Cognitive load balanced teaching PPT automatic typesetting method and system

Abstract

The invention relates to the technical field of automatic typesetting, in particular to a teaching PPT automatic typesetting method and system with cognitive load balance, which comprises the following steps of: based on a teaching PPT page scene, analyzing information unit spatial layout and adjacent interval change, extracting spatial distribution characteristics, combining cognitive load balance, analyzing knowledge point paragraphs and character spans, and obtaining a teaching PPT page scene; dividing knowledge levels, screening visual level conflicts, and outputting page arrangement configuration parameters. According to the method, by analyzing page information unit distribution and knowledge point span characteristics, hierarchical aggregation and primary and secondary partition of knowledge content are achieved, page layout parameters are actively adjusted according to different changes of teaching content, priority judgment and configuration matching are automatically completed when parameter conflicts are processed, and the method is high in practicability. Page display with clear knowledge structure logic and clear primary and secondary contents is displayed, intelligent typesetting with balanced cognitive load is helped to be realized, key knowledge is guaranteed to be efficiently conveyed, and the orderliness and scene adaptability of PPT contents are improved.

Description

Technical Field

[0001] This invention relates to the field of automated typesetting technology, and in particular to a method and system for automatic typesetting of teaching PPTs with balanced cognitive load. Background Technology

[0002] Automated typesetting technology involves the structured, hierarchical, and spatial arrangement of text and graphics information on digital media. This includes the identification and parsing of content elements, the formulation of typesetting logic rules, the generation and adaptation of layout schemes, and its widespread application in the automated presentation of various information expression media such as electronic documents, presentations, and web interfaces. Traditional teaching slide layout methods involve manually distributing text, images, and other information across pages based on the teaching content during the creation of teaching PPTs. The order of content and the layout structure are adjusted subjectively based on experience to highlight key teaching points and clearly convey information. This method relies heavily on manual arrangement rules and template design to configure content, typically employing methods such as manually adjusting content position and font size, selecting page structures based on preset templates, and manually dividing teaching units according to chapter logic to complete the slide layout task.

[0003] Existing technologies primarily rely on manual rules and template-driven approaches. During operation, they fail to perceive and dynamically manage the content distribution, knowledge point span, and page hierarchy in real time. The presentation order and structural division of knowledge units lack precision, easily leading to unclear boundaries between key information and auxiliary content, and recurring issues such as excessively high content density or uneven distribution. Visual logic in layout is often chaotic, and there is a frequent problem of failing to distinguish between primary and secondary elements. In complex or ever-changing teaching scenarios, manual adjustments are inefficient and difficult to dynamically respond to the layout needs of different content structures. The coordination between information carrying and knowledge expression is poor, easily causing poor transmission of teaching information and difficulty in interpreting page content. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and propose a method and system for automatic layout of teaching PPTs with balanced cognitive load.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a method for automatically formatting teaching PPTs with balanced cognitive load, comprising the following steps: S1: Based on the teaching PPT page scenario, analyze the spatial distribution of each information unit in the coordinate system, calculate the adjacent interval of the same type of unit, compare the interval data before and after the page content adjustment, judge the distribution change trend, and obtain the spatial distribution change characteristics. S2: Based on the spatial distribution variation characteristics and combined with the cognitive load balancing target, analyze the marked knowledge point paragraphs, locate the character range of each knowledge point in the text, and calculate the span between knowledge points in turn to obtain the knowledge span distribution parameters. S3: Based on the knowledge span distribution parameters and combined with cognitive load balancing, count the number of knowledge points in the paragraph, match the number with the span data, analyze the knowledge structure characteristics, divide the main, supplementary and extended categories according to the proportion, and obtain the knowledge hierarchical structure identifier. S4: Based on the knowledge hierarchical structure identifier, for the automatic typesetting requirements, analyze the page partition structure parameters, integrate information density and visual hierarchy order, compare the correspondence between knowledge hierarchy and structure parameters, screen out inconsistent blocks, and obtain structure mapping feature groups. S5: Based on the structure mapping feature group, for parameter conflicts in the page layout area, according to the hierarchical priority rule of structural component type, the parameter order is compared one by one, the parameter combination with the lower order is eliminated, the optimal layout configuration is retained, and the page layout configuration parameters are obtained.

