Digital education resource matching method

Abstract

The invention relates to the technical field of data resource matching, in particular to a digital education resource matching method. The method comprises the following steps: acquiring educational resource data; performing data preprocessing on the educational resource data to generate an educational resource input data set; obtaining digital education content data; performing multi-dimensional portrait drawing on the educational resource input data set by using a preset model to generate digital education multi-dimensional portrait data; performing content feature matching on the digital education content data by using the digital education multi-dimensional portrait data to generate education resource content matching data; therefore, according to the digital education resource matching method based on multi-dimensional feature analysis, context awareness and dynamic weight allocation, the problems that a traditional matching mode is poor in context adaptability and low in resource allocation efficiency are solved, and the dynamic adaptability and matching precision of the digital education resources are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data resource matching, and in particular to a digital education resource matching method. Background Art

[0002] Traditional systems often rely on static data processing methods and cannot adapt to changes in user needs in real time, which makes the system insufficient when dealing with dynamic needs. Traditional matching algorithms are usually based on simple similarity calculations, such as recommendations based on keywords or user ratings. These methods often ignore the multi-dimensional characteristics of educational content and user needs, and therefore cannot fully consider the personalized needs of users and changes in the learning environment. Furthermore, traditional systems usually lack contextual awareness and fail to fully consider the specific context of the user (such as device type, usage environment, etc.), which makes it difficult for the system to provide effective resource recommendations when facing complex contextual changes. In addition, many traditional systems lack dynamic adjustment mechanisms when matching educational resources, and cannot be dynamically adjusted according to real-time data changes and user feedback, resulting in recommendation results that may not meet actual needs. Finally, traditional technologies often do not work well when processing multi-dimensional data, and it is difficult to integrate user behavior data, learning goals, environmental factors, etc., resulting in limited resource recommendation results. Summary of the invention

[0003] Based on this, it is necessary to provide a digital educational resource matching method to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a digital educational resource matching method is provided, the method comprising the following steps:

[0005] Step S1: obtaining educational resource data; performing data preprocessing on the educational resource data to generate an educational resource input data set;

[0006] Step S2: Acquire digital education content data; Use a preset model to draw a multi-dimensional portrait of the education resource input data set to generate digital education multi-dimensional portrait data; Use the digital education multi-dimensional portrait data to match the content features of the digital education content data to generate education resource content matching data;

[0007] Step S3: acquiring level meter data and height sensor data; performing scenario analysis on the level meter data and the height sensor data to generate scenario perception data; performing scenario perception scheme generation on the scenario perception data to generate a scenario perception scheme;

[0008] Step S4: Perform multi-dimensional dynamic matching of educational resource content matching data and scenario perception solutions to generate digital educational resource dynamic matching data, thereby completing digital educational resource matching.

[0009] The beneficial effect of the present invention is that through a multi-dimensional digital education resource matching method, efficient and accurate dynamic adaptation of educational resources and learning needs is achieved from the data level. Using educational resource input data sets and digital education content data, digital education multi-dimensional portrait data is constructed through multi-dimensional portrait drawing and feature priority sorting and other technologies. This process can not only capture the complex characteristics of educational resources in high-dimensional space, but also ensure the accuracy of resource matching in different situations through feature weight allocation and priority sorting. Through the generation of scenario perception schemes, combined with the data of level meters and height sensor meters, the scene perception ability is further enhanced. Through steps such as nonlinear mapping and data source combination, accurate judgment of scenario status is achieved, ensuring the pertinence of learning resource recommendations in different situations. At the same time, the system conducts multi-dimensional dynamic weight adjustment and weighted allocation through dynamic adaptability evaluation, combined with scenario perception schemes and educational resource content matching data, and effectively optimizes the adaptability of educational resources. By deeply analyzing the supply and demand relationship of educational resources, using supply and demand relationship data graphs and similarity calculations, the system can adjust the resource recommendation scheme in a timely manner when learning needs change, ensuring that learners obtain the most suitable resources. The application of technologies such as sparse coding and embedded vectors not only simplifies the complexity of multi-dimensional portrait data, but also ensures the real-time and accuracy of resource recommendations. Therefore, the present invention solves the problems of poor situational adaptability and low resource allocation efficiency of traditional matching methods through a digital education resource matching method based on multi-dimensional feature analysis, situational perception and dynamic weight allocation, and improves the dynamic adaptability and matching accuracy of digital education resources.

[0010] Preferably, step S1 comprises the following steps:

[0011] Step S11: Acquire educational resource data;

[0012] Step S12: performing data cleaning on the educational resource data to generate educational resource cleansing data;

[0013] Step S13: performing format standardization processing on the education resource cleansing data to generate digital education resource standardized data;

[0014] Step S14: labeling the input data set of the digital educational resources standardization to generate an educational resource input data set.

[0015] The present invention obtains the original educational resource data, which provides a rich resource input for the system. The original data has problems such as noise, redundancy and missing, and a data cleaning method is introduced to generate high-quality educational resource cleaned data by filtering, correcting and supplementing incomplete data. This step not only removes the noise data that affects the accuracy of the data, but also ensures the integrity and effectiveness of the resource content. Through format standardization processing, the format of the educational resource data is unified, the format inconsistency problem caused by different data sources or input methods is solved, and digital educational resource standardized data is generated. The core value of this step is to ensure the uniformity of the data structure, which is convenient for feature extraction and matching operations in subsequent processing steps. Through the annotation operation of standardized data, the attributes and classification of each data set are further clarified, and an accurate educational resource input data set is generated. This annotation process not only improves the operability of the data, but also enhances the matching efficiency of resources and learning needs.

[0016] Preferably, step S2 comprises the following steps:

[0017] Step S21: Acquire digital education content data and learning demand data;

[0018] Step S22: Use a preset model to draw a multi-dimensional portrait of the educational resource input data set to generate digital education multi-dimensional portrait data;

[0019] Step S23: Use the digital education multi-dimensional portrait data to perform two-dimensional graph data mapping on the learning demand data to generate a digital education supply and demand relationship data graph; use the digital education supply and demand relationship data graph to perform similarity calculation on the digital education content data to generate digital education similarity data; perform content feature matching on the digital education similarity data to generate educational resource content matching data.

[0020] The present invention obtains digital education content data and learning demand data, ensures the extensiveness of educational resources and the diversity of learning needs, and provides a rich source of input data for subsequent matching. The preset model is used to draw a multi-dimensional portrait of the educational resource input data set, and digital education multi-dimensional portrait data is generated. This step systematically integrates the content, form, difficulty and other information of the resources by comprehensively analyzing the various dimensional characteristics of the educational resources, constructs a high-dimensional feature portrait, and provides a more accurate data basis for personalized recommendation. The digital education multi-dimensional portrait data is further used to map the learning demand data to a two-dimensional graph data, and a digital education supply and demand relationship data graph is generated. This mapping operation simplifies complex multi-dimensional information into a visual two-dimensional graph, which not only improves the intuitiveness of data processing, but also enhances the clarity of the supply and demand relationship. By calculating the similarity of the supply and demand relationship graph, digital education similarity data is generated, so that the system can efficiently quantify the matching degree between resources and learning needs, thereby ensuring the accuracy of the recommended content. Finally, the similarity data is used to match the content features, and the final educational resource content matching data is generated, realizing the accurate docking of resources and needs.

[0021] Preferably, step S22 includes the following steps:

[0022] Step S221: performing high-dimensional mapping on the educational resource input data set through high-dimensional feature space mapping to generate a high-dimensional feature data set;

[0023] Step S222: performing dynamic weight allocation and quantification processing on the high-dimensional feature data set to generate digital education feature weight data; using a feature priority sorting algorithm to sort the importance of the digital education feature weight data to generate a feature priority sorting table;

[0024] Step S223: Use the preset model to draw a multi-dimensional portrait of the educational resource input data set and the feature priority ranking table, and perform sparse processing through the sparse coding method to generate digital education multi-dimensional portrait data.

