Teaching resource dynamic allocation method and system based on education software platform data

Abstract

The invention discloses a teaching resource dynamic allocation method and system based on education software platform data, and relates to the technical field of informationization, and the method comprises the steps: S1, obtaining the geographic position information of education resources, and carrying out the standardization processing of the position data, and obtaining a structured spatial distribution frame; s2, acquiring logical relationship data according to the spatial distribution framework, performing association mapping on the logical relationship data and the geographical location information, and determining a multi-dimensional resource attribute set; s3, aiming at the multi-dimensional resource attribute set, integrating the geographic position information and the logical relationship data into a resource distribution presentation layer to obtain a dynamically updated resource distribution view; according to the teaching resource dynamic allocation method and system based on the educational software platform data, the resource balance is continuously evaluated according to the adjusted distribution state, the optimization process is repeated when necessary, finally, dynamic optimal allocation of the educational resources is achieved, the resource utilization efficiency is improved, and educational fairness is promoted.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to a method and system for dynamically allocating teaching resources based on educational software platform data. Background Art

[0002] In the field of educational resource management and optimization, building an efficient resource distribution and allocation system is crucial. This area is directly related to educational equity and resource utilization efficiency, and is an indispensable part of promoting educational modernization. The rational allocation of physical and virtual resources can effectively improve teaching quality and narrow regional disparities. However, current research and practice often expose deep-seated shortcomings when faced with complex resource management demands.

[0003] Existing resource management methods often focus on single-dimensional information display or static recording, making them incapable of addressing the dynamic changes in resource distribution and the need for multi-party collaboration. In particular, the integrated presentation of resource location, ownership, and usage status lacks an effective linkage mechanism, hindering managers from quickly grasping the overall situation and making flexible adjustments based on actual needs. This information silo phenomenon leads to inefficient resource allocation, often resulting in idle resources or uneven resource allocation. In this context, the primary challenge is to deeply integrate geographic location information with logical relationship information to achieve a visual representation of resource distribution. Because geographic location information has spatial attributes, while logical relationship information involves multiple attributes such as resource ownership and usage frequency, combining the two requires striking a balance between data integration and visualization. Furthermore, the complexity of this integration presents another challenge: how to implement real-time resource allocation and status monitoring in a dynamic environment to ensure the rationality of allocation paths and convenient operation. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for dynamic allocation of teaching resources based on educational software platform data, and to construct a teaching resource distribution map that supports intuitive display and dynamic allocation based on the integration of geographic information and logical relationships.

[0005] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: a method for dynamically allocating teaching resources based on educational software platform data, comprising S1, obtaining geographic location information of educational resources, standardizing the location data, and obtaining a structured spatial distribution framework; S2, according to the spatial distribution framework, obtaining logical relationship data, and associating and mapping the logical relationship data with the geographic location information to determine a multi-dimensional resource attribute set; S3, for the multi-dimensional resource attribute set, integrating the geographic location information and the logical relationship data into a resource distribution presentation layer to obtain a dynamically updated resource distribution view; S4, from the dynamically updated resource distribution view, obtaining resource usage frequency and real-time status monitoring data, judging whether there is idle or uneven resource distribution, and determining the resources that need to be adjusted. Node; S5. If a resource node that needs to be adjusted is detected, the optimal allocation path is calculated to obtain a path plan for resource reallocation; S6. According to the path plan for resource reallocation, the execution parameters of the dynamic allocation mechanism are obtained, and automated allocation instructions are generated to determine the specific time and scope of allocation execution; S7. Through automated allocation instructions, real-time status monitoring data is obtained, the resource distribution presentation layer is updated, and the adjusted resource distribution status record is obtained; S8. According to the adjusted resource distribution status record, the latest data on resource usage frequency is obtained, the balance of resource distribution is re-analyzed, and it is determined whether the preset allocation balance threshold is reached; S9. If the preset allocation balance threshold is not reached, the allocation instruction is regenerated, the resource distribution view is updated, and the final resource optimization distribution result is obtained.

[0006] Preferably, the S1 includes obtaining the geographic location records of educational resources through a pre-established geographic information database, performing preliminary cleaning on the geographic location records, and using general tools to fill in and eliminate missing values ​​and outliers in the records to obtain a sorted location information set; based on the sorted location information set, using spatial attribute analysis tools to standardize the location information set, and performing unified conversion on data formats from different sources. If the data format does not meet the preset threshold, it is remapped to determine a standardized location data set; through the standardized location data set, the spatial distribution law of the data set is mined in combination with the location analysis tool to obtain the spatial aggregation characteristics and distribution differences in the data set, and determine the geographical distribution hot spots of educational resources; based on the geographical distribution hot spots, using data integration tools to construct a structured spatial distribution framework, and performing multi-dimensional fusion on the distribution data set to obtain the final geographical distribution data set.

[0007] Preferably, the S2 includes obtaining classification information of educational resources from pre-established resource attribution classification data, performing preliminary matching on the classification information and the geographic location information, and using a data fusion tool to associate the attribution classification data with the location information to obtain a preliminary resource attribute set; obtaining relevant records of resource usage frequency through the preliminary resource attribute set, cleaning incomplete data in the records, and using a general data completion tool to fill in the missing parts to determine a complete resource usage frequency data set; if the value in the complete resource usage frequency data set exceeds a preset threshold, a data standardization tool is used to adjust the range of the data set, and the adjusted data is secondary fused with the preliminary resource attribute set to determine a multi-dimensional resource attribute association set; based on the multi-dimensional resource attribute association set, a data mapping tool is used to integrate the resource attribution classification, usage frequency and geographic location information to obtain a final multi-dimensional resource attribute set.

[0008] Preferably, the S3 includes using a data mapping tool to perform preliminary correspondence processing on the resource distribution and geographic coordinate information based on pre-established resource distribution data and geographic coordinate information, integrating the correspondence data to obtain a preliminary integrated distribution data set; using the preliminary integrated distribution data set, using a visual drawing tool to match the distribution data set and the distribution layer; if there is an inconsistency between the location mark and the layer information, adjusting the inconsistent part to determine the adjusted distribution layer data set; obtaining the adjusted distribution layer data set, using an interface rendering tool to adapt the distribution layer data set to the interface display requirements, dynamically presenting the adapted data to obtain a dynamically presented resource distribution view; for the dynamically presented resource distribution view, using a timed refresh tool in combination with the update frequency data to periodically collect and process the view update content; if the update frequency exceeds a preset threshold, the view content is immediately adjusted to determine the final resource distribution view.

[0009] Preferably, the S4 includes obtaining relevant information of resource frequency and status monitoring from the dynamically updated resource distribution view, classifying and arranging the information using a data extraction tool, classifying the data of resource status and abnormal nodes, and obtaining a classified resource status set; according to the classified resource status set, using a comparative analysis tool to compare the data of abnormal nodes and idle phenomena one by one, if the data of the abnormal node exceeds a preset threshold, it is marked as a potential uneven distribution point, and the marked abnormal distribution data set is determined; for the marked abnormal distribution data set, obtaining relevant content of node distribution and real-time data, using a data mapping tool to correspond the data set with the content presented by the distribution, and obtaining a corresponding node adjustment directory; through the corresponding node adjustment directory, combined with the requirements of monitoring frequency and view update, using a timed refresh tool to periodically check the nodes in the directory, if it is found that the resource status is unevenly distributed, the relevant nodes are prioritized, and the resource node directory that ultimately needs to be adjusted is determined.

[0010] Preferably, the S5 includes relevant data divided according to geographic location and region, and uses a map data processing tool to hierarchically organize the node distribution in the target area, and matches the geographic location information of the node with the resource ownership to obtain a hierarchical node distribution data set; through the hierarchical node distribution data set, combined with the data on resource demand and node priority, a path calculation tool is used to compare the path length and delivery time from the starting node to the target node to determine a preliminary path planning scheme; if the path length or delivery time in the preliminary path planning scheme exceeds a preset threshold, a dynamic adjustment tool is used to combine real-time monitoring data and transportation cost information to locally optimize the path to obtain an adjusted path plan; for the adjusted path plan, relevant indicators of allocation efficiency are obtained, and a data comparison tool is used to perform comparison and analysis to determine the final resource reallocation path plan.

[0011] Preferably, the S6 includes obtaining the mapping relationship between execution parameters and regional divisions from a pre-established repository based on the relevant data of the path plan and dynamic allocation, and using a data sorting tool to hierarchically process the parameters in order to facilitate operation, thereby obtaining a hierarchical parameter set; using the hierarchical parameter set, in combination with the real-time monitoring data and the basis for priority sorting, the matching tool is used to perform an initial division of the allocation time and allocation scope to determine the initial time and scope plan; if the initial time and scope plan does not meet the demand matching standard, the path adjustment tool is used to partially correct the plan in combination with the real-time status of resource allocation, and the revised time and scope match is determined; for the revised time and scope match, the generation rule of the automation instruction is obtained, and the instruction construction tool is used to integrate the match and the dynamic allocation conditions to generate the final automation allocation instruction.

