorithm rigidity existing in traditional systems, effectively improves the efficiency of educational resource allocation, and endows the teaching affairs system with the excellent characteristic of continuous evolution. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The disclosure of the present invention will be described with reference to the accompanying drawings. It should be understood that the drawings are for illustrative purposes only and are not intended to limit the scope of protection of the present invention. In the drawings, the same reference numerals are used to refer to the same components. Among them:

[0031] Figure 1 Schematically shows the overall schematic diagram of the method flow proposed according to an embodiment of the present invention;

[0032] Figure 2 Schematically shows one of the partial schematic diagrams of the method flow proposed according to an embodiment of the present invention;

[0033] Figure 3 Schematically shows another partial schematic diagram of the method flow proposed according to an embodiment of the present invention;

[0034] Figure 4 Schematically shows a third partial schematic diagram of the method flow proposed according to an embodiment of the present invention. Detailed Embodiments

[0035] It is easily understood that according to the technical solution of the present invention, without changing the essential spirit of the present invention, those of ordinary skill in the art can propose various alternative structural forms and implementation methods. Therefore, the following detailed embodiments and the accompanying drawings are only exemplary descriptions of the technical solution of the present invention and should not be regarded as the whole of the present invention or as a limitation or restriction on the technical solution of the present invention.

[0036] According to an embodiment of the present invention, in combination with the accompanying drawings, a one-stop integrated management system for intelligent educational administration data is shown, including a data architecture. The data architecture outputs a data source that can be automatically optimized after constructing and optimizing an educational administration data model. A data acquisition unit, a data modeling unit, a data verification unit, and a data optimization unit are provided in the data architecture. The data architecture collects structured and unstructured data streams of teaching scenarios through the data acquisition unit, establishes a knowledge graph that can semantically associate teaching entity relationships through the data modeling unit, verifies and extracts valid data features in the knowledge graph through the data verification unit, and the data verification unit optimizes the valid data features through the data optimization unit and cyclically purifies and upgrades them. The data architecture further includes a data lake sub-architecture, and a timestamp mark for verifying valid data features is provided in the data lake sub-architecture. Therefore, in this embodiment, an intelligent data base with global perception is constructed, and structured and unstructured data streams of teaching scenarios are collected in real time through distributed IoT terminals, including three-dimensional data spaces such as student biometric data streams, classroom behavior heat maps, and teaching resource interaction trajectories. The dynamic ontology modeling technology is used to construct an educational knowledge graph to achieve semantic-level association of teaching entity relationships, and data self-purification and feature self-extraction are completed through edge computing nodes. A blockchain-anchored data lake architecture is specially designed to enable the teaching process data to have a verifiable timestamp mark, and a data entropy increase suppression algorithm is introduced to ensure the self-evolution ability of the data ecosystem.

[0037] Other embodiments of the present invention: It further includes an algorithm architecture, and an algorithm model and an algorithm engine are provided in the algorithm architecture. The algorithm model generates an evolving sandbox that can be processed by the algorithm engine by distilling the data source using quantization evaluation technology. The algorithm model includes a hybrid intelligent framework and a multi-modal time series prediction and causal inference algorithm. The algorithm model constructs a quantization evaluation model of teaching quality through the hybrid intelligent framework and the multi-modal time series prediction and causal inference algorithm. It further includes a decision rule tree. The algorithm engine generates a decision rule tree that can run the algorithm model through a neural network. After the decision rule tree and the evolving sandbox are paired, self-optimization construction is carried out through genetic programming technology. Therefore, in this embodiment, an innovative teaching cognition-enhanced hybrid intelligent framework is proposed, which integrates a multi-modal time series prediction and causal inference algorithm to construct a quantization evaluation model of teaching quality. A dynamic teaching strategy engine is constructed using deep reinforcement learning, and the adaptive adjustment of teaching parameters is realized through the dynamic game between course difficulty and student performance. A lightweight inference model cluster based on knowledge distillation is developed, which can compress complex teaching analysis algorithms into interpretable decision rule trees to achieve the Pareto optimum of computing power requirements and decision accuracy. The system has a built-in algorithm evolving sandbox to support the self-optimization of the algorithm architecture based on genetic programming to ensure that the model continuously adapts to changes in education policies.