[0025] The present invention performs high-dimensional mapping of the educational resource input data set through high-dimensional feature space mapping to generate a high-dimensional feature data set. This operation enables the complex features in the original educational resource data to be mapped to the high-dimensional space, thereby retaining more information dimensions and ensuring the comprehensiveness and diversity of resource features. The high-dimensional feature data set is subjected to dynamic weight allocation quantification processing to generate digital education feature weight data. This process gives each feature a flexible weight allocation method through quantification processing, effectively reflects the importance of different features, and enhances the dynamic adaptability of the model. At the same time, these weight data are sorted using a feature priority sorting algorithm to generate a feature priority sorting table, which clearly arranges the importance of each feature and optimizes the feature utilization efficiency in subsequent processing steps. The educational resource input data set and the feature priority sorting table are multi-dimensionally portrayed using a preset model, and the multi-dimensional portrait is sparsely processed by a sparse coding method to generate digital education multi-dimensional portrait data. Through sparse processing, not only the interference of redundant information is reduced, but also the prominence of key features is improved, making the portrait data more compact and representative.

[0026] Preferably, step S23 includes the following steps:

[0027] Step S231: using the embedding vector method to perform high-dimensional deconstruction on the digital education multi-dimensional portrait data to generate digital education portrait deconstruction data; using the digital education portrait deconstruction data to perform two-dimensional graph data mapping on the learning demand data to generate a digital education supply and demand relationship data graph;

[0028] Step S232: optimizing the relationship topology of the digital education supply and demand relationship data graph to generate a digital education supply and demand topology graph; performing similarity calculation on the digital education content data using the digital education supply and demand relationship data graph to generate digital education similarity data;

[0029] Step S233: Perform content feature matching on the digital education similarity data to generate education resource content matching data.

[0030] The present invention uses the embedding vector method to perform high-dimensional deconstruction on the multi-dimensional portrait data of digital education, and generates digital education portrait deconstruction data. This process separates the key features of the multi-dimensional portrait data by high-dimensional deconstruction, so that the core features of educational resources can correspond to learning needs more intuitively. The learning demand data is mapped to a two-dimensional graph data using these deconstructed data to generate a digital education supply and demand relationship data graph. This mapping process enables the complex supply and demand relationship to be visualized, and enhances the system's understanding of the matching relationship between learning needs and educational resources. The digital education supply and demand relationship data graph is optimized by the relationship topology graph, and a digital education supply and demand topology graph is generated. This optimization further improves the matching efficiency of resources and needs by adjusting the structure of nodes and edges in the graph, reduces invalid matching paths, and enhances the connectivity and adaptability of the supply and demand relationship. The digital education content data is similarity calculated through the supply and demand relationship data graph, and digital education similarity data is generated. This calculation accurately quantifies the matching degree between educational content and learning needs, thereby ensuring the accuracy of the subsequent matching process. Finally, the similarity data was matched with content features to generate educational resource content matching data, achieving a closed-loop optimization from similarity calculation to content matching.

[0031] Preferably, step S3 comprises the following steps:

[0032] Step S31: Acquire data of the level meter and data of the height sensor;

[0033] Step S32: Performing level height information enhancement processing on the level meter data and the height sensor data to generate level-height enhanced data; performing scenario analysis on the level-height enhanced data to generate a scenario state determination result;

[0034] Step S33: generating a situation awareness solution based on the situation state determination result.

[0035] The present invention obtains the user's position information and somatosensory data in the learning process by collecting data from a level meter and a height sensor, thereby ensuring the accuracy and integrity of the data. By performing horizontal height information enhancement processing on the level meter and height sensor data, horizontal-height enhanced data is generated. This step improves the system's perception of the user's position and posture by performing nonlinear enhancement on the data, so that the system can not only identify static learning states, but also capture dynamic learning behaviors. In addition, combined with scenario analysis, the system generates scenario state determination results based on the enhanced data. This process converts complex user location information into actionable scenario states, such as mobile handheld states and desk stable states, by introducing advanced scenario analysis algorithms, thereby providing a reliable decision-making basis for subsequent scenario perception. Finally, based on the scenario state determination results, a scenario perception solution is generated. The generation of this solution fully considers the needs of users in different learning scenarios. For example, in a mobile handheld state, the system tends to recommend resources suitable for fragmented learning, while in a desk stable state, the system will give priority to recommending content suitable for deep learning. Through this intelligent perception solution, the system can dynamically adjust the allocation strategy of educational resources to ensure the matching degree of learning resources and the optimization of user experience.

[0036] Preferably, step S32 includes the following steps:

[0037] Step S321: performing nonlinear level height information enhancement mapping on the level meter data and the height sensor data to generate level-height enhancement data; performing stable interval identification on the level-height enhancement data to generate level-height stable period data;

[0038] Step S322: combining the data sources of the horizontal-altitude enhancement data and the horizontal-altitude stable period data to generate horizontal-altitude combined source data;

[0039] Step S323: Use the preset horizontal-altitude threshold to perform scenario state determination on the horizontal-altitude joint source data to generate a scenario state determination result, wherein the scenario state determination result includes a moving handheld result and a desk stable state result.

[0040] The present invention generates level-height enhanced data by mapping the level meter data and the height sensor data through nonlinear level height information enhancement. The use of nonlinear mapping can fully explore the complex dynamic change characteristics in the data, ensure that the system can identify the user's slight changes in different postures and spatial positions, and greatly improve the sensitivity to the user's motion state. The enhanced data is processed by stable interval recognition to generate level-height stable period data. This step effectively distinguishes the difference between dynamic changes and stable states, and provides a more accurate time interval for subsequent judgment. By combining the data sources of the level-height enhanced data and the level-height stable period data, the level-height joint source data is generated. The core of this step is to integrate the information of multi-source data, and improve the integrity and robustness of the data through joint processing. The combination of multi-source data not only enhances the system's multi-dimensional perception of the user's current state, but also further optimizes the system's judgment accuracy for the transition between different states by introducing stable period data, ensuring that the system can distinguish the specific needs of users in different scenarios. The joint source data is judged by the preset level-height threshold to generate scene state judgment results such as mobile handheld results and desk stable state results. This judgment process sets reasonable threshold standards, allowing the system to accurately classify the user based on the user's current posture and environment.

[0041] Preferably, step S33 includes the following steps:

[0042] Step S331: generating a situation awareness solution based on the situation state determination result, wherein the situation awareness solution includes a mobile handheld solution and a desktop stable learning solution;

[0043] Step S332: When the result of the scenario status determination is a mobile handheld result, the digital education content data is short-term adaptively filtered to generate fragmented learning resource data; the learning demand data is short-term prioritized to generate short-term priority data; the fragmented learning resource data and the short-term priority data are cross-resource validated using the digital education multi-dimensional portrait data to generate mobile handheld cross-validation data; the mobile handheld cross-validation data is mobile adaptively filtered to generate a mobile handheld recommended resource set, thereby obtaining a mobile handheld solution;

[0044] Step S333: When the scenario state judgment result is a desk stable state result, deep learning feature analysis is performed on the digital education multi-dimensional portrait data to generate deep learning feature data; learning demand matching is performed on the deep learning feature data and the digital education supply and demand relationship data graph to generate deep learning matching data; multi-dimensional learning content recommendation is performed on the deep learning matching data using the digital education supply and demand topology graph to generate multi-dimensional learning content recommendation data; desktop device resource recommendation is performed on the multi-dimensional learning content recommendation data using the digital education portrait deconstruction data to generate a desktop device recommended resource set, thereby obtaining a desktop stable learning plan.