[0012] Preferably, the S7 includes obtaining the dynamic status information of the resource nodes from a pre-established monitoring platform through automated instructions, performing preliminary sorting on the characteristics of the node changes, and obtaining an initial status data set; based on the initial status data set, using a data comparison tool to analyze the distribution presentation of the resource nodes; if the node changes exceed a preset threshold, triggering the distribution presentation update process and determining the updated priority sequence; after obtaining the priority sequence, using a presentation layer generation tool to locally adjust the distribution presentation based on the current status of the resource distribution to obtain updated presentation layer data; using the updated presentation layer data, combined with real-time feedback of the dynamic status, using a support vector machine algorithm to classify the adjustment status of the resource distribution to determine whether the classified distribution meets the requirements of the deployment instruction; if the classified distribution meets the requirements of the deployment instruction, integrating the adjustment status with the distribution record through a status recording tool to obtain a final status record file; based on the final status record file, using a data mapping tool to generate an update log of the distribution record based on the dynamic status changes of the resource nodes to determine the latest status presentation of the resource distribution; using the update log, combined with the real-time monitoring data stream, continuously calibrating the presentation layer of the resource distribution to obtain a calibrated distribution status record.

[0013] Preferably, the S8 includes obtaining the latest data stream of resource usage frequency from a pre-established monitoring platform based on the adjusted resource distribution status record, and performing preliminary sorting on the characteristics of the frequency change to obtain a summary data set of the frequency change; using a data fusion tool to integrate the summary data set with the current status of the resource distribution, extracting features based on the distribution adjustment requirements during the integration, and determining the fused distribution feature data set; using the distribution feature data set, using a data comparison tool to recalculate the balance of the resource distribution, and if the calculation result does not reach the preset distribution balance threshold, triggering the balance adjustment process to determine the adjusted balance distribution status; based on the adjusted balance distribution status, combined with the real-time data stream, using a status comparison tool to continuously verify the resource distribution and the distribution balance target to obtain the final distribution balance record.

[0014] The teaching resource dynamic allocation system based on educational software platform data is used to implement the steps of the teaching resource dynamic allocation method based on educational software platform data. The system includes a location information acquisition module for acquiring the geographical location information of educational resources and standardizing the location data to construct a structured spatial distribution framework; a data association module for acquiring the logical relationship data of resources according to the spatial distribution framework, and associating and mapping the logical relationship data with the geographical location information to determine a multi-dimensional resource attribute set; a layer generation module for integrating the geographical location information with the logical relationship data based on the multi-dimensional resource attribute set to generate a resource distribution presentation layer and form a dynamically updated resource distribution view; a status analysis module for obtaining the resource usage frequency and real-time status monitoring data from the resource distribution view to determine whether there is idle or uneven resource distribution. And determine the resource nodes that need to be adjusted; the path planning module is used to calculate the optimal allocation path and generate a path plan for resource reallocation when the resource nodes that need to be adjusted are detected; the instruction generation module is used to obtain the execution parameters of the dynamic allocation mechanism according to the path plan, generate automatic allocation instructions, and determine the specific time and scope of allocation execution; the allocation execution module is used to execute resource allocation operations according to the automatic allocation instructions, and obtain updated real-time status monitoring data, update the resource distribution presentation layer, and form an adjusted resource distribution status record; the analysis feedback module is used to obtain the latest data on resource usage frequency based on the adjusted resource distribution status record, analyze the balance of resource distribution, and determine whether the preset allocation balance threshold is reached; the optimization iteration module is used to regenerate allocation instructions and update the resource distribution view when the allocation balance threshold is not reached, and output the final resource optimization distribution result.

[0015] It can be seen from the above technical solution that the present invention has the following beneficial effects: This method and system for dynamic allocation of teaching resources based on educational software platform data obtains educational resource location information by establishing a geographic information database, combines spatial attribute analysis to form a unified geographic distribution data set, and associates and maps with logical relationship data such as resource ownership and usage frequency to generate a visual resource distribution view. Based on this view, the present invention analyzes resource usage status, identifies idle or unevenly distributed resource nodes, calculates the optimal allocation plan using allocation path optimization technology, and executes resource reallocation through automated instructions. The present invention can also continuously evaluate resource balance based on the adjusted distribution status, repeat the optimization process when necessary, and ultimately achieve dynamic optimization allocation of educational resources, improve resource utilization efficiency, and promote educational equity. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 Flow chart of the method of the present invention; Figure 2This is a system connection diagram of the present invention. DETAILED DESCRIPTION

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only 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 ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0018] like Figure 1 As shown, the present invention provides a technical solution: a method for dynamically allocating teaching resources based on educational software platform data, including S1, obtaining geographical location information of educational resources, standardizing the location data, and obtaining a structured spatial distribution framework; S2, according to the spatial distribution framework, obtaining logical relationship data, and associating and mapping the logical relationship data with the geographical location information to determine a multi-dimensional resource attribute set; S3, for the multi-dimensional resource attribute set, integrating the geographical location information and the logical relationship data into a resource distribution presentation layer to obtain a dynamically updated resource distribution view; S4, from the dynamically updated resource distribution view, obtaining resource usage frequency and real-time status monitoring data, judging whether there is idle or uneven resource distribution, and determining the resource nodes that need to be adjusted ; S5. If a resource node that needs to be adjusted is detected, calculate the optimal allocation path and obtain the path plan for resource reallocation; S6. According to the path plan for resource reallocation, obtain the execution parameters of the dynamic allocation mechanism, generate automated allocation instructions, and determine the specific time and scope of allocation execution; S7. Through automated allocation instructions, obtain real-time status monitoring data, update the resource distribution presentation layer, and obtain the adjusted resource distribution status record; S8. According to the adjusted resource distribution status record, obtain the latest data on resource usage frequency, re-analyze the balance of resource distribution, and determine whether the preset allocation balance threshold is reached; S9. If the preset allocation balance threshold is not reached, regenerate the allocation instruction, update the resource distribution view, and obtain the final resource optimization distribution result.

[0019] This implementation achieves structured resource management by constructing a spatial distribution framework for educational resources and multi-dimensionally associating logical relationship data with geographic location information. Dynamic visualization using a resource distribution presentation layer enables real-time understanding of the geographical distribution and usage status of teaching resources. By analyzing resource usage frequency and real-time status monitoring data, it identifies idle or unevenly distributed resources. It further calculates the optimal resource allocation path, generates automated allocation instructions, and executes dynamic allocation operations. After allocation is complete, the resource status is updated, and a determination is made again as to whether the resource distribution is balanced, thereby achieving closed-loop optimization control.

[0020] S1 includes obtaining the geographic location records of educational resources through a pre-established geographic information database, performing preliminary cleaning on the geographic location records, and using general tools to fill in and remove missing values ​​and outliers in the records to obtain a sorted location information set; based on the sorted location information set, using spatial attribute analysis tools to standardize the location information set, uniformly converting data formats from different sources, and remapping if the data format does not meet the preset threshold to determine a standardized location data set; using the standardized location data set, combining the location analysis tool to mine the spatial distribution patterns of the data set, obtaining spatial aggregation characteristics and distribution differences in the data set, and determining the geographic distribution hotspots of educational resources; based on the geographic distribution hotspots, using data integration tools to construct a structured spatial distribution framework, and performing multi-dimensional fusion on the distribution data set to obtain the final geographic distribution data set. In this embodiment, the geographic information database is first connected through the data interface preset by the education platform. The location information of educational resources stored in the database is organized in a structured record format. Each record contains a unique resource number, corresponding administrative division code, resource name, longitude value, and latitude value. All records are retrieved and stored in the "initial location information set". This information set then enters a two-step data cleansing process. The first step involves missing value handling. Any record whose longitude or latitude field is empty or contains invalid characters (such as non-numeric characters) is considered missing. Spatial interpolation is performed on all missing records. This method involves selecting the five closest valid location records from the administrative division to which the current record belongs, calculating the average of their longitude and latitude, and replacing the missing fields with this average. The second step involves outlier removal. The outlier criterion is that the average spatial distance between a single record and other records in the same division exceeds a set threshold of 20 kilometers. Spatial distance is calculated using a spherical distance formula. Considering the Earth's radius of 6,371 kilometers, the longitude and latitude of each pair of records are converted to spherical coordinates, and then the arc length distance is calculated. If the average distance between a record and its five neighboring records exceeds 20 kilometers, it is marked as an outlier and removed. After cleansing, the resulting location information set enters the standardization phase. The first step in standardization is format unification, which converts all location record formats into floating-point longitude and latitude coordinates, retains 6 decimal places, and ensures accuracy at the meter level. The second step is range verification, setting the legal longitude range to 60 to 140, and the legal latitude range to 10 to 55. Records outside this range are directly sent to the exception handling process. The exception handling process is remapping, and the operation method is: using the resource number of the record as the key field, search the historical archive table for old records with the same number but coordinate values ​​within the legal range. If found, the old coordinates are used to overwrite the current values. If not found, the record will be added to the exception log and marked as requiring manual review. After completing the format unification and remapping operations, the standardized location data is sent to the spatial aggregation analysis module.Cluster analysis uses a fixed-radius scanning method, with a scanning radius of 5 kilometers. The scanning window slides across the entire spatial range in steps of 0.5 kilometers. At each center point, the number of records within the coverage area is counted. If a center point covers at least 10 records within the radius, it is defined as a spatial cluster. Regional aggregation is performed on all clusters, and adjacent clusters with a spatial distance of no more than 5 kilometers are grouped into a hotspot cluster. All hotspot clusters are further divided into square grids with a side length of 1 kilometer. The number of resources contained in each grid is calculated. A grid is identified as a hotspot if the number of records contained is greater than 1.5 times the average number of resources for all grids in the region. The average number is calculated by dividing the total number of resources in all valid grids by the total number of grids, and then multiplying by 1.5 to obtain the threshold. Next, the minimum enclosing rectangular area is calculated based on the boundaries of all hotspot grids to form the hotspot boundary. Resource attribute data is extracted within each hotspot area, including resource type (e.g., elementary school classroom, multimedia classroom, laboratory), year of operation, average daily usage time, and number of people served. All attributes are written to a spatial distribution table through field aggregation operations, with each hotspot unit receiving a set of multi-dimensional attribute values. This structured table is sorted by hotspot number and administrative division number, forming a complete structured spatial distribution framework. This framework serves as the basic input data for the resource allocation algorithm, enabling subsequent logical association and path optimization. For missing value processing, five adjacent points are selected. This is achieved by sorting all records in ascending latitude and longitude order and then calculating the average of the five points preceding and following the point being processed, plus the point itself, to ensure spatial continuity. The anomaly distance threshold is set at 20 kilometers. This is based on statistical analysis of resource distribution data from districts and counties across China over the past three years. 90% of pairs of resources within the same division are within 20 kilometers of each other, thus serving as the anomaly detection boundary. For spatial cluster analysis, the 5-kilometer scanning radius and 10-point density threshold are derived from resource density distribution characteristic curves analyzed in a test environment for sample areas of 10 major cities, including Beijing, Guangzhou, and Chengdu, to ensure coverage of approximately 95% of valid cluster points in a typical urban area. The hotspot grid standard is set at 1.5 times the average number of resources. This is calculated by multiplying the standard deviation of the number of resources in all valid grids by 0.5 and adding the average value, reflecting the principle of significance. A grid side length of 1 km was chosen to balance execution efficiency and cluster detection accuracy during performance testing. A grid side length of 1 km performed optimally, with calculation times under 2 seconds and accuracy coverage exceeding 92%. These parameters are fixed in the configuration file and can be adjusted based on regional deployment strategies.