[0038] Other embodiments of the present invention: It further includes a fusion architecture, which includes a fusion twin and a fusion decision-making unit. The fusion twin provides data training for the fusion decision-making unit and assists the fusion decision-making unit in generating an intelligent educational administration super brain system. The fusion twin generates an autonomous game network of multi-data after synchronization, collaboration, and intelligent data. The fusion decision-making unit generates an entropy weight TOPSIS model for the fusion decision-making unit by combining the Nash equilibrium matching algorithm. The entropy weight TOPSIS model is implanted into the intelligent educational administration super brain system for use. Therefore, this embodiment pioneers an educational digital twin fusion engine, constructs a campus-level teaching twin through virtual-real mapping technology, and realizes cognitive synchronization between the physical space and the digital space. Adopting a multi-agent collaboration mechanism under the federated learning framework, enables subsystems such as curriculum arrangement, resource allocation, and quality assessment to form an autonomous game collaboration network. Develop a teaching element matching algorithm based on multi-objective optimization to dynamically optimize the Nash equilibrium solution in scenarios such as teacher scheduling, classroom allocation, and curriculum arrangement. Through knowledge injection-enhanced analysis technology, encode the experience knowledge of education experts into computable cognitive modules, and finally form an intelligent educational administration super brain system with the ability to reduce global decision entropy. Its unique teaching resource allocation decision tree supports a multi-dimensional scheme optimization mechanism based on entropy weight TOPSIS.

[0039] In summary, this application creatively introduces frontier technologies such as quantization evaluation, digital twin cognitive synchronization, and federated multi-agent game into the field of education management, constructs an educational decision-making ecosystem with self-evolution ability, forms breakthrough innovations at the three levels of data governance, algorithm architecture, and system integration, solves problems such as decision-making black boxes, resource misallocation, and algorithm rigidity existing in traditional systems, effectively improves the efficiency of educational resource allocation, and endows the educational administration system with the excellent characteristic of continuous evolution.

[0040] A one-stop comprehensive management system for intelligent educational administration data and its usage method, including

[0041] Step S1: Construct an intelligent data base with global perception to achieve reliable collection and dynamic governance of multi-modal data in teaching scenarios. Specifically, first, collect students' biometric features, classroom behavior heat maps, and teaching resource interaction trajectories in real time through a distributed Internet of Things terminal network, then construct an educational knowledge graph with semantic associations using dynamic ontology modeling technology, combine with a blockchain-anchored data lake to ensure data traceability, and finally deploy a data entropy increase suppression algorithm to realize self-extraction of data features and dynamic cleaning of outliers through edge computing nodes;

[0042] Step S2: Establish a cognitive enhancement algorithm system, and generate interpretable teaching decision rules through a hybrid intelligent framework. Specifically, first construct a quantized evaluation model that fuses multi-modal time series prediction and causal reasoning to quantitatively analyze the dynamic correlation of teaching effects. Then, based on deep reinforcement learning, construct a teaching strategy game engine to dynamically adjust the parameters of course difficulty and teaching rhythm. Finally, use knowledge distillation technology to compress complex algorithms into lightweight decision trees, and use genetic programming to achieve self-optimization of the algorithm architecture;

[0043] Step S3: Integrate and form an educational administration super-brain system with global decision entropy reduction, and output an optimal resource allocation plan based on multi-objective dynamic game. Specifically, first create a campus-level educational digital twin, and achieve cognitive synchronization between virtual and real spaces through a federated learning framework. Then, construct a multi-agent collaborative game network, and use the Nash equilibrium algorithm to optimize the dynamic matching of teachers, classrooms, and courses. Finally, inject the experience and knowledge of educational experts to form a computable cognitive module, and generate a set of decision-making plans in combination with the entropy weight TOPSIS mechanism.

[0044] Working principle: This application creatively introduces frontier technologies such as quantized evaluation, digital twin cognitive synchronization, and federated multi-agent game into the field of education management, constructs an educational decision-making ecosystem with self-evolution ability, and achieves breakthrough innovations in three levels: data governance, algorithm architecture, and system integration, solving problems such as decision-making black boxes, resource misallocation, and algorithm rigidity existing in traditional systems, effectively improving the efficiency of educational resource allocation, and endowing the educational administration system with the excellent characteristic of continuous evolution.

[0045] The technical scope of the present invention is not limited to the content described above. Those skilled in the art can make various deformations and modifications to the above embodiments without departing from the technical idea of the present invention, and these deformations and modifications should all fall within the protection scope of the present invention.