[0006] The present invention is improved in that the spatial distribution variation characteristics include distribution trend type, change range, and clustering characteristics; the knowledge span distribution parameters include span type, span statistical results, and association pattern; the knowledge hierarchical structure identifier includes main level, supplementary level, and extended level; the structure mapping feature group includes partition category, density ratio, and visual sorting; and the page layout configuration parameters include layout method, priority sequence, and content matching result.

[0007] The present invention is improved in that the step of obtaining the spatial distribution variation characteristics is specifically as follows: S111: Based on the teaching PPT page scenario, detect the page coordinates and boundaries of text segments, images, charts and tables within each page, collect the horizontal and vertical axis distribution parameters, calculate the shortest interval distance of similar information units in the page coordinate system, integrate various interval data, and obtain the information interval distribution within the page. S112: Based on the information interval distribution within the page, compare the shortest interval data of the same type of information units on the same page before and after the page content adjustment, statistically analyze the distribution state of interval increase or decrease, and classify it into three categories: interval shortening, interval widening, and no change, to obtain the type of page interval change. S113: Based on the type of page interval variation, statistically analyze the interval variation categories of information units on the horizontal and vertical axes of each page, determine whether the information units on the entire page are clustered, diffused, or fluctuating, and obtain the spatial distribution variation characteristics.

[0008] The present invention is improved in that the steps for obtaining the knowledge span distribution parameter are as follows: S211: Based on the spatial distribution variation characteristics, analyze the knowledge point paragraphs marked on the teaching PPT page, extract the start and end positions of the characters of the knowledge points in each paragraph, analyze the coordinates of each knowledge point in the character sequence according to the character order, calculate the character span between adjacent knowledge points, and obtain the knowledge point span interval sequence. S212: Based on the knowledge point span interval sequence, sort the character span intervals according to the order of knowledge points, perform summation and difference calculation of maximum and minimum intervals on the span intervals respectively, and summarize the central trend of the data to obtain paragraph span interval statistical parameters; S213: Based on the paragraph span interval statistical parameters, combined with the number of knowledge points, character density diffusion rate and character span fluctuation rate of each paragraph, calculate the distribution dispersion of each paragraph under the character arrangement structure to obtain the knowledge span distribution parameters.

[0009] The present invention is improved in that the steps for obtaining the knowledge hierarchical structure identifier are as follows: S311: Based on the knowledge span distribution parameters, sequentially identify the start and end positions of characters of knowledge points in paragraph text, calculate the character distance between each pair of adjacent knowledge points, aggregate character span data within paragraph, and count the cumulative span and quantity to obtain paragraph span aggregation quantity; S312: Based on the paragraph span aggregation amount, retrieve the number of knowledge points in each paragraph, calculate the knowledge point density offset, characterize the proportion matching features of each paragraph, compare the structural difference distribution trend, and obtain the paragraph matching trend amount. S313: Based on the paragraph matching trend quantity, combine the paragraph distribution characteristics, and divide the trend quantity into the main level, supplementary level and extended level in sequence according to the distribution type to obtain the knowledge hierarchical structure identifier.

[0010] The present invention is improved in that the step of obtaining the structure mapping feature group is specifically as follows: S411: Based on the knowledge hierarchical structure identifier, according to the structural component information in each partition of the page, detect the number of information units in each partition, obtain the actual page area occupied by the partition, calculate the number of information units per unit area of ​​each partition, record the information unit density measure of the corresponding partition, and obtain the information density set. S412: Based on the information density set, the page coordinates and dimensions of each structural component are statistically analyzed. A visual sorting sequence for each partition is established based on the coordinates and dimensions. The visual sorting sequence of each partition is compared with the information density set in turn. Partitions with inconsistent visual sorting and information density sorting are filtered out to obtain the density and sorting difference group. S413: Based on the density and sorting difference group, extract the knowledge level type, visual sorting sequence and information density measure of each partition, refer to the knowledge level structure parameters, perform partition calculation according to the degree of sorting deviation, obtain the visual density sorting deviation magnitude, and obtain the structure mapping feature group.