[0021] S2 includes obtaining classification information of educational resources from pre-established resource attribution classification data, performing preliminary matching of the classification information with the geographic location information, and using a data fusion tool to associate the attribution classification data with the location information to obtain a preliminary resource attribute set; obtaining relevant records of resource usage frequency through the preliminary resource attribute set, cleaning incomplete data in the records, and using a general data completion tool to fill in the missing parts to determine a complete resource usage frequency data set; if the value in the complete resource usage frequency data set exceeds a preset threshold, a data standardization tool is used to adjust the range of the data set, and the adjusted data is secondary fused with the preliminary resource attribute set to determine a multi-dimensional resource attribute association set; based on the multi-dimensional resource attribute association set, a data mapping tool is used to integrate the resource attribution classification, usage frequency and geographic location information to obtain a final multi-dimensional resource attribute set.

[0022] In this embodiment, the preloaded resource attribution classification database is first accessed. Each resource record in the database contains fields including resource number, name of the affiliated school, campus identification code, administrative division code, resource function type (such as ordinary classroom, multimedia classroom, laboratory, etc.), course attribute label (such as Chinese, mathematics, physics, etc.) and other information. Based on the resource number field as the unique index key, the classification information is matched one by one with the standardized geographic location information set obtained in step S1. The matching method is to perform an equivalue connection on the two data sets according to the resource number field. If a record exists in both data sets, the geographic location information and the classification information are merged to form a complete record and added to the "preliminary resource attribute set". In this step, each merged record contains fields such as longitude, latitude, resource type, course category, affiliated school and its administrative division, and the total number of fields is not less than 6 items.

[0023] After completing the preliminary attribute collection, the resource usage frequency dataset processing process begins. First, resource usage records from the past 30 days are extracted from the platform log database. The following fields are archived by resource number and statistically analyzed: daily usage count, average duration per usage (in minutes), maximum consecutive usage days over the past 30 days, and weekly usage frequency (number of days per week). If a field contains a null value or an abnormal format (such as a negative number or non-numeric string), the data is considered incomplete and enters the completion process. Missing records are selected for completion using a two-level filtering mechanism: the first level filters for "same resource function type" and the second level filters for "geographic distance no more than 10 kilometers." This filtering method first identifies records with identical function types from the resource attribute collection, then calculates the geographic spherical distance between them and the current record, selecting the five records with the smallest distances as reference samples. If there are fewer than five records, all matching records are selected. The arithmetic mean of the field is then taken and written into the current record as the fill-in value for the missing field, completing the data completion process. Once the completion is complete, a "complete resource usage frequency dataset" is generated. Subsequently, a data validation process is performed to verify the validity of each field value in the complete dataset. The legal value range is set based on historical platform statistics: daily usage must not exceed 12 times, the average duration of each session must not exceed 90 minutes, the total daily usage must not exceed 600 minutes, and the maximum number of consecutive days of use must not exceed 30 days. These thresholds are determined based on the following: the 12-time limit represents the theoretical maximum frequency of use based on the daily schedule of courses and activities; 90 minutes represents the duration of two standard classes; 600 minutes represents the limit of 10 hours of work time; and 30 days represents the maximum window of historical records. If a field in a record exceeds these thresholds, it is sent to the normalization module. Normalization uses linear normalization to compress field values ​​to a range between 0 and 1. Specifically, for each field, the minimum and maximum values ​​in the current dataset are calculated. The minimum value is then subtracted from each field value, and the result is divided by the difference between the maximum and minimum values, with the result rounded to three decimal places. After processing, all numeric fields are converted to normalized values ​​to avoid attribute fusion bias due to different scales. After standardization, the processed frequency dataset is again joined with the preliminary resource attribute set using the resource number as the key, producing a "standardized resource attribute fusion data table." Multidimensional attribute coupling analysis is then performed on the fused data table, enumerating the frequency of combinations of fields such as "resource type," "course attribute," and "frequency of use" across all resource records. If a combination occurs in 10% or more of all records, it is considered a high-frequency correlation and entered into the "multidimensional resource attribute correlation set." For example, if the combination "Multimedia classroom - Physics course - Daily usage exceeds 10 times" appears 120 times out of 1000 records, exceeding the 10% threshold of 100 occurrences, the combination is recorded as a strongly correlated attribute combination.The 10% threshold was established through statistical analysis of high-frequency resource coupling characteristics from past projects, enabling the stable selection of representative resource attribute structures. Finally, a data mapping tool was used to structurally fuse the "multi-dimensional resource attribute association set" with the original attribution classification information, usage frequency data, and geographic location information. This fusion was performed by vertically aligning fields, ensuring that each record contained the resource number, administrative division, resource type, course attribute label, standardized usage frequency value, and longitude and latitude information. Once the fusion was complete, the "final multi-dimensional resource attribute set" was generated.

[0024] S3 includes using a data mapping tool to perform preliminary correspondence processing on the resource distribution and geographic coordinate information based on pre-established resource distribution data and geographic coordinate information, integrating the correspondence data to obtain a preliminary integrated distribution data set; using the preliminary integrated distribution data set, using a visual drawing tool to match the distribution data set and the distribution layer; if there is an inconsistency between the location mark and the layer information, adjusting the inconsistent part to determine the adjusted distribution layer data set; obtaining the adjusted distribution layer data set, using an interface rendering tool to adapt the distribution layer data set to the interface display requirements, dynamically presenting the adapted data to obtain a dynamically presented resource distribution view; for the dynamically presented resource distribution view, using a timed refresh tool in combination with the update frequency data to periodically collect and process the view update content; if the update frequency exceeds the preset threshold, the view content is immediately adjusted to determine the final resource distribution view.