[0045] The present invention is divided into a mobile handheld solution and a desktop stable learning solution based on the situation state determination result through a situation perception solution. Through this segmentation, the system can distinguish different learning scenarios, and then generate corresponding resource allocation strategies according to the different needs of users in dynamic mobile or static desktop learning, so as to meet the learning needs of users in different states. When the system determines that the user is in a mobile handheld state, the digital education content data is firstly short-term adaptively filtered to generate fragmented learning resource data. This processing method ensures that the system can provide short and easy-to-digest learning content in a mobile state, avoiding the cognitive burden of lengthy information on users. At the same time, the learning demand data is sorted in a short time priority, and short-term priority data is generated, so that the system can give priority to meeting the user's learning goals in a short time. Then, the fragmented learning resource data and the short-term priority data are cross-resource verified using the multi-dimensional portrait data of digital education, further ensuring the high matching degree between the learning resources and the short-term needs of users. By mobile adaptive filtering of the mobile handheld cross-validation data, a mobile handheld recommended resource set is finally generated, which realizes the efficient scheduling of learning resources, ensures that users can obtain fragmented learning resources suitable for the current state when learning on the move, and completes the mobile handheld solution. When the user is in a stable state at the desk, the system adopts a strategy of deep learning feature analysis and resource matching. By performing deep learning feature analysis on the multi-dimensional portrait data of digital education and generating deep learning feature data, the system can dig out the deep learning needs of users in a stable learning state, and match the learning needs according to the digital education supply and demand relationship data graph, generate deep learning matching data, and ensure that the learning content provided is highly consistent with user needs. At the same time, the system uses the digital education supply and demand topology graph to recommend multi-dimensional learning content for the deep learning matching data, and generates multi-dimensional learning content recommendation data, so that users can perform more in-depth and complex learning tasks in a stable state. In addition, by using the digital education portrait deconstruction data to recommend desktop device resources for the multi-dimensional learning content recommendation data, the system can recommend the best desktop device usage plan for users, thereby further improving the learning experience and efficiency, and ultimately completing the desktop stable learning plan.

[0046] Preferably, step S4 comprises the following steps:

[0047] Step S41: Perform dynamic adaptability evaluation using digital education multi-dimensional portrait data and scenario perception scheme to generate digital education resource adaptability evaluation data; perform weighted allocation again on the digital education resource adaptability evaluation data and education resource content matching data to generate dynamic weight matching data;

[0048] Step S42: Perform multi-dimensional dynamic matching on the dynamic weight matching data to generate digital education resource dynamic matching data, thereby completing digital education resource matching.

[0049] The present invention uses digital education multi-dimensional portrait data and situational awareness scheme to perform dynamic adaptability evaluation and generate digital education resource adaptability evaluation data. This process dynamically evaluates the adaptability of existing resources by combining the multi-dimensional portrait of the user with the needs in different learning situations, thereby ensuring the efficient application of educational content in different situations. This evaluation method can quickly adjust the resource allocation strategy according to the real-time changing learning scenarios and user needs to avoid the mismatch problem between resources and needs. , Dynamic weight matching data is generated by re-weighting the adaptability evaluation data of digital education resources and the matching data of educational resource content. The weighted allocation process ensures the accuracy of resource allocation by introducing a dynamic weight mechanism. Specifically, the system dynamically adjusts the weight of each educational resource according to the current learning status and personalized needs of the user, so that resources with high relevance and high priority obtain greater weights. This not only ensures the reasonable allocation of resources, but also makes the learning content closer to the user's immediate learning needs, thereby enhancing the learning effect. In addition, dynamic weighting can also adjust the priority and weight distribution of resources in real time according to the user's learning progress and state changes, ensuring the flexibility and adaptability of resource allocation. The system performs multi-dimensional dynamic matching of dynamic weight matching data to generate dynamic matching data for digital educational resources, thereby completing the matching of digital educational resources. This process uses a multi-dimensional dynamic matching algorithm, combined with user portraits, learning needs, and situational status in different dimensions, to accurately allocate educational resources that are suitable for current needs. Through this multi-dimensional matching mechanism, the system can quickly find the most suitable content when resources are abundant, and dynamically adjust according to feedback from users during the learning process to ensure a high degree of adaptability between learning resources and user needs. This dynamic and adaptive matching of educational resources not only improves the personalization of learning content, but also makes the learning experience smoother, greatly enhancing the flexibility and intelligence of the system.

[0050] Preferably, step S41 includes the following steps:

[0051] Step S411: Using the digital education multi-dimensional portrait data and the situational awareness solution to perform situational factor adaptability scoring, and generate digital education resource adaptability scoring data;

[0052] Step S412: Perform fuzzy rule adjustment on the digital education resource adaptability scoring data to generate digital education resource dynamic weight adjustment data;

[0053] Step S413: Use a matrix decomposition algorithm to perform multi-dimensional dynamic matching on the dynamic weight adjustment data of digital educational resources to generate dynamic matching data of digital educational resources, thereby completing the matching of digital educational resources.

[0054] The present invention generates digital education resource adaptability scoring data by using digital education multi-dimensional portrait data and scenario perception scheme to perform scenario factor adaptability scoring. This process quantitatively evaluates the educational needs and resource characteristics under different scenarios, so that each resource can be adapted according to specific scenario factors (such as user learning style, learning environment, etc.), thereby reflecting the degree of adaptation of the resource in practical applications. Through this scoring mechanism, it is possible to accurately evaluate which resources are more suitable for a specific learning scenario, thereby providing a data basis for subsequent resource optimization. The digital education resource adaptability scoring data is adjusted by fuzzy rules to generate digital education resource dynamic weight adjustment data. The fuzzy rule adjustment enables the scoring data to be reasonably adjusted in consideration of uncertainty and subjectivity in this process, and the generated dynamic weight adjustment data can better reflect the actual applicability of educational resources in different scenarios. This dynamic adjustment mechanism has significant advantages in coping with the complex variability in practical applications, and helps to improve the accuracy and applicability of matching results. The digital education resource dynamic weight adjustment data is dynamically matched in multiple dimensions using a matrix decomposition algorithm to generate digital education resource dynamic matching data. This algorithm achieves accurate resource matching by converting dynamic weight adjustment data into weight matrices in multiple dimensions and decomposing them. The matrix decomposition algorithm can effectively mine the potential structure and patterns in the data and provide more accurate resource matching results. This method not only improves the matching accuracy, but also maintains high computational efficiency when processing large-scale data.

[0055] The beneficial effect of the present invention is that through a multi-dimensional digital education resource matching method, efficient and accurate dynamic adaptation of educational resources and learning needs is achieved from the data level. Using educational resource input data sets and digital education content data, digital education multi-dimensional portrait data is constructed through multi-dimensional portrait drawing and feature priority sorting and other technologies. This process can not only capture the complex characteristics of educational resources in high-dimensional space, but also ensure the accuracy of resource matching in different situations through feature weight allocation and priority sorting. Through the generation of scenario perception schemes, combined with the data of level meters and height sensor meters, the scene perception ability is further enhanced. Through steps such as nonlinear mapping and data source combination, accurate judgment of scenario status is achieved, ensuring the pertinence of learning resource recommendations in different situations. At the same time, the system conducts multi-dimensional dynamic weight adjustment and weighted allocation through dynamic adaptability evaluation, combined with scenario perception schemes and educational resource content matching data, and effectively optimizes the adaptability of educational resources. By deeply analyzing the supply and demand relationship of educational resources, using supply and demand relationship data graphs and similarity calculations, the system can adjust the resource recommendation scheme in a timely manner when learning needs change, ensuring that learners obtain the most suitable resources. The application of technologies such as sparse coding and embedded vectors not only simplifies the complexity of multi-dimensional portrait data, but also ensures the real-time and accuracy of resource recommendations. Therefore, the present invention solves the problems of poor situational adaptability and low resource allocation efficiency of traditional matching methods through a digital education resource matching method based on multi-dimensional feature analysis, situational perception and dynamic weight allocation, and improves the dynamic adaptability and matching accuracy of digital education resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A flowchart of the steps of a digital educational resource matching method is provided;

[0057] Figure 2 for Figure 1 Detailed implementation steps of step S2 in the flowchart;

[0058] Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG.

[0059] Figure 4 for Figure 1 Detailed implementation steps of step S4 in FIG.