Claims

1. A one-stop comprehensive management system for intelligent educational administration data, characterized in that, It includes a data architecture which outputs an automatically optimizable data source after constructing and optimizing an educational administration data model; An algorithm architecture which has an algorithm model and an algorithm engine inside. The algorithm model generates an evolving sandbox that can be processed by the algorithm engine after distilling the data source and using quantization evaluation technology; And A fusion architecture which includes a fusion twin and a fusion decision unit. The fusion twin provides data for the fusion decision unit to train and assists the fusion decision unit in generating an intelligent educational administration super brain system.

2. The one-stop integrated management system for intelligent educational administration data according to claim 1, wherein The data architecture has a data collection unit, a data modeling unit, a data verification unit, and a data optimization unit. The data architecture collects structured and unstructured data streams of teaching scenarios through the data collection unit. The data architecture builds a knowledge graph that can semantically associate teaching entity relationships through the data modeling unit. The data architecture verifies and extracts valid data features in the knowledge graph through the data verification unit. The data verification unit optimizes the valid data features through the data optimization unit and cyclically purifies and upgrades them.

3. The one-stop integrated management system for intelligent educational administration data according to claim 2, characterized in that, The data architecture also includes a data lake sub-architecture which has a timestamp mark for verifying valid data features.

4. The one-stop integrated management system for intelligent educational administration data according to claim 3, characterized in that, The algorithm model includes a hybrid intelligent framework and a multi-modal time series prediction and causal inference algorithm. The algorithm model constructs a quantization evaluation model of teaching quality through the hybrid intelligent framework and the multi-modal time series prediction and causal inference algorithm.

5. The one-stop integrated management system for intelligent educational administration data according to claim 4, characterized in that, It also includes a decision rule tree. The algorithm engine generates a decision rule tree that can run the algorithm model through a neural network. After the decision rule tree and the evolving sandbox are paired, they are self-optimized and constructed through genetic programming technology.

6. The one-stop integrated management system for intelligent educational administration data according to claim 5, characterized in that, The fusion twin generates an autonomous game network of multi-data after synchronizing, collaborating, and intelligentizing data. The fusion decision unit generates an entropy weight TOPSIS model for the fusion decision unit to use in combination with the Nash equilibrium matching algorithm. The entropy weight TOPSIS model is implanted into the intelligent educational administration super brain system for use.

7. A one-stop integrated management system for intelligent educational administration data and its usage method according to claim 6, characterized in that, It includes Step S1: Build an intelligent data base for global perception to achieve trustworthy collection and dynamic governance of multi-modal data in teaching scenarios; Step S2: Establish a cognitive enhanced algorithm system to generate interpretable teaching decision rules through a hybrid intelligent framework; Step S3: Integrate and form an educational administration super brain system with global decision entropy reduction, and output an optimal resource allocation plan based on multi-objective dynamic game.

8. A one-stop integrated management system and its usage method for intelligent educational administration data according to claim 7, characterized in that, The said Step S1 includes Step S11: Real-time collect students' biometric features, classroom behavior heat maps, and teaching resource interaction trajectories through a distributed Internet of Things terminal network; Step S12: Use dynamic ontology modeling technology to construct a semantically associated educational knowledge graph, and combine a blockchain-anchored data lake to ensure data traceability; Step S13: Deploy a data entropy increase suppression algorithm to achieve self-extraction of data features and dynamic cleaning of outliers through edge computing nodes.

9. A one-stop integrated management system and its usage method for intelligent educational administration data according to claim 8, characterized in that, The said Step S2 includes Step S21: Construct a quantization evaluation model integrating multi-modal time series prediction and causal inference to quantitatively analyze the dynamic association of teaching effects; Step S22: Build a teaching strategy game engine based on deep reinforcement learning to dynamically adjust the parameters of course difficulty and teaching rhythm; Step S23: Compress the complex algorithm into a lightweight decision tree through knowledge distillation technology, and use genetic programming to achieve self-optimization of the algorithm architecture.

10. A one-stop integrated management system and its usage method for intelligent educational administration data according to claim 9, characterized in that, The said step S3 includes Step S31: Create a campus-level educational digital twin, and achieve cognitive synchronization between the virtual and real spaces through the federated learning framework; Step S32: Construct a multi-agent collaborative game network, and use the Nash equilibrium algorithm to optimize the dynamic matching of teachers, classrooms, and courses; Step S33: Inject the expert experience knowledge of education to form a computable cognitive module, and generate a decision solution set in combination with the entropy weight TOPSIS mechanism.

Images