[0025] In this embodiment, the static description information of all teaching resources is first called from the educational resource distribution database. This information includes the unique number of each resource, the administrative division code, the resource classification level and the classified resource density label. At the same time, the longitude and latitude geographic coordinates of each resource are read from the coordinate database. Both data tables are matched with the resource number as the primary key field. Through the data mapping module, an inner join operation is performed on the two tables, and all resources existing in the two tables are formed into a new record set in a one-to-one correspondence. Each record contains 7 fields: resource number, longitude, latitude, administrative division code, resource category, density level, and distribution label. The data mapping tool constructs a matching index table based on the hash index to achieve O(1) level data docking efficiency. This set is defined as the "preliminary integrated distribution data set". After the preliminary integration is completed, the map layer matching stage is entered. First, the base map is loaded, that is, the administrative area boundary vector layer. This layer uses GeoJSON or Shapefile format and contains the coordinates of the boundary polygon vertices and the corresponding area identification code. Subsequently, the resource points are mapped onto the base map. The drawing method is as follows: for each resource, its latitude and longitude coordinate values ​​are extracted as the center of the layer drawing point, the radius is set to 5 pixels, and the color coding rules are set according to the density level field. The color coding rules are defined in the configuration as follows: red is used for high density level, orange is used for medium density level, and green is used for low density level. The color values ​​are defined using RGB values, which are 255-0-0, 255-165-0, and 0-128-0 respectively. After the initial drawing of the layer is completed, the layer consistency verification module is called to perform a spatial inclusion judgment on the geographic coordinates of the drawn resource points and the boundaries of the administrative regions to which they belong. The judgment criterion is whether the coordinate point of the resource point is located inside the polygon of the corresponding region. This operation uses the ray method to determine whether the point falls into the polygon. If a point falls outside the regional boundary, it is considered that the location mark is inconsistent with the layer information. The deviation threshold is set to no more than 0.0005 degrees of difference in longitude or latitude coordinates, which is approximately equivalent to 55 meters of ground distance in a straight line. This is based on test data on resource distribution in major urban areas across China, which shows that over 95% of registered resources fall within this range. Any deviation outside this range is considered significant. For offset records, the automatic adjustment module is activated. Specifically, it checks whether the administrative region code to which it belongs is incorrectly filled or mismatched. If so, the coordinates of the polygon's center point are calculated from the region's vector boundary data. The longitude and latitude values ​​of this center point are used to directly replace the original coordinates of the resource, completing the position correction. All corrected records are merged into the "Adjusted Distribution Layer Dataset." After the layer adjustment, the interface rendering adaptation phase begins. According to the user interface layout specifications, the resource layer is scaled to the interface resolution. The layer scale is set to 80% of the map container width, and the point layer marker level is set to the top of the layer stack to ensure it is not obscured by the background map.All point markers are accompanied by a floating label containing the resource number and type. The label font is set to Microsoft YaHei, size 12, and the same color as the point marker to ensure visual consistency. The adapted layer data is displayed in real time by the front-end visualization module using the WebGL rendering engine, creating a "dynamically rendered resource distribution view." To ensure the real-time performance of the dynamic view, a refresh control mechanism is implemented. This mechanism is implemented using a scheduled refresh tool with a default refresh cycle of 30 seconds. Before each cycle is triggered, the resource status comparison module executes, comparing the current resource data with the previous data snapshot. Comparison fields include location, type, and density level. If the number of changed records exceeds 10% of the total number of resource records, the change rate is considered excessive, the current refresh wait cycle is interrupted, and a full layer refresh is immediately performed. The 10% threshold was determined based on a seven-day statistical test of resource fluctuations in 30 sample areas before the platform's actual deployment. This threshold can be defined in the configuration file and defaults to 0.1. Under this threshold, the refresh frequency is controlled to no more than twice per minute, and the average CPU and GPU load does not exceed 60%, effectively ensuring stability and responsiveness. After the instant update is completed, the adjusted and redrawn layer will be returned to the user interface as the "final resource distribution view". This view contains the distribution location, usage status and classification labels of all currently valid resources for further analysis by the allocation engine.

[0026] S4 includes obtaining relevant information on resource frequency and status monitoring from a dynamically updated resource distribution view, classifying and arranging the information using a data extraction tool, categorizing the data on resource status and abnormal nodes, and obtaining a classified resource status set; based on the classified resource status set, using a comparative analysis tool to compare the data on abnormal nodes and idle phenomena one by one; if the data on the abnormal node exceeds a preset threshold, it is marked as a potential uneven distribution point, and the marked abnormal distribution data set is determined; for the marked abnormal distribution data set, obtaining relevant content on node distribution and real-time data, using a data mapping tool to correspond the data set with the content presented in the distribution, and obtaining a corresponding node adjustment directory; through the corresponding node adjustment directory, combined with the requirements of monitoring frequency and view update, using a timed refresh tool to periodically check the nodes in the directory; if it is found that the resource status is unevenly distributed, the relevant nodes are prioritized to determine the resource node directory that ultimately needs to be adjusted. In this implementation, first, at the beginning of each resource status refresh cycle, the real-time status data of all available teaching resources is extracted from the dynamic resource distribution view. Each data record contains at least five fields: resource number, number of times used in the past 30 minutes, cumulative number of days not used, average length of a single-day usage period, and current active status identifier. After reading the above fields using the data extraction tool, the records are classified according to the resource number, and each resource is classified into the categories of "normal use", "abnormal use" or "long-term idleness" through a logical judgment function. The classification criteria are: if the number of times a resource has been used in the past 30 minutes is 0, and it has not been used for no less than 10 consecutive days, it is judged as "long-term idleness"; if the number of times a resource has been used in the past 30 minutes is less than 1, and the absolute value of the deviation value compared with the average usage frequency of resources of the same category in the area exceeds 40%, it is judged as "abnormal use"; other cases are classified as "normal use". Deviation is calculated by filtering all records of the same resource type from the resource collection within the same administrative region and calculating the arithmetic mean of their 30-minute average usage times. The average usage of the current resource is then subtracted from this average and divided by the average to obtain the percentage deviation, rounded to two decimal places. If this value is greater than or equal to positive 40%, it is marked as over-utilized; if it is less than or equal to negative 40%, it is marked as under-utilized. This 40% deviation threshold is derived from data collected from seven consecutive days of monitoring of 4,000 resource samples across 10 prefecture-level cities before the platform was deployed. Distribution density analysis confirmed that this threshold accurately identifies resources with utilization significantly deviating from the average. After the classification process is completed, resources marked as "abnormally used" and "long-term idle" are added to the "abnormal node temporary list" and "idle node temporary list," respectively, and merged to form the "abnormal distribution dataset." Each record in this dataset contains fields including the resource ID, actual usage times, average usage times, deviation value, number of consecutive days of non-use, geographic coordinates, resource type, status category, and generation timestamp.After the dataset is generated, the data mapping module is invoked to match the resource ID in each dataset record with the layer ID field in the resource distribution layer. This information is used to obtain the actual rendering location of each resource on the layer and to supplement its layer index, annotation label, and layer zoom level fields. All mapped records are output as a "node adjustment catalog." This catalog is then fed into the scheduled refresh verification module. The refresh module uses a 30-minute refresh cycle, configured uniformly by the platform's operational maintenance policy, and executes node recheck logic once per cycle. First, all resources in the catalog are traversed, their latest monitoring data is retrieved, and their deviation from the regional mean and the cumulative number of days unused are recalculated. If any metric still meets the abnormality criteria—i.e., the deviation value still exceeds 40% or the number of days unused remains at least 10—the resource remains in the "resources requiring verification list" with its status unchanged. Otherwise, it is removed from the catalog. After verification, a priority score is applied to the resources remaining in the catalog. The scoring model uses three evaluation dimensions: First, the frequency of use deviation score: 100 points for deviations exceeding 60%, 80 points for 50% to 60%, and 60 points for 40% to 50%; second, the unused time score: 100 points for unused resources exceeding 15 days, and 80 points for 10 to 15 days; and third, the regional stress score: This score is calculated based on the average utilization of all resources of the same type in the same region. If the average utilization of a resource of the same type in the current region is 20% higher than the global average, the region is considered resource-strained and scored 100 points. If the utilization is between 10% and 20%, it is scored 80 points, and if it is less than 10%, it is scored 60 points. The weighted average of the three scoring results is calculated using a weighting ratio of 4:3:3, with a total score ranging from 0 to 100. Resources with a final score of 70 or higher are marked as "priority resources for adjustment" and included in the "Resource Node Directory for Final Adjustment." The above scoring weights were derived through simulated scheduling experiments. In a simulation of resource allocation across 50 regions, this scoring method improved resource balancing by over 8%. A score of 70 was empirically determined to be the minimum possible allocation threshold; allocations below this threshold were ineffective. Ultimately, this directory serves as direct input to the scheduling module, enabling the next stage of path optimization and resource allocation instruction generation based on its node information. This ensures that resource allocation, based on status monitoring and logical judgment, has a complete, accurate, and dynamically updated basis.

[0027] S5 includes relevant data divided according to geographic location and region, and uses map data processing tools to hierarchically organize the node distribution in the target area, and matches the geographic location information of the nodes with the resource ownership to obtain a hierarchical node distribution data set; through the hierarchical node distribution data set, combined with data on resource demand and node priority, a path calculation tool is used to compare the path length and delivery time from the starting node to the target node to determine a preliminary path planning scheme; if the path length or delivery time in the preliminary path planning scheme exceeds the preset threshold, a dynamic adjustment tool is used to combine real-time monitoring data and transportation cost information to locally optimize the path and obtain an adjusted path plan; for the adjusted path plan, relevant indicators of allocation efficiency are obtained, and data comparison tools are used for comparison and analysis to determine the final resource reallocation path plan.