[0060] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0061] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0062] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0063] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0064] To achieve this, please refer to Figures 1 to 4 , a digital educational resource matching method, the method comprising the following steps:

[0065] Step S1: obtaining educational resource data; performing data preprocessing on the educational resource data to generate an educational resource input data set;

[0066] Step S2: Acquire digital education content data; Use a preset model to draw a multi-dimensional portrait of the education resource input data set to generate digital education multi-dimensional portrait data; Use the digital education multi-dimensional portrait data to match the content features of the digital education content data to generate education resource content matching data;

[0067] Step S3: acquiring level meter data and height sensor data; performing scenario analysis on the level meter data and the height sensor data to generate scenario perception data; performing scenario perception scheme generation on the scenario perception data to generate a scenario perception scheme;

[0068] Step S4: Perform multi-dimensional dynamic matching of educational resource content matching data and scenario perception solutions to generate digital educational resource dynamic matching data, thereby completing digital educational resource matching.

[0069] The beneficial effect of the present invention is that through a multi-dimensional digital education resource matching method, efficient and accurate dynamic adaptation of educational resources and learning needs is achieved from the data level. Using educational resource input data sets and digital education content data, digital education multi-dimensional portrait data is constructed through multi-dimensional portrait drawing and feature priority sorting and other technologies. This process can not only capture the complex characteristics of educational resources in high-dimensional space, but also ensure the accuracy of resource matching in different situations through feature weight allocation and priority sorting. Through the generation of scenario perception schemes, combined with the data of level meters and height sensor meters, the scene perception ability is further enhanced. Through steps such as nonlinear mapping and data source combination, accurate judgment of scenario status is achieved, ensuring the pertinence of learning resource recommendations in different situations. At the same time, the system conducts multi-dimensional dynamic weight adjustment and weighted allocation through dynamic adaptability evaluation, combined with scenario perception schemes and educational resource content matching data, and effectively optimizes the adaptability of educational resources. By deeply analyzing the supply and demand relationship of educational resources, using supply and demand relationship data graphs and similarity calculations, the system can adjust the resource recommendation scheme in a timely manner when learning needs change, ensuring that learners obtain the most suitable resources. The application of technologies such as sparse coding and embedded vectors not only simplifies the complexity of multi-dimensional portrait data, but also ensures the real-time and accuracy of resource recommendations. Therefore, the present invention solves the problems of poor situational adaptability and low resource allocation efficiency of traditional matching methods through a digital education resource matching method based on multi-dimensional feature analysis, situational perception and dynamic weight allocation, and improves the dynamic adaptability and matching accuracy of digital education resources.

[0070] In the embodiment of the present invention, reference Figure 1 The above is a schematic diagram of the steps of a digital education resource matching method of the present invention. In this example, the digital education resource matching method includes the following steps:

[0071] Step S1: obtaining educational resource data; performing data preprocessing on the educational resource data to generate an educational resource input data set;

[0072] In the embodiment of the present invention, by acquiring educational resource data, preliminary data collection is performed to lay the foundation for subsequent analysis. The data acquisition stage involves the comprehensive collection of educational resources, including but not limited to educational materials in various forms such as teaching materials, teaching videos, online courses, interactive exercises, etc. The data at this stage comes from multiple sources, such as educational platforms, academic databases, educational institutions, etc., and the data types include text, images, videos, etc. The diversity of data sources is crucial for subsequent data processing and analysis. The purpose of data preprocessing is to generate an educational resource input data set. Data preprocessing includes a series of systematic operation steps, mainly including data cleaning, data integration, data conversion and data normalization. The process of data cleaning aims to identify and correct errors, inconsistencies and missing values ​​in the data, for example, removing duplicate entries, correcting incorrect formats and filling missing data. This step ensures the accuracy and completeness of the data and lays a solid foundation for subsequent analysis. Data integration involves aggregating and integrating data from different sources to ensure the consistency and uniformity of the data. It is often necessary to convert and standardize the data format so that it can be processed and analyzed on the same platform. In the data conversion stage, the data needs to be formatted and converted into a structure suitable for subsequent analysis. For example, text data is converted into a structured data table, or video data is extracted into a specific feature vector. In the process of data normalization, the data is standardized so that the data are comparable.

[0073] Step S2: Acquire digital education content data; Use a preset model to draw a multi-dimensional portrait of the education resource input data set to generate digital education multi-dimensional portrait data; Use the digital education multi-dimensional portrait to match the content features of the digital education content data to generate education resource content matching data;

[0074] In an embodiment of the present invention, obtaining digital education content data involves collecting and integrating different types of data from various digital education resources, including e-books, online courses, teaching videos, interactive tools, etc. The acquisition of digital education content data covers the comprehensive collection and preprocessing of data to ensure its effectiveness and consistency in subsequent analysis. A multi-dimensional portrait is drawn for the education resource input data set using a preset model. The "preset model" here refers to a model used for data analysis and feature extraction, such as a collaborative filtering model, a deep learning model, or a support vector machine. These models are used to perform multi-dimensional feature analysis on the education resource input data set to draw a portrait of each education resource in multiple dimensions. For example, the semantic features of the text, the visual features of the video, and the user behavior features of the interactive data can be extracted through a deep learning model, thereby generating multi-dimensional portrait data of the education resource. This multi-dimensional portrait not only covers the basic information of the resource, but also includes comprehensive information such as its performance in different education scenarios, frequency of use, and user evaluation. The digital education content data is matched with content features using the multi-dimensional portrait of digital education. This process involves using feature information in the multi-dimensional portrait data to identify and match corresponding features in the education content data. Specifically, content feature matching can be achieved by calculating the similarity between feature vectors, for example, using cosine similarity, Euclidean distance and other metrics to evaluate the similarity between different contents. Through this matching process, educational resource content matching data can be generated, that is, the matching degree between each educational content and the preset standard or target user needs.

[0075] Step S3: acquiring level meter data and height sensor data; performing scenario analysis on the level meter data and the height sensor data to generate scenario perception data; performing scenario perception scheme generation on the scenario perception data to generate a scenario perception scheme;

[0076] In an embodiment of the present invention, obtaining level meter data and height sensor data involves collecting data related to the horizontal posture and height change of the device from sensors or measuring devices. The level meter is used to measure the horizontality of the device or structure, while the height sensor is used to detect and record height changes. The data collection of the two provides a basic data source for subsequent scenario analysis. Scenario analysis is performed on the level meter data and the height sensor data. Scenario analysis aims to identify the situation or environmental features by processing and parsing the original sensor data. Specific technical means include data fusion and pattern recognition. Data fusion technology integrates data from different sensors into a unified data model to eliminate noise and improve the accuracy of the data. Pattern recognition technology identifies specific patterns or abnormalities in the data by applying statistical analysis and machine learning algorithms. These technologies can help construct scenario-aware data to reflect the state changes of the device under different environmental conditions, such as whether the device is in a stable state, a mobile state, or other specific scenarios. After generating scenario-aware data, a scenario-aware solution is further generated for it. This step analyzes the scenario-aware data by applying a decision support system or a rule engine to formulate a corresponding scenario-aware solution. The solution includes operation instructions or adjustment strategies for different scenario conditions. For example, based on the stability status of the equipment, corresponding operation suggestions can be generated to guide users to take appropriate measures in different scenarios.

[0077] Step S4: Perform multi-dimensional dynamic matching of educational resource content matching data and scenario perception solutions to generate digital educational resource dynamic matching data, thereby completing digital educational resource matching.