[0028] In this implementation, the longitude and latitude coordinates of each teaching resource node and its administrative division code are first called, and the standardized geographic boundary layer data is loaded. The map data processing tool takes the coordinates of the resource node as input, determines the administrative area boundary to which each node belongs through the spatial containment relationship, and then performs multi-level sorting according to the administrative division level and the functional classification field of the resource. The specific processing process is to first divide the resources into the districts and counties according to the administrative division code, and then group them according to the resource's affiliated unit field within the district and county, and then perform the final stratification according to the resource use field. In the output "layered node distribution data set", each record contains fields such as resource number, resource type, geographic coordinates, administrative division code, affiliated unit code, node stratification identifier, etc., totaling no less than 6 items.

[0029] Next, new resource requirements are extracted from the dispatch request task queue. Each requirement record contains the administrative code of the target area, the required resource type, the expected receipt time, and the dispatch urgency level. Based on the type of requirement, a set of source nodes with matching resource types and a dispatchable status are selected from the aforementioned hierarchical node data set. For each source node, sort from high to low according to the priority score field in step S4, and select the top three source nodes as alternative dispatch sources. Then, use the path calculation tool to calculate the shortest path length and estimated delivery time from each source node to the center point of the target area. The path length is based on the shortest driving distance returned by the map API interface, in meters; the delivery time is the path length divided by the average real-time driving speed corresponding to the path, in minutes. The speed is determined by the real-time road speed provided by the traffic monitoring sub-unit, in kilometers per hour, and converted to meters per minute for calculation.

[0030] The acceptable threshold for route length is set at 20,000 meters, and the acceptable threshold for delivery time is set at 30 minutes. These two parameters are fixed configuration items, based on the 85th percentile of dispatch records extracted from 10 typical cities nationwide on normal working days, covering most urban resource dispatch scenarios. If a route's length or time exceeds either threshold, it is rejected and enters the route optimization process. Route optimization is performed by a dynamic adjustment tool. It first uses the real-time traffic data module to obtain the current speed of each section of the original route and identify any high-congestion sections with speeds below 15 kilometers per hour. If so, the route is replanned to exclude these sections and the new route length and time are calculated. The adjusted route must reduce delivery time by at least 20% and increase transportation costs by no more than 10%. The transportation cost is calculated as the route length divided by 1000 and multiplied by the unit price of 2 yuan per kilometer. If the new route meets both of these conditions, it is replaced with the original route and marked as an "optimized route solution." For each adjusted route plan, dispatch efficiency metrics are then calculated, including actual delivery time, transportation cost, and dispatch timeliness. Average values ​​for similar scenarios are retrieved from a historical database as a baseline for comparison. If the current route plan outperforms the historical average in at least two of these three metrics, the route is marked as the "final resource reallocation route." The judgment logic is: if the delivery time is at least 2 minutes faster than the historical average, or the transportation cost is at least 5 yuan lower than the average, or the dispatch timeliness is at least 5 percentage points higher than the historical average, it is considered superior. The final route plan is recorded and submitted to the dispatch scheduling engine as part of the dispatch execution plan. All of the above route parameters, time metrics, and thresholds are set as adjustable parameters in the configuration template and are supported by clear data sources. The 20,000-meter route length threshold is based on the maximum conventional dispatch distance for urban resources. The 30-minute delivery time is the upper limit of acceptable cross-regional dispatch in most cities. 15 kilometers per hour is defined as the critical speed for urban road congestion. The 20% time optimization threshold is the minimum percentage required to ensure that route optimization generates substantial benefits. The 10% cost limit is the platform's maximum tolerance for dynamic optimization cost fluctuations. All parameters are set uniformly by the administrator, and support customized configuration by city and administrative level.

[0031] S6 includes obtaining the mapping relationship between execution parameters and regional divisions from a pre-established repository based on the relevant data of the path plan and dynamic allocation, and using a data sorting tool to hierarchically process the parameters in order to facilitate operation, thereby obtaining a hierarchical parameter set; using the hierarchical parameter set, in combination with real-time monitoring data and priority sorting basis, using a matching tool to initially divide the allocation time and allocation scope, and determine the initial time and scope plan; if the initial time and scope plan does not meet the demand matching standard, then using a path adjustment tool, in combination with the real-time status of resource allocation, partially correct the plan and determine the corrected time and scope match; for the corrected time and scope match, obtaining the generation rules of the automation instruction, and using the instruction construction tool to integrate the match and dynamic allocation conditions to generate the final automation allocation instruction.

[0032] In this implementation, after the path plan is determined, parameter data matching the current deployment area is first extracted from the configured execution parameter repository. This repository uses the region code as the primary search key. Each record contains eight fields: task type, deployment level, administrative division number, default deployment time period, default deployment radius, deployment trigger method, minimum execution response time, and maximum execution wait time, totaling eight items. A data organization tool is then used to group the extracted parameter set into three layers: the first layer is based on the task type field, such as general teaching deployment, emergency resource deployment, and temporary supplementary deployment; the second layer is based on administrative division level, such as prefecture-level cities, counties, and sub-districts; and the third layer is based on resource level, such as first-level resources being in short supply and second-level resources being common resources. Once the stratification is complete, a "stratified parameter set" is generated, which serves as the basis for setting task parameters. Subsequently, the initial deployment time and range division is performed, combining the target node's real-time status information, including whether the resource is idle, available time period, and whether it is currently locked by other tasks. Time partitioning is implemented by rolling backward in 15-minute increments, starting with the current time. The search is performed to find the first consecutive idle time window within the target node's free time period that is at least equal to the required allocation duration. The allocation duration is determined by the resource type: standard classroom resources are set to be at least 45 minutes, multimedia resources are set to be at least 60 minutes, and specialized resources such as laboratories are set to be at least 90 minutes. If no matching time window is found for the target node within six backward rollbacks (i.e., 90 minutes), the node is considered time mismatched. Simultaneously with time partitioning, spatial range partitioning is performed. A circular buffer with an initial radius of 3000 meters is constructed, centered at the target node's latitude and longitude. The radius is taken from the "Default Allocation Radius" field in the hierarchical parameter set. All candidate source nodes are traversed, and the spherical distance between them and the target node is calculated using a simplified spherical trigonometry method in meters. All source nodes within the buffer are retained. If the number of results is less than three, the range is considered insufficient for allocation. If any of the conditions is not met, that is, there is no available time window or insufficient source nodes, the path adjustment process will be initiated. First, the search radius is increased by 20% of the original radius, that is, from 3,000 meters to 3,600 meters. This threshold is taken from the statistical results of urban resource distribution density. In most urban contexts, 3,600 meters covers more than 90% of school nodes. Then shorten the rolling time granularity from 15 minutes to 10 minutes, and extend the search cycle from 6 times to 12 times, that is, the maximum search range is 120 minutes. By expanding the spatial radius and time window granularity, try to generate new feasible combinations. If there are multiple combinations, the solution with the earliest time and the shortest source node distance is selected as the "corrected time and range combination." After completing the time and range combination, start the automatic allocation instruction generation module.Extract the field structure from the standard instruction template, including task number, execution time, starting node number, target node number, path number, allocation resource type, execution priority, receipt interface address, etc., a total of 8 fields. Fill in the corrected parameters into the corresponding fields one by one to build a complete automated execution command. The command is encapsulated in JSON format, and the fields are all standardized fields. The format is uniformly defined by the platform interface protocol document to ensure consistency across calls. After the construction is completed, the integrity of the instruction structure is checked. If there are no missing fields and the time field format is legal, the instruction will be pushed to the scheduling engine for execution. All key parameters are generated by a large amount of historical allocation data and actual measurement statistics in the early stage of platform operation. 3,000 meters is the standard allocation radius for urban areas, derived from an analysis of the maximum coverage radius of daily cross-school resource allocation in six cities. 45, 60, and 90 minutes are the minimum usage periods for common teaching resources, determined based on course length standards in the academic affairs department. Rolling intervals of 15 and 10 minutes are the platform's minimum task scheduling cycles, synchronized with the scheduling thread's processing capacity and the pace of teacher feedback. A minimum of three source nodes is the selection threshold, ensuring redundancy and fault tolerance in path selection. All of these settings are fixed in the configuration table and can be adjusted as needed by administrators.

[0033] S7 includes obtaining dynamic status information of resource nodes from a pre-established monitoring platform through automated instructions, performing preliminary sorting of node change characteristics to obtain an initial status data set; analyzing the distribution presentation of resource nodes using a data comparison tool based on the initial status data set; triggering a distribution presentation update process and determining an update priority sequence if node changes exceed a preset threshold; after obtaining the priority sequence, locally adjusting the distribution presentation based on the current status of the resource distribution using a presentation layer generation tool to obtain updated presentation layer data; classifying the adjusted status of the resource distribution using a support vector machine algorithm based on the updated presentation layer data in combination with real-time feedback of the dynamic status to determine whether the classified distribution meets the requirements of the allocation instruction; if the classified distribution meets the requirements of the allocation instruction, integrating the adjusted status with the distribution record using a status recording tool to obtain a final status record file; generating an update log of the distribution record based on the final status record file based on the dynamic status changes of the resource nodes using a data mapping tool to determine the latest status presentation of the resource distribution; and continuously calibrating the resource distribution presentation layer using the update log in combination with the real-time monitoring data stream to obtain a calibrated distribution status record.

[0034] In this implementation, based on the generated automated deployment instructions, dynamic state data of resource nodes related to the current deployment is first obtained in real time from the central resource monitoring platform via a pre-set monitoring interface. This data includes at least six fields, including the node's unique code, state type (idle, occupied, faulty), last change timestamp, geographic coordinates, region code, and bound task number. All state information is stored in a state cache, and a structured state table is constructed with the node code as the primary key. Next, the number of state changes for each node within a 5-minute window is counted. If a node experiences two or more state changes, it is marked as a node with high state volatility. This data is then combined into an "initial state dataset." A structured comparison tool is then used to compare the current initial state dataset with the previous layer state record, focusing on changes to the resource state and location fields. The number of nodes experiencing state changes is compared and divided by the total number of monitored nodes to calculate the state change ratio. If the state change ratio is greater than 0.1, meaning that the percentage of nodes experiencing state changes exceeds 10%, the resource layer is deemed to require an update. This threshold was determined based on an analysis of historical deployment records during the platform's deployment phase. In over 90% of normal deployments, the percentage of nodes with status changes was less than 10%. Therefore, 10% was set as the default minimum threshold for triggering changes. Once the layer update conditions are met, all nodes with status changes are scored according to the following criteria: if the changed node is located in the core teaching resource area and its status changes from "idle" to "occupied," it receives a score of 5; if its status changes from "idle" to "faulted," it receives a score of 4; if it is located in a non-core area, it receives a score of 3; and all other changes receive a score of 2. An "update priority sequence" is generated based on the scores, sorted from high to low. Subsequently, the layer grid area containing the top 10 scoring nodes is located, with a fixed grid length of 1 kilometer. The layer rendering engine is invoked to perform a local layer refresh on the target area, outputting the "updated layer dataset." The platform then loads the updated layer data and all status records collected within the past five minutes, using a support vector machine algorithm to classify the status features. After loading the updated layer data and all resource node status records collected within the past five minutes, the platform first extracts the feature variables used for classification for each status record. There are four characteristic variables in total, namely: state change frequency, that is, the number of times the node changes its state in the past 5 minutes, with a value range of 0 to 5 times; state switching path, that is, the type of transition of the node from the previous state to the current state, such as "idle to occupied", "idle to faulty", "occupied to idle", etc., which is quantified using a set of fixed codes; geographic location change amplitude, that is, the change value of the latitude and longitude coordinates of the node during the state change process, calculating the spherical distance between its previous and next coordinates, in meters; time interval from the last state change to the current moment, in seconds.The above features are sequenced to construct a feature vector, which is then fed into a trained support vector machine model for classification prediction. The model uses a radial basis function kernel. The training dataset is based on 10,000 real-world state change data from historical deployment tasks collected before the platform went live. The classification label is "whether deployment instructions are met," with a value of "yes" or "no." Each node state record is independently predicted, and the output is a binary classification. A "yes" prediction indicates that the node's current state can support the normal execution of the current deployment instruction. The classification results for all nodes are statistically summarized. If the proportion of nodes predicted as "yes" accounts for more than 90% of the total number of nodes involved in the classification, the current layer state is considered to meet the deployment objectives, thus confirming that the current state adjustment is valid and no further re-deployment is required. Training features include state change frequency, state transition path, geographic location change magnitude, and the time interval between the last state change and the current state. The trained support vector machine model is used to perform classification predictions on each node record, outputting whether the node "meets deployment instructions." If more than 90% of the predicted nodes are judged to "meet deployment requirements," the current state adjustment is considered valid. Afterward, the status recording tool is called to generate a "final status record file," which includes fields such as the node number, information before and after the status change, the change time, the assigned deployment task number, and a confirmation flag. Based on this record file, the data mapping module writes the final status of each node to the historical status trajectory table in the resource database, simultaneously generating an "update log." This log is automatically generated every 30 seconds, recording all nodes with status changes, including the change time, previous and next status information, and the grid number in which they are located. The log is formatted as structured JSON. Finally, a calibration operation is performed based on this log and the current monitoring data stream. Calibration is performed every 60 seconds. The node number is used as the primary key, and the latest status in the log is compared with the record in the current layer cache. If a timestamp difference of less than 30 seconds indicates a state inconsistency, a local layer refresh is performed. This process ensures that the layer state is always consistent with the monitoring data. The resulting calibration output is a "calibrated distribution state record" and stored in the resource platform view history table for subsequent analysis. The above parameters were determined during the platform's initial development phase based on historical deployment data and operational load testing results: the 10% state change threshold represents the minimum threshold for affecting deployment plan stability; scores of 5, 4, 3, and 2 are derived from expert-calibrated node impact levels; the 90% classification judgment value for the support vector machine model is derived from cross-validation accuracy analysis; a 1km grid edge length for layer updates represents the optimal computational unit for balancing accuracy and performance; a 30-second state recording interval and a 60-second calibration cycle represent the minimum task window for platform scheduling threads. These parameters are clearly defined in the configuration table and can be adjusted dynamically by administrators based on policy requirements or load.

[0035] S8 includes obtaining the latest data stream of resource usage frequency from a pre-established monitoring platform according to the adjusted resource distribution status record, performing preliminary sorting on the characteristics of frequency changes, and obtaining a summary data set of frequency changes; integrating the summary data set with the current state of resource distribution using a data fusion tool, extracting characteristics based on the demand for distribution adjustment during integration, and determining a fused distribution feature data set; recalculating the balance of resource distribution using a data comparison tool based on the distribution feature data set, and if the calculation result does not reach the preset distribution balance threshold, triggering the balance adjustment process and judging the adjusted balance distribution state; based on the adjusted balance distribution state, combining the real-time data stream, using the state comparison tool to continuously verify the resource distribution and the distribution balance target, and obtaining the final distribution balance record. In this embodiment, first, by retrieving the latest distribution status record from the resource platform, extracting all resource node codes involved in this round of dynamic allocation, and obtaining the usage frequency data of these nodes in the past 5 minutes from the resource monitoring platform in real time, specifically recording the state once per minute, with the state "occupied" counted as 1 and the state "unoccupied" counted as 0, and finally obtaining a binary sequence of 5 time points. The sequence for each node is summed and divided by 5 to obtain the node's average usage frequency during that time period, with a value ranging from 0 to 1. For example, if a node's sequence is "1, 1, 0, 1, 1," its frequency value is 0.8. The frequency values ​​of all nodes form a "frequency change summary dataset," sorted by node number. The historical scheduling records and current distribution status of the nodes are then synchronized and mapped to fields such as the node's region, resource type, current status, and task binding relationship. This fused "distribution feature dataset" is then generated. Next, the nodes are divided into several subsets based on region code. Within each region, the average frequency of all nodes is calculated, and its standard deviation is calculated. The standard deviation is calculated by taking the square root of the sum of the squared differences between the frequency of each node and the regional average. The ratio of the standard deviation to the average is calculated for each region to assess the distribution balance within that region. This ratio is called the "distribution fluctuation coefficient." If the distribution fluctuation coefficient for a region exceeds 0.5, it indicates severe resource inequality within the region, and this value serves as the pre-set warning threshold. This threshold is based on the platform's analysis of the balance of historical usage data across different types of regions across the country. Under normal deployment conditions, the fluctuation coefficient of more than 95% of regions is less than 0.5, so this value is set as the default judgment standard. Once the fluctuation coefficient of a region is detected to exceed 0.5, the balance adjustment process is immediately triggered. Nodes with a frequency value higher than 0.2 of the regional average are prioritized and marked as "high-load nodes." Nodes with a frequency value lower than 0.2 of the regional average are simultaneously marked as "low-load nodes."The threshold of 0.2 is derived from an analysis of the fluctuation range of 3,000 sets of sample task frequency change data. The results show that when the frequency difference exceeds 0.2, the efficiency of resource allocation between nodes is significantly affected. An allocation path matching matrix is ​​constructed based on the distance between nodes, resource compatibility, and historical task conflicts. Path reconstruction is performed for high- and low-load node pairs within a distance of 20 kilometers and without mutually bound tasks in the past 24 hours, and allocation path plans are generated first. After the initial path allocation is completed, the latest frequency data stream is collected again for the readjusted resource status, and the standard deviation to mean ratio is recalculated. If the calculation results show that the deviation between the node frequency and the target frequency of their area for more than 90% is within 5%, it means that the allocation has achieved the preset balance target. This 90% ratio is the platform allocation execution stability assessment standard. It is set based on the analysis of the closed-loop efficiency of past tasks during the operation phase and can effectively balance the algorithm convergence speed and platform resource stability. Ultimately, all calculation results, node change data, regional balance indicators, and the proportion of successfully deployed nodes are combined to form a "final distribution balance record" and stored in the platform's historical database for subsequent strategy analysis and scheduling parameter optimization. All thresholds and calculation rules used are clearly defined in the parameter configuration table and dynamically maintained and adjusted by operations administrators based on actual operating conditions and policy requirements.