[0078] In an embodiment of the present invention, multi-dimensional dynamic matching of educational resource content matching data and scenario-aware solutions. This step involves comprehensive analysis of the two types of data through a dynamic matching algorithm to generate dynamic matching data for digital educational resources. The technical means of this process include data integration, dynamic matching algorithms and their optimization. Educational resource content matching data contains various feature information about educational resources, such as themes, difficulty levels, applicable grades, etc., while scenario-aware solutions involve relevant information about the environment and user behavior, such as learning scenarios, user status, etc. In order to achieve multi-dimensional dynamic matching, it is necessary to integrate these two types of data and connect the characteristics of educational resources with the data level of scenario-awareness. Data integration technologies such as data fusion and data alignment play a key role here. Through these technologies, data from different sources can be unified into a common framework to ensure the compatibility and consistency of the data. A dynamic matching algorithm is used for specific data processing. Dynamic matching algorithms, such as dynamic time warping (DTW) or a recommendation algorithm based on matrix decomposition, comprehensively consider the content characteristics of educational resources and scenario-aware solutions. These algorithms identify the best matching combination by calculating the similarity or correlation between different data dimensions. For example, the dynamic time warping algorithm can handle the comparison of time series data and find the best matching point by optimizing the path. The matrix decomposition algorithm improves the accuracy and efficiency of matching by breaking down complex matching problems into multiple simple sub-problems and solving them step by step. Optimization technology is also indispensable in the dynamic matching process. By adjusting the parameters of the algorithm, adding regularization terms, or using heuristic methods, the quality and accuracy of the matching results can be further improved. For example, constraints can be introduced to ensure the rationality of the matching results in practical applications, or adaptive algorithms can be used to dynamically adjust the matching strategy based on real-time data feedback.

[0079] Preferably, step S1 comprises the following steps:

[0080] Step S11: Acquire educational resource data;

[0081] Step S12: performing data cleaning on the educational resource data to generate educational resource cleansing data;

[0082] Step S13: performing format standardization processing on the education resource cleansing data to generate digital education resource standardized data;

[0083] Step S14: labeling the input data set of the digital educational resources standardization to generate an educational resource input data set.

[0084] In the embodiment of the present invention, data acquisition is involved, and its core technical means include data collection and data storage. The acquisition of educational resource data depends on multiple data sources, such as database extraction, API interface call, and file system reading. This process includes extracting data in various forms such as text, images, and videos from various educational resource management systems to ensure the comprehensiveness and diversity of data sources to meet subsequent processing requirements. Data cleaning is performed on educational resource data. Data cleaning technology aims to identify and correct errors and inconsistencies in data. It mainly includes missing value processing, outlier detection and processing, and duplicate data deletion. Common methods include filling missing values ​​(such as using mean filling or regression interpolation), outlier correction (such as outlier detection based on statistical methods), and data deduplication (such as deduplication based on hash algorithms). These operations can be implemented through data cleaning tools and programming libraries (such as Pandas, R language dplyr package, etc.) to ensure the accuracy and consistency of the data. It involves data format standardization processing. This step mainly converts the cleaned data into a unified format for subsequent analysis and processing. Specific techniques include data format conversion (such as unifying the date and time format to the ISO 8601 standard), data type conversion (such as converting text data into categorical variables), and field naming standardization (such as ensuring the consistency of field names). This is achieved through data conversion tools and scripts, using methods such as regular expressions and conversion functions to standardize the data into a predefined format. Finally, the input data set of digital educational resources is annotated. The annotation process involves adding labels or classification information to each piece of data to enhance the usability of the data. Specific technical means include manual annotation and automatic annotation. Manual annotation involves data experts labeling data according to business needs, while automatic annotation uses machine learning algorithms (such as named entity recognition in natural language processing) to automatically classify data. This process involves the use of annotation tools (such as LabelMe, Labelbox, etc.), as well as training data classification models to improve the accuracy of annotation.

[0085] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes:

[0086] Step S21: Acquire digital education content data and learning demand data;

[0087] Step S22: Use a preset model to draw a multi-dimensional portrait of the educational resource input data set to generate digital education multi-dimensional portrait data;

[0088] Step S23: Use the digital education multi-dimensional portrait data to perform two-dimensional graph data mapping on the learning demand data to generate a digital education supply and demand relationship data graph; use the digital education supply and demand relationship data graph to perform similarity calculation on the digital education content data to generate digital education similarity data; perform content feature matching on the digital education similarity data to generate educational resource content matching data.

[0089] In the embodiment of the present invention, the collection of digital education content data and learning demand data is involved. Technical means include data capture (such as using web crawler technology to obtain data from the education platform), data extraction (such as extracting data of specific fields from a structured database), and data integration (such as merging data from different sources into a unified data storage system). Commonly used tools and technologies include BeautifulSoup and Scrapy frameworks in Python, and SQL query language, etc., which can effectively obtain the required data from a variety of resources. Use a preset model to draw a multi-dimensional portrait of the educational resource input data set to generate digital education multi-dimensional portrait data. This process mainly relies on multi-dimensional data modeling and visualization technology. Multi-dimensional portrait drawing involves building a complex data model to capture various aspects of the data, including dimensional analysis (such as the subject, difficulty, type, etc. of educational resources), and visualization technology (such as using data visualization tools D3.js, Tableau, etc. to display multi-dimensional data). These models and tools can map educational resource data in a multi-dimensional space, thereby generating comprehensive multi-dimensional portrait data. Technical means include two-dimensional graph data mapping of digital education multi-dimensional portrait data, and similarity calculation of digital education supply and demand relationship data graph. Two-dimensional graph data mapping involves converting high-dimensional data into a two-dimensional space through dimensionality reduction techniques (such as principal component analysis PCA, t-SNE, etc.) for easy visualization and analysis. Similarity calculations rely on distance metrics (such as cosine similarity, Euclidean distance, etc.), which can measure the degree of similarity between educational content data and learning needs data. In specific implementations, numerical computing libraries (such as NumPy, SciPy) can be used to perform these calculations. Subsequently, content feature matching is performed on the similarity data. Technical means include feature extraction (such as using the TF-IDF algorithm to extract text features) and matching algorithms (such as matching-based sorting algorithms) to identify educational resources that best match learning needs.

[0090] Preferably, step S22 includes the following steps:

[0091] Step S221: performing high-dimensional mapping on the educational resource input data set through high-dimensional feature space mapping to generate a high-dimensional feature data set;

[0092] Step S222: performing dynamic weight allocation and quantification processing on the high-dimensional feature data set to generate digital education feature weight data; using a feature priority sorting algorithm to sort the importance of the digital education feature weight data to generate a feature priority sorting table;

[0093] Step S223: Use the preset model to draw a multi-dimensional portrait of the educational resource input data set and the feature priority ranking table, and perform sparse processing through the sparse coding method to generate digital education multi-dimensional portrait data.

[0094] In the embodiment of the present invention, "high-dimensional feature space mapping" refers to mapping the educational resource input data set to a high-dimensional feature space in order to more comprehensively capture the complex features of the data. Technical means include high-dimensional feature extraction technology and mapping algorithms. Commonly used high-dimensional feature extraction technologies include principal component analysis (PCA), linear discriminant analysis (LDA) and kernel principal component analysis (KPCA). These methods can map input data from a low-dimensional space to a high-dimensional feature space to improve the ability of feature representation. In the process of generating a high-dimensional feature data set, mathematical tools and programming libraries (such as NumPy, SciPy, Scikit-learn) are required to handle the calculation and mapping of the feature space. The "dynamic weight allocation quantization processing" performed involves assigning weights to each feature in the high-dimensional feature data set, and converting the weight values ​​into a data format that can be used for subsequent analysis through a quantization method. This process includes dynamic weight allocation algorithms (such as weighted least squares, dynamic adjustment algorithms) and quantization processing techniques. Feature priority sorting algorithms (such as information gain-based sorting, weighted average method) are used to sort features according to the degree of influence of features on the final goal and generate a feature priority sorting table. This ranking table can reflect the importance of each feature to the data model or analysis results, thus providing a basis for feature selection and data analysis. In the process of generating the feature priority ranking table, the algorithm library (such as the feature selection module in Scikit-learn) is used to implement the ranking operation. "Multidimensional portrait drawing" is to visualize the educational resource input data set and the feature priority ranking table using a preset model, with the aim of showing the multidimensional features of the data and their interrelationships. This process involves the use of multidimensional data modeling techniques (such as multidimensional scaling analysis MDS, t-SNE, etc.) and visualization tools (such as Matplotlib, Seaborn) to generate digital education multidimensional portrait data. Sparse coding methods are used to sparse multidimensional data to improve the efficiency of data processing and the interpretability of the model. Sparse coding methods (such as dictionary learning and sparse representation) achieve effective compression of features and extraction of important information by reducing the redundant parts of the data, and finally generate sparse digital education multidimensional portrait data. This process involves the use of programming libraries (such as Scikit-learn, TensorFlow) to implement sparse coding technology.

[0095] Preferably, step S23 includes the following steps:

[0096] Step S231: using the embedding vector method to perform high-dimensional deconstruction on the digital education multi-dimensional portrait data to generate digital education portrait deconstruction data; using the digital education portrait deconstruction data to perform two-dimensional graph data mapping on the learning demand data to generate a digital education supply and demand relationship data graph;

[0097] Step S232: optimizing the relationship topology of the digital education supply and demand relationship data graph to generate a digital education supply and demand topology graph; performing similarity calculation on the digital education content data using the digital education supply and demand relationship data graph to generate digital education similarity data;

[0098] Step S233: Perform content feature matching on the digital education similarity data to generate education resource content matching data.

[0099] In an embodiment of the present invention, the "embedded vector method" is used to perform high-dimensional deconstruction of digital education multi-dimensional portrait data to generate digital education portrait deconstruction data. The embedded vector method (such as Word2Vec, GloVe in word embedding technology or BERT in deep learning) embeds high-dimensional data into a low-dimensional vector space to retain the semantic and contextual information in the data, thereby realizing high-dimensional deconstruction and dimensionality reduction processing of the data. This process can extract the potential features in the data and generate a low-dimensional representation for subsequent analysis. The generated digital education portrait deconstruction data provides a basis for the two-dimensional graph data mapping of learning demand data. By performing two-dimensional mapping on the deconstructed data, a digital education supply and demand relationship data graph can be generated, which shows the relationship between learning needs and educational resources. The digital education supply and demand relationship data graph performs "relationship topology graph optimization", which involves using a graph optimization algorithm (such as a graph convolutional network GCN, a shortest path algorithm, and a minimum spanning tree algorithm) to structurally optimize the supply and demand relationship graph to enhance the connectivity and information transfer efficiency of the graph. The optimized graph, namely the digital education supply and demand topology graph, can more clearly show the complex relationship between educational resources and learning needs, and improve the accuracy of supply and demand matching. In addition, using the optimized supply and demand relationship data graph to perform "similarity calculation" on digital education content data involves using similarity measurement methods (such as cosine similarity, Euclidean distance, Jaccard similarity) to calculate the similarity between content and generate digital education similarity data. The similarity calculation process can quantify the correlation between content, thereby providing a basis for subsequent content matching. The "content feature matching" operation is performed on the digital education similarity data, and feature matching algorithms (such as K-nearest neighbor algorithm KNN, support vector machine SVM, or more complex matching algorithms such as neural networks) are used to match similarity data with educational resources. The educational resource content matching data generated by this process reflects the specific matching between resources and needs, and can effectively identify the educational resources that best meet learning needs. The implementation of feature matching requires data processing and machine learning frameworks (such as TensorFlow, PyTorch) to ensure the accuracy and efficiency of the matching results.

[0100] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes:

[0101] Step S31: Acquire data of the level meter and data of the height sensor;

[0102] Step S32: Performing level height information enhancement processing on the level meter data and the height sensor data to generate level-height enhanced data; performing scenario analysis on the level-height enhanced data to generate a scenario state determination result;

[0103] Step S33: generating a situation awareness solution based on the situation state determination result.

[0104] In an embodiment of the present invention, basic data collection is performed by "obtaining level data and height sensor data". The level is used to measure the angle and inclination on the horizontal plane, while the height sensor measures the height change and position in space. These sensor data provide raw information about the environment, which exists in the form of sensor readings and includes noise and errors, so they need to be further optimized and enhanced in subsequent processing. The "horizontal height information enhancement processing" technology is applied to process the collected level data and height sensor data. This step mainly uses signal processing and data enhancement technology to improve data quality and information density. For example, noise is removed by a filter (such as a Kalman filter), or data interpolation technology is used to fill in data gaps, and the data expression ability is improved by enhancement algorithms (such as feature enhancement and data expansion). The generated "horizontal-height enhanced data" not only contains more accurate spatial position and angle information, but also has improved data dimensions and resolution. Based on these enhanced data, "scenario analysis" is performed to generate "scenario state determination results", in which scenario analysis models (such as machine learning classifiers or rule engines) are used to interpret data and identify different environmental scenarios. The core of scenario analysis is to judge the current environment state based on the characteristic patterns of the data, for example, to identify whether the user is in a static learning state or a mobile state. "Scenario-aware solution generation" is performed through the "scenario state judgment results", which is the process of converting the results of environmental state analysis into specific application strategies. The generated "scenario-aware solution" determines the appropriate operation or adjustment strategy based on the scenario state judgment results. This process involves solution design techniques (such as decision trees, rule engines, and optimization algorithms) to formulate corresponding operation strategies according to different environmental scenarios. The generation of scenario-aware solutions ensures that the system can dynamically adapt to environmental changes, thereby achieving more accurate and effective system responses.

[0105] Preferably, step S32 includes the following steps:

[0106] Step S321: performing nonlinear level height information enhancement mapping on the level meter data and the height sensor data to generate level-height enhancement data; performing stable interval identification on the level-height enhancement data to generate level-height stable period data;

[0107] Step S322: combining the data sources of the horizontal-altitude enhancement data and the horizontal-altitude stable period data to generate horizontal-altitude combined source data;

[0108] Step S323: Use the preset horizontal-altitude threshold to perform scenario state determination on the horizontal-altitude joint source data to generate a scenario state determination result, wherein the scenario state determination result includes a moving handheld result and a desk stable state result.

[0109] In an embodiment of the present invention, "non-linear level height information enhancement mapping" is implemented to process the level meter data and the height sensor data. Non-linear mapping technology, such as support vector machine regression (SVR), artificial neural network (ANN) or other non-linear transformation methods, is used to perform complex transformations on the original data to improve the expressive power and information density of the data. This mapping can effectively capture the nonlinear characteristics in the data and generate more descriptive "level-height enhancement data". Subsequently, the "stable interval identification" technology is applied to identify the stable interval in the data through statistical analysis or dynamic time warping (DTW) and other algorithms to generate "level-height stable period data". The purpose of this stability analysis is to find the interval with less change in the data, reflect the stable environmental state, and provide a reliable basis for subsequent analysis. The "data source union" technology is used to fuse the "level-height enhancement data" and the "level-height stable period data". The data source union uses data fusion algorithms such as weighted average, principal component analysis (PCA) or deep learning methods to merge data from different sources into a unified data set-"level-height joint source data". This process not only improves the comprehensiveness and consistency of the data, but also enables the data set to more comprehensively reflect environmental information, laying the foundation for further analysis. The preset "horizontal-altitude threshold" is used to determine the situational state of the "horizontal-altitude joint source data". The "threshold" here refers to the predefined numerical boundary in a specific scenario, which is used to distinguish different states. Based on these thresholds, the joint source data is analyzed by a threshold determination algorithm (such as threshold logic, classification model) to generate "situational state determination results". These results include "mobile handheld results" and "desk stable state results", which correspond to the state of the user on a mobile handheld device or stable learning on the desktop. This determination process uses a machine learning classifier or rule engine to compare the data with the preset threshold to determine the current environmental state.