[0036] S9 includes obtaining the current distribution state data stream of the resource allocation from a pre-established monitoring platform, comparing the distribution state with a preset balance threshold, determining whether there is a deviation area, and obtaining the deviation distribution record; if the deviation distribution record shows that the preset balance threshold is not reached, re-planning the allocation instructions through the path optimization tool, generating a new instruction data set in combination with a dynamic mechanism, determining whether the instructions cover all deviation areas, and obtaining an adjusted instruction set; based on the adjusted instruction set, using the resource update tool to adjust the distribution state in real time, generating a new distribution view for the adjusted state, and obtaining the updated view data; using the updated view data, using the view update tool to verify the optimization results, and combining the real-time adjusted data stream to determine the final resource distribution view, and obtaining a distribution state record that meets the balance threshold. In this embodiment, the current distribution state data stream of the resource allocation is first obtained in real time through the monitoring platform interface. The data stream includes the current state, geographical location, usage frequency, regional affiliation, and task execution record of all resource nodes. Nodes are grouped by regional code, and the usage frequency of nodes in each region is counted. For each set of regional data, the average frequency of all nodes is first calculated. This is done by adding the frequency values ​​of all nodes in the region and dividing it by the number of nodes. The average of the squared differences between the frequencies of all nodes and the regional average is then calculated. Finally, the square root of this average is taken to obtain the standard deviation of the frequency for that region. Dividing the standard deviation by the average yields the regional allocation deviation coefficient, which is used to measure the balance of resource usage frequency. If the deviation coefficient for a region is greater than 0.5, it indicates an imbalance in resource usage frequency distribution. The region is marked as a "deviation region" and recorded in the "deviation distribution record." The 0.5 threshold is a reasonable upper limit determined by the platform's statistical evaluation of resource allocation effectiveness across multiple sample regions. Historically, over 95% of stable allocation regions have deviation coefficients below this value, so it serves as a fixed reference for determining balance. When a deviation region is detected, the allocation instruction reconstruction module is activated. This module first sorts the frequencies of all nodes within the deviation region, marking nodes with frequencies 0.2 above the regional average as high-load nodes and nodes with frequencies 0.2 below the regional average as low-load nodes. The path calculation module is then called to calculate feasible paths between each high-load node and a possible low-load node. Calculation parameters include path length, required time, traffic conditions, and dispatch costs. Path length is calculated by calculating the spherical distance between the two nodes' longitude and latitude. Transportation time is based on historical traffic data and current traffic conditions. Acceptable paths are defined as those with a length of no more than 20 kilometers and a transportation time of no more than 30 minutes. If the calculated path does not meet these requirements, the match is discarded.All paths and node matching relationships that meet the conditions are integrated into a new allocation instruction data set. The instruction fields include source node, target node, resource type, execution time window and allocation policy tag. The above allocation instruction data set is reviewed for integrity. If there are nodes in the deviation area that have not been covered by the allocation instruction, the instructions are supplemented, and finally a complete set of adjusted allocation instructions is formed. According to this set, the allocation operation is performed on each resource node in turn, and the resource status, belonging area and task binding relationship of the node are updated. The map layer data is synchronously updated by the layer rendering module to generate a new distribution view. The distribution view uses color scale to encode frequency difference. The node status is reflected by the change of icon color and shape. The new layer data is integrated with the real-time monitoring status to ensure that all displayed information is synchronized with the actual node operation status. After the view update is completed, the current area status is re-analyzed for the allocation deviation coefficient. The ratio of the frequency mean to the standard deviation is calculated for each region. If the deviation coefficient is less than 0.5 for more than 90% of the regions and there are no new deviation regions, the current allocation is considered effective and the current round of resource optimization is deemed complete. The final resource distribution view and detailed distribution status records are stored for subsequent strategy evaluation and operational analysis. This 90% judgment standard is derived from the platform's closed-loop quality analysis of the past 1,000 rounds of resource allocation results. The results show that when the deviation is controlled within this range, resource utilization efficiency and user satisfaction are optimized. All parameters are set by the platform's data governance team based on operational evaluation standards.

[0037] like Figure 2As shown, a dynamic allocation system for teaching resources based on educational software platform data is also provided, which is used to implement the steps of the dynamic allocation method for teaching resources based on educational software platform data. The system includes a location information acquisition module, which is used to obtain the geographical location information of educational resources and standardize the location data to build a structured spatial distribution framework; a data association module, which is used to obtain the logical relationship data of resources according to the spatial distribution framework, and associate and map the logical relationship data with the geographical location information to determine a multi-dimensional resource attribute set; a layer generation module, which is used to integrate the geographical location information and the logical relationship data based on the multi-dimensional resource attribute set, generate a resource distribution presentation layer, and form a dynamically updated resource distribution view; a status analysis module, which is used to obtain the resource usage frequency and real-time status monitoring data from the resource distribution view to determine whether there are idle resources or uneven distribution. phenomenon and determine the resource nodes that need to be adjusted; the path planning module is used to calculate the optimal allocation path and generate a path plan for resource reallocation when a resource node that needs to be adjusted is detected; the instruction generation module is used to obtain the execution parameters of the dynamic allocation mechanism according to the path plan, generate automatic allocation instructions, and determine the specific time and scope of allocation execution; the allocation execution module is used to execute resource allocation operations according to the automatic allocation instructions, and obtain updated real-time status monitoring data, update the resource distribution presentation layer, and form an adjusted resource distribution status record; the analysis feedback module is used to obtain the latest data on resource usage frequency based on the adjusted resource distribution status record, analyze the balance of resource distribution, and determine whether the preset allocation balance threshold is reached; the optimization iteration module is used to regenerate allocation instructions and update the resource distribution view when the allocation balance threshold is not reached, and output the final resource optimization distribution result.

[0038] This system is based on a modular design, with functional modules working together to achieve closed-loop management of the dynamic allocation of educational resources. The location information acquisition module, accessing the educational resource database, the geographic information database, and the positioning interface, collects the geographic coordinates of each resource unit at a set interval. After standardization, the module formats the information in a unified format and constructs a spatial distribution framework based on longitude and latitude, with a grid length of 1 kilometer. The data association module uses this distribution framework as an index to retrieve multidimensional data, including logical classification information and usage frequency, for resources in the corresponding area. Field matching is performed using cleansing rules to form a structured resource attribute table. The layer generation module uses this attribute table as the foundation for layer rendering. Through the data visualization interface, it generates map layers, identifying resource nodes, status popularity, and distribution density, and maintains a refreshable layer model on the backend. The status analysis module monitors resource usage fluctuations in the distribution layer in real time. By comparing the current status with usage records from the past 24 hours, it calculates the standard deviation and mean difference in each node's utilization rate, and uses this information to determine whether there are idle or overloaded nodes. When resource allocation in a region is found to be uneven or significantly idle, the path planning module activates, calculates path loss weights, and prioritizes the path construction plan with the lowest cost and fastest deployment time. The instruction generation module then reads this path plan and the system's predefined parameter table to automatically generate the deployment task's execution parameters, including resource quantity, start and end nodes, deployment method, time window, and execution priority. These parameters are then encapsulated in a structured format as a deployment instruction set. Based on this instruction set, the deployment execution module issues control instructions for the actual resource status on the platform. Changes in resource status are fed back to the layer system in real time, generating a record of the adjusted resource distribution. The analysis and feedback module uses this record as a basis for retrieving the latest 5-minute frequency data, performing a regional balance analysis, and calculating the allocation deviation coefficient. If the analysis results indicate that the deviation coefficient for any region exceeds the system-set threshold of 0.5, the deployment is deemed to be incompletely balanced and enters the optimization iteration module. This module continues to execute compensatory allocations until the system determines that the allocation deviation coefficient for more than 90% of regions is below 0.5. Finally, it outputs a resource distribution view and an optimization results report.

[0039] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for dynamically allocating teaching resources based on educational software platform data, characterized in that: include: S1. Obtain the geographic location information of educational resources, standardize the location data, and obtain a structured spatial distribution framework; S2. Obtain logical relationship data based on the spatial distribution framework, associate and map the logical relationship data with geographic location information, and determine a multi-dimensional resource attribute set; S3. For a multi-dimensional resource attribute set, geographic location information and logical relationship data are integrated into a resource distribution presentation layer to obtain a dynamically updated resource distribution view; S4. Obtain resource usage frequency and real-time status monitoring data from the dynamically updated resource distribution view to determine whether there is idle or uneven resource allocation, and identify resource nodes that need adjustment; S5. If a resource node that needs to be adjusted is detected, the optimal allocation path is calculated to obtain a path plan for resource reallocation; S6. Based on the resource reallocation path plan, obtain the execution parameters of the dynamic allocation mechanism, generate automated allocation instructions, and determine the specific time and scope of allocation execution; S7. Obtain real-time status monitoring data through automated deployment instructions, update the resource distribution presentation layer, and obtain the adjusted resource distribution status record; S8. Based on the adjusted resource distribution status record, obtain the latest data on resource usage frequency, re-analyze the balance of resource distribution, and determine whether the preset allocation balance threshold is reached; S9. If the preset allocation balance threshold is not reached, regenerate the allocation instruction, update the resource distribution view, and obtain the final resource optimization distribution result.