[0110] Preferably, step S33 includes the following steps:

[0111] Step S331: generating a situation awareness solution based on the situation state determination result, wherein the situation awareness solution includes a mobile handheld solution and a desktop stable learning solution;

[0112] Step S332: When the result of the scenario status determination is a mobile handheld result, the digital education content data is short-term adaptively filtered to generate fragmented learning resource data; the learning demand data is short-term prioritized to generate short-term priority data; the fragmented learning resource data and the short-term priority data are cross-resource validated using the digital education multi-dimensional portrait data to generate mobile handheld cross-validation data; the mobile handheld cross-validation data is mobile adaptively filtered to generate a mobile handheld recommended resource set, thereby completing the mobile handheld solution;

[0113] Step S333: When the scenario state judgment result is a desk stable state result, deep learning feature analysis is performed on the digital education multi-dimensional portrait data to generate deep learning feature data; learning demand matching is performed on the deep learning feature data and the digital education supply and demand relationship data graph to generate deep learning matching data; multi-dimensional learning content recommendation is performed on the deep learning matching data using the digital education supply and demand topology graph to generate multi-dimensional learning content recommendation data; desktop device resource recommendation is performed on the multi-dimensional learning content recommendation data using the digital education portrait deconstruction data to generate a desktop device recommended resource set, thereby completing the desktop stable learning plan.

[0114] In the embodiment of the present invention, the technical means of "context-aware solution generation" is used to convert the result of the situation state determination into a specific operation plan. This process includes generating a "mobile handheld solution" and a "desktop stable learning solution" according to the situation state determination result. The generation of the situation-aware solution depends on the rule engine or decision tree model. These models dynamically generate processing solutions for different situations based on the input situation state determination results to ensure that the system can be adaptively adjusted for different usage scenarios. When the situation state determination result is a "mobile handheld result", the "digital education content data" is first subjected to "short-term adaptive filtering". This technical means includes a real-time data processing algorithm, such as a sliding window method or an adaptive filter, which is used to extract relevant information in a short period of time from a large amount of data to generate "fragmented learning resource data". Then, the "learning demand data" is sorted by "short-term priority sorting", which uses a priority sorting algorithm, such as a priority queue or a weighted sorting method, to generate "short-term priority data". Combined with the "digital education multi-dimensional portrait data", "cross-resource verification" is performed, and the fragmented learning resources and short-term priority data are integrated and verified using cross-validation technology to ensure the relevance of resources and the accuracy of priorities, and generate "mobile handheld cross-validation data". Finally, the cross-validation data is further processed through the "mobile adaptive filtering" technology to generate a "mobile handheld recommended resource set". This process involves the use of filters and recommendation algorithms, such as content-based recommendation algorithms or collaborative filtering, to optimize and generate resource sets suitable for mobile learning, thereby completing the "mobile handheld solution". When the scenario state judgment result is "desk stable state result", the "digital education multi-dimensional portrait data" is first subjected to "deep learning feature analysis". This uses deep learning models, such as convolutional neural networks (CNN) or long short-term memory networks (LSTM), to extract features from multi-dimensional data and generate "deep learning feature data". Subsequently, these feature data and the "digital education supply and demand relationship data graph" are used for "learning demand matching", and "deep learning matching data" is generated through matching algorithms, such as similarity calculation or optimized matching methods. Multi-dimensional learning content recommendation is performed on the "deep learning matching data" according to the "digital education supply and demand topology graph". This process involves graph analysis and recommendation system technologies, such as graph-based recommendation algorithms, to generate "multi-dimensional learning content recommendation data". Finally, the "digital education portrait deconstruction data" is used to perform "desktop device resource recommendations" on the recommended data. This process uses recommendation algorithms, such as content-based recommendations or collaborative filtering methods, to generate a "desktop device recommended resource set" to complete the "desktop stable learning plan."

[0115] As an example of the present invention, refer to Figure 4 As shown, in this example, step S4 includes:

[0116] Step S41: Perform dynamic adaptability evaluation using digital education multi-dimensional portrait data and scenario perception scheme to generate digital education resource adaptability evaluation data; perform weighted allocation again on the digital education resource adaptability evaluation data and education resource content matching data to generate dynamic weight matching data;

[0117] Step S42: Perform multi-dimensional dynamic matching on the dynamic weight matching data to generate digital education resource dynamic matching data, thereby completing digital education resource matching.

[0118] In the embodiment of the present invention, "digital education multi-dimensional portrait data" and "context perception scheme" are used to perform "dynamic adaptability evaluation". This process involves applying an adaptability evaluation model, such as a weighted decision model or an adaptive learning algorithm, to evaluate the adaptability of educational resources in different situations. This includes real-time analysis of the performance of educational resources in specific situations to generate "digital education resource adaptability evaluation data". Adaptability evaluation dynamically evaluates the applicability of resources by comprehensively considering various factors in multi-dimensional portrait data and context perception schemes, thereby quantitatively evaluating the effectiveness of resources. "Digital education resource adaptability evaluation data" and "education resource content matching data" are "re-weighted and allocated", a process that includes recalculating the weights of resources using weighted algorithms (such as weighted average, weighted regression model) to reflect the new evaluation results. This weighted processing uses a dynamic weight allocation algorithm to generate "dynamic weight matching data" taking into account the latest information and actual needs in the evaluation data. This step ensures that the weight allocation of resources can accurately reflect their current adaptability and is adjusted according to the evaluation results to optimize the matching degree of resources. The "dynamic weight matching data" is subjected to "multi-dimensional dynamic matching". This process uses multi-dimensional matching algorithms (such as multi-dimensional scaling analysis, clustering algorithms, or multi-objective optimization techniques) to match the updated resource weights with actual needs. These technical means can handle complex data relationships, taking into account the data characteristics of different dimensions and their dynamic changes, and generate "digital education resource dynamic matching data". Through dynamic matching, the system can adjust and optimize the allocation of resources in real time to ensure a high degree of fit between educational resources and user needs, thereby achieving effective resource matching.

[0119] Preferably, step S41 includes the following steps:

[0120] Step S411: Using the digital education multi-dimensional portrait data and the situational awareness solution to perform situational factor adaptability scoring, and generate digital education resource adaptability scoring data;

[0121] Step S412: Perform fuzzy rule adjustment on the digital education resource adaptability scoring data to generate digital education resource dynamic weight adjustment data;

[0122] Step S413: Use a matrix decomposition algorithm to perform multi-dimensional dynamic matching on the dynamic weight adjustment data of digital educational resources to generate dynamic matching data of digital educational resources, thereby completing the matching of digital educational resources.

[0123] In the embodiment of the present invention, "digital education multi-dimensional portrait data" and "context perception scheme" are used to perform "context factor adaptability scoring" to generate "digital education resource adaptability scoring data". This process conducts a comprehensive evaluation of resources by quantitatively evaluating various types of information in the multi-dimensional portrait data, such as the type of educational content, the user's learning habits, environmental factors, etc., combined with the specific contexts in the context perception scheme (such as learning environment, time limit, etc.). This relies on a rule-based scoring model or adaptive scoring algorithm to calculate the adaptability score of resources by considering multiple context factors to ensure that the resources can meet specific learning needs and environmental conditions. "Fuzzy rule adjustment" is performed on the "digital education resource adaptability scoring data" to generate "digital education resource dynamic weight adjustment data". The fuzzy rule adjustment adopts a fuzzy logic system to convert the adaptability scoring data into dynamic weights by establishing a fuzzy rule base. The fuzzy logic system can handle uncertainty and fuzzy data, and convert the scoring results into specific dynamic weight values ​​to adapt to changing learning needs and contexts. This process effectively adjusts the weights of resources to reflect their applicability and priority in different contexts, thereby optimizing resource allocation. The "digital education resource dynamic weight adjustment data" is "dynamically matched in multiple dimensions" using a "matrix decomposition algorithm" to generate "digital education resource dynamic matching data". Matrix decomposition algorithms, such as singular value decomposition (SVD) or non-negative matrix factorization (NMF), are used to decompose complex relationships in dynamic weight adjustment data and identify potential matching patterns. These algorithms can decompose high-dimensional data matrices into low-dimensional matrices to reveal the underlying structure and correlation in the data. In this way, the algorithm can match resources in multiple dimensions to ensure that each educational resource can optimally match specific learning needs and situations, thereby completing accurate resource matching.

[0124] The beneficial effect of the present invention is that through a multi-dimensional digital education resource matching method, efficient and accurate dynamic adaptation of educational resources and learning needs is achieved from the data level. Using educational resource input data sets and digital education content data, digital education multi-dimensional portrait data is constructed through multi-dimensional portrait drawing and feature priority sorting and other technologies. This process can not only capture the complex characteristics of educational resources in high-dimensional space, but also ensure the accuracy of resource matching in different situations through feature weight allocation and priority sorting. Through the generation of scenario perception schemes, combined with the data of level meters and height sensor meters, the scene perception ability is further enhanced. Through steps such as nonlinear mapping and data source combination, accurate judgment of scenario status is achieved, ensuring the pertinence of learning resource recommendations in different situations. At the same time, the system conducts multi-dimensional dynamic weight adjustment and weighted allocation through dynamic adaptability evaluation, combined with scenario perception schemes and educational resource content matching data, and effectively optimizes the adaptability of educational resources. By deeply analyzing the supply and demand relationship of educational resources, using supply and demand relationship data graphs and similarity calculations, the system can adjust the resource recommendation scheme in a timely manner when learning needs change, ensuring that learners obtain the most suitable resources. The application of technologies such as sparse coding and embedded vectors not only simplifies the complexity of multi-dimensional portrait data, but also ensures the real-time and accuracy of resource recommendations. Therefore, the present invention solves the problems of poor situational adaptability and low resource allocation efficiency of traditional matching methods through a digital education resource matching method based on multi-dimensional feature analysis, situational perception and dynamic weight allocation, and improves the dynamic adaptability and matching accuracy of digital education resources.

[0125] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0126] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A digital educational resource matching method, characterized in that: The following steps are involved: Step S1: obtaining educational resource data; performing data preprocessing on the educational resource data to generate an educational resource input data set; Step S2: Acquire digital education content data; Use a preset model to draw a multi-dimensional portrait of the education resource input data set to generate digital education multi-dimensional portrait data; Use the digital education multi-dimensional portrait data to match the content features of the digital education content data to generate education resource content matching data; Step S3: acquiring level meter data and height sensor data; performing scenario analysis on the level meter data and the height sensor data to generate scenario perception data; performing scenario perception scheme generation on the scenario perception data to generate a scenario perception scheme; Step S4: Perform multi-dimensional dynamic matching of educational resource content matching data and scenario-aware solutions to generate digital educational resource dynamic matching data, thereby completing digital educational resource matching.

2. The digital educational resource matching method according to claim 1, characterized in that: Step S1 specifically includes: Step S11: Acquire educational resource data; Step S12: performing data cleaning on the educational resource data to generate educational resource cleansing data; Step S13: performing format standardization processing on the education resource cleansing data to generate digital education resource standardized data; Step S14: labeling the input data set of the digital educational resources standardization to generate an educational resource input data set.

3. The digital educational resource matching method according to claim 1, characterized in that: Step S2 specifically includes: Step S21: Acquire digital education content data and learning demand data; Step S22: Use a preset model to draw a multi-dimensional portrait of the educational resource input data set to generate digital education multi-dimensional portrait data; Step S23: Use the digital education multi-dimensional portrait data to perform two-dimensional graph data mapping on the learning demand data to generate a digital education supply and demand relationship data graph; use the digital education supply and demand relationship data graph to perform similarity calculation on the digital education content data to generate digital education similarity data; perform content feature matching on the digital education similarity data to generate educational resource content matching data.

4. The digital educational resource matching method according to claim 3, characterized in that: Step S22 specifically includes: Step S221: performing high-dimensional mapping on the educational resource input data set through high-dimensional feature space mapping to generate a high-dimensional feature data set; Step S222: performing dynamic weight allocation and quantification processing on the high-dimensional feature data set to generate digital education feature weight data; using a feature priority sorting algorithm to sort the importance of the digital education feature weight data to generate a feature priority sorting table; Step S223: Use the preset model to draw a multi-dimensional portrait of the educational resource input data set and the feature priority ranking table, and perform sparse processing through the sparse coding method to generate digital education multi-dimensional portrait data.

5. The digital educational resource matching method according to claim 3, characterized in that: Step S23 specifically includes: Step S231: using the embedding vector method to perform high-dimensional deconstruction on the digital education multi-dimensional portrait data to generate digital education portrait deconstruction data; using the digital education portrait deconstruction data to perform two-dimensional graph data mapping on the learning demand data to generate a digital education supply and demand relationship data graph; Step S232: optimizing the relationship topology of the digital education supply and demand relationship data graph to generate a digital education supply and demand topology graph; performing similarity calculation on the digital education content data using the digital education supply and demand relationship data graph to generate digital education similarity data; Step S233: Perform content feature matching on the digital education similarity data to generate education resource content matching data.

6. The digital educational resource matching method according to claim 1, characterized in that: Step S3 specifically includes: Step S31: Acquire data of the level meter and data of the height sensor; Step S32: Performing level height information enhancement processing on the level meter data and the height sensor data to generate level-height enhanced data; performing scenario analysis on the level-height enhanced data to generate a scenario state determination result; Step S33: generating a situation awareness solution based on the situation state determination result.

7. The digital educational resource matching method according to claim 6, characterized in that: Step S32 specifically includes: Step S321: performing nonlinear level height information enhancement mapping on the level meter data and the height sensor data to generate level-height enhancement data; performing stable interval identification on the level-height enhancement data to generate level-height stable period data; Step S322: combining the data sources of the horizontal-altitude enhancement data and the horizontal-altitude stable period data to generate horizontal-altitude combined source data; Step S323: Use the preset horizontal-altitude threshold to perform scenario state determination on the horizontal-altitude joint source data to generate a scenario state determination result, wherein the scenario state determination result includes a moving handheld result and a desk stable state result.

8. The digital educational resource matching method according to claim 6, characterized in that: Step S33 specifically includes: Step S331: generating a situation awareness solution based on the situation state determination result, wherein the situation awareness solution includes a mobile handheld solution and a desktop stable learning solution; Step S332: When the result of the scenario status determination is a mobile handheld result, the digital education content data is short-term adaptively filtered to generate fragmented learning resource data; the learning demand data is short-term prioritized to generate short-term priority data; the fragmented learning resource data and the short-term priority data are cross-resource validated using the digital education multi-dimensional portrait data to generate mobile handheld cross-validation data; the mobile handheld cross-validation data is mobile adaptively filtered to generate a mobile handheld recommended resource set, thereby obtaining a mobile handheld solution; Step S333: When the scenario state judgment result is a desk stable state result, deep learning feature analysis is performed on the digital education multi-dimensional portrait data to generate deep learning feature data; learning demand matching is performed on the deep learning feature data and the digital education supply and demand relationship data graph to generate deep learning matching data; multi-dimensional learning content recommendation is performed on the deep learning matching data using the digital education supply and demand topology graph to generate multi-dimensional learning content recommendation data; desktop device resource recommendation is performed on the multi-dimensional learning content recommendation data using the digital education portrait deconstruction data to generate a desktop device recommended resource set, thereby obtaining a desktop stable learning plan.

9. The digital educational resource matching method according to claim 1, characterized in that: Step S4 specifically includes: Step S41: Perform dynamic adaptability evaluation using digital education multi-dimensional portrait data and scenario perception scheme to generate digital education resource adaptability evaluation data; perform weighted allocation again on the digital education resource adaptability evaluation data and education resource content matching data to generate dynamic weight matching data; Step S42: Perform multi-dimensional dynamic matching on the dynamic weight matching data to generate digital education resource dynamic matching data, thereby completing digital education resource matching.

10. The digital educational resource matching method according to claim 9, characterized in that: Step S41 specifically includes: Step S411: Using the digital education multi-dimensional portrait data and the situational awareness solution to perform situational factor adaptability scoring, and generate digital education resource adaptability scoring data; Step S412: Perform fuzzy rule adjustment on the digital education resource adaptability scoring data to generate digital education resource dynamic weight adjustment data; Step S413: Use a matrix decomposition algorithm to perform multi-dimensional dynamic matching on the dynamic weight adjustment data of digital educational resources to generate dynamic matching data of digital educational resources, thereby completing the matching of digital educational resources.

Images