2. The method for dynamically allocating teaching resources based on educational software platform data according to claim 1, characterized in that: Said S1 includes obtaining geographical location records of educational resources through a pre-established geographical information database, performing preliminary cleaning on the geographical location records, and using common tools to fill in and remove missing values ​​and outliers in the records to obtain a sorted location information set; Based on the collated location information set, spatial attribute analysis tools are used to standardize the location information set. Data formats from different sources are uniformly converted. If the data format does not meet the preset threshold, it is remapped to determine a standardized location data set. The standardized location data set is combined with location analysis tools to mine the spatial distribution patterns of the data set, obtain the spatial aggregation characteristics and distribution differences in the data set, and determine the geographical distribution hotspots of educational resources. According to the geographical distribution hotspots, data integration tools are used to build a structured spatial distribution framework, and multi-dimensional fusion is performed on the distribution data set to obtain the final geographical distribution data set.

3. The method for dynamically allocating teaching resources based on educational software platform data according to claim 1, characterized in that: S2 includes obtaining classification information of educational resources from pre-established resource attribution classification data, performing preliminary matching between the classification information and geographic location information, and using a data fusion tool to associate the attribution classification data with the location information to obtain a preliminary resource attribute set; obtaining relevant records of resource usage frequency through the preliminary resource attribute set, cleaning incomplete data in the records, and filling in the missing parts using a general data completion tool to determine a complete resource usage frequency data set; If the value in the complete resource usage frequency data set exceeds the preset threshold, the data standardization tool is used to adjust the range of the data set, and the adjusted data is secondary integrated with the preliminary resource attribute set to determine the multi-dimensional resource attribute association set; based on the multi-dimensional resource attribute association set, the data mapping tool is used to integrate the resource ownership classification, usage frequency and geographic location information to obtain the final multi-dimensional resource attribute set.

4. The method for dynamically allocating teaching resources based on educational software platform data according to claim 1, characterized in that: The S3 includes using a data mapping tool to perform preliminary correspondence processing on the resource distribution and geographic coordinate information based on pre-established resource distribution data and geographic coordinate information, integrating the correspondence data to obtain a preliminary integrated distribution data set; using the preliminary integrated distribution data set, using a visual drawing tool to match the distribution data set and the distribution layer; if there is an inconsistency between the location mark and the layer information, adjusting the inconsistent part to determine the adjusted distribution layer data set; obtaining the adjusted distribution layer data set, using an interface rendering tool to adapt the distribution layer data set to the interface display requirements, and dynamically presenting the adapted data to obtain a dynamically presented resource distribution view; for the dynamically presented resource distribution view, using a timed refresh tool in combination with update frequency data to periodically collect and process the view update content; if the update frequency exceeds a preset threshold, the view content is immediately adjusted to determine the final resource distribution view.

5. The method for dynamically allocating teaching resources based on educational software platform data according to claim 1, characterized in that: The S4 includes obtaining relevant information on resource frequency and status monitoring from the dynamically updated resource distribution view, classifying and arranging the information using a data extraction tool, categorizing the resource status and abnormal node data into different categories, and obtaining a classified resource status set; using a comparative analysis tool to compare the abnormal node and idle phenomenon data one by one based on the classified resource status set; if the data of the abnormal node exceeds a preset threshold, it is marked as a potential uneven distribution point, and a marked abnormal distribution data set is determined; for the marked abnormal distribution data set, relevant content of the node distribution and real-time data is obtained, and a data mapping tool is used to match the data set with the content presented by the distribution to obtain a corresponding node adjustment directory; Through the corresponding node adjustment directory, combined with the monitoring frequency and view update requirements, a timed refresh tool is used to periodically check the nodes in the directory. If it is found that the resource status is unevenly distributed, the relevant nodes are prioritized to determine the resource node directory that ultimately needs to be adjusted.

6. The method for dynamically allocating teaching resources based on educational software platform data according to claim 1, characterized in that: S5 includes using a map data processing tool to hierarchically organize the node distribution within the target area based on the relevant data of geographic location and regional division, matching the geographic location information of the nodes with the resource ownership, and obtaining a hierarchical node distribution data set; using the hierarchical node distribution data set, combined with the data of resource demand and node priority, using a path calculation tool to compare the path length and delivery time from the starting node to the target node to determine a preliminary path planning solution; If the route length or delivery time in the initial route planning solution exceeds the preset threshold, a dynamic adjustment tool is used to combine real-time monitoring data and transportation cost information to locally optimize the route and obtain an adjusted route solution; For the adjusted path plan, obtain relevant indicators of allocation efficiency, use data comparison tools to perform comparative analysis, and determine the final resource reallocation path plan.

7. The method for dynamically allocating teaching resources based on educational software platform data according to claim 1, characterized in that: S6 includes obtaining a mapping relationship between execution parameters and regional divisions from a pre-established repository based on the path plan and the relevant data of the dynamic allocation, and hierarchically processing the parameters using a data sorting tool to obtain a hierarchical parameter set for ease of operation; Through the layered parameter set, combined with real-time monitoring data and priority sorting basis, the matching tool is used to initially divide the allocation time and allocation scope to determine the initial time and scope plan; if the initial time and scope plan does not meet the demand matching standards, the path adjustment tool is used to combine the real-time status of resource allocation to partially correct the plan and determine the revised time and scope combination; for the revised time and scope combination, the generation rules of the automated instructions are obtained, and the instruction construction tool is used to integrate the combination with the dynamic allocation conditions to generate the final automated allocation instructions.

8. The method for dynamically allocating teaching resources based on educational software platform data according to claim 1, characterized in that: The S7 includes obtaining dynamic status information of resource nodes from a pre-established monitoring platform through automated instructions, and performing preliminary sorting on the characteristics of node changes to obtain an initial status data set; based on the initial status data set, using a data comparison tool to analyze the distribution presentation of resource nodes; if the node change exceeds a preset threshold, triggering the distribution presentation update process, and determining the updated priority sequence; after obtaining the priority sequence, using a presentation layer generation tool to locally adjust the distribution presentation based on the current status of the resource distribution to obtain updated presentation layer data; using the updated presentation layer data, combined with real-time feedback on the dynamic status, using a support vector machine algorithm to classify the adjustment status of the resource distribution to determine whether the classified distribution meets the requirements of the deployment instruction; if the classified distribution meets the requirements of the deployment instruction, integrating the adjustment status with the distribution record through a status recording tool to obtain a final status record file; Based on the final status record file, a data mapping tool is used to generate an update log of the distribution record for the dynamic status changes of the resource nodes to determine the latest status presentation of the resource distribution; through the update log, combined with the real-time monitoring data stream, the resource distribution presentation layer is continuously calibrated to obtain the calibrated distribution status record.

9. The method for dynamically allocating teaching resources based on educational software platform data according to claim 1, characterized in that: The S8 includes obtaining the latest data stream of resource usage frequency from a pre-established monitoring platform based on the adjusted resource distribution status record, and preliminarily sorting out the characteristics of frequency changes to obtain a summary data set of frequency changes; using a data fusion tool to integrate the summary data set with the current status of resource distribution, extracting features based on the distribution adjustment requirements during integration, and determining a fused distribution feature data set; using the distribution feature data set, using a data comparison tool to recalculate the balance of resource distribution; if the calculation result does not reach a preset distribution balance threshold, the balance adjustment process is triggered to determine the adjusted balance distribution status; based on the adjusted balance distribution status, combined with the real-time data stream, a status comparison tool is used to continuously verify the resource distribution and the distribution balance target to obtain a final distribution balance record.

10. A system for dynamically allocating teaching resources based on educational software platform data, for implementing the steps of the method for dynamically allocating teaching resources based on educational software platform data according to any one of claims 1 to 9, characterized in that: The system comprises: The location information acquisition module is used to obtain the geographical location information of educational resources and standardize the location data to build a structured spatial distribution framework; A data association module is used to obtain logical relationship data of resources according to the spatial distribution framework, and associate and map the logical relationship data with the geographical location information to determine a multi-dimensional resource attribute set; A layer generation module is used to integrate the geographic location information with the logical relationship data based on a multi-dimensional resource attribute set to generate a resource distribution presentation layer and form a dynamically updated resource distribution view; The status analysis module is used to obtain resource usage frequency and real-time status monitoring data from the resource distribution view, determine whether there is idle resource or uneven resource allocation, and identify resource nodes that need adjustment; The path planning module is used to calculate the optimal allocation path and generate a path plan for resource reallocation when a resource node that needs to be adjusted is detected; The instruction generation module is used to obtain the execution parameters of the dynamic allocation mechanism according to the path plan, generate automatic allocation instructions, and determine the specific time and scope of allocation execution; The deployment execution module is used to execute resource deployment operations according to the automated deployment instructions, obtain updated real-time status monitoring data, update the resource distribution presentation layer, and form a record of the adjusted resource distribution status; An analysis and feedback module is used to obtain the latest data on resource usage frequency based on the adjusted resource distribution status records, analyze the balance of resource distribution, and determine whether the preset allocation balance threshold is reached; The optimization iteration module is used to regenerate allocation instructions and update the resource distribution view when the allocation balance threshold is not reached, and output the final resource optimization distribution result.

Images