[0011] The present invention is improved in that the steps for obtaining the page layout configuration parameters are specifically as follows: S511: Based on the structure mapping feature group, call the partition category of the page layout area, analyze the density ratio and visual sorting parameters in each area, classify the structural component types item by item, establish a sorting sequence using the hierarchical priority rule, compare the priority order of each parameter, filter out combinations that do not meet the priority rule, and obtain the parameter priority structure sequence. S512: Based on the parameter priority structure sequence, perform order comparison on inconsistent arrangement items in each region, determine the positions where there are contradictions, adjust the sorting position of the associated parameter items in turn, and combine and update the order of structural components and arrangement parameters to obtain the visual structure rearrangement sequence. S513: Based on the visual structure rearrangement sequence, identify the density ratio and visual sorting of the structural component types, screen the combinations with visual conflicts or structural repetitions one by one, retain the parameter combinations with priority in the arrangement order, and obtain the page layout configuration parameters.

[0012] An automatic PPT formatting system for cognitive load balancing, the system comprising: The spatial feature extraction module is based on the teaching PPT page scenario. It analyzes the spatial layout characteristics of information units in each page, calculates the adjacent intervals of similar information units in the page coordinate system, compares the changes of various intervals before and after page content adjustment, judges the spatial distribution change trend, summarizes the changes in maximum interval, minimum interval and dispersion, and obtains the spatial distribution change characteristics. Based on the aforementioned spatial distribution variation characteristics, the knowledge span aggregation module analyzes the knowledge point paragraphs marked on the teaching PPT page around the goal of balancing cognitive load, locates the start and end character ranges of each knowledge point in the text, calculates the character span between knowledge points in turn, and aggregates the span calculation results of each paragraph in order to obtain the knowledge span distribution parameters. The hierarchical structure division module, based on the knowledge span distribution parameters and combined with the cognitive load balancing requirements, counts the number of knowledge points in each paragraph, matches the number of knowledge points in each paragraph with the span data in proportion, analyzes the knowledge structure characteristics of the paragraph, and divides the main, supplementary and extended levels according to the matching results to obtain the knowledge hierarchical structure identifier. Based on the knowledge hierarchical structure identifier, the structure mapping generation module analyzes the structural component information of each section of the page according to the automatic layout requirements, integrates information density and visual hierarchy order, compares the correspondence between knowledge hierarchy and parameter items, screens out areas where density and visual hierarchy order do not match, and obtains structure mapping feature groups. Based on the structure mapping feature group, the layout configuration optimization module addresses parameter conflicts in the page layout area by comparing the parameter order one by one according to the hierarchical priority rule of structural component types, eliminating parameter combinations that are ranked lower, and retaining the optimal layout configuration to obtain the page layout configuration parameters.

[0013] Compared with the prior art, the advantages and positive effects of the present invention are as follows: This invention analyzes the distribution of page information units and the span characteristics of knowledge points to achieve hierarchical aggregation and primary / secondary distinction of knowledge content. During the layout process, it dynamically adapts the visual hierarchy based on the distribution status and structural mapping. It proactively adjusts page layout parameters according to different changes in teaching content, and automatically completes priority determination and configuration matching when handling parameter conflicts. This results in a page display with a clear knowledge structure and distinct primary and secondary content, helping to achieve intelligent layout with balanced cognitive load, ensuring efficient transmission of key knowledge, improving the organization and scene adaptability of PPT content, promoting the scientific and automated level of teaching resource organization, and significantly facilitating the adaptive arrangement and intelligent presentation of diverse teaching content. Attached Figure Description

[0014] Figure 1 This is a flowchart of the main steps of the present invention; Figure 2 This is a flowchart illustrating the acquisition of spatial distribution variation characteristics in this invention. Figure 3 This is a flowchart illustrating the acquisition of knowledge span distribution parameters in this invention. Figure 4 This is a flowchart illustrating the process of obtaining the knowledge hierarchy structure identifier in this invention; Figure 5 This is a flowchart illustrating the process of obtaining structural mapping feature groups in this invention. Figure 6 This is a flowchart illustrating the process of obtaining page layout configuration parameters in this invention. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0016] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0017] Example Please see Figure 1This invention provides a technical solution: an automatic layout method for teaching PPTs with balanced cognitive load, comprising the following steps: S1: Based on the teaching PPT page scenario, analyze the spatial layout characteristics of information units in each page, calculate the adjacent intervals of similar information units in the page coordinate system, compare the changes of various intervals before and after page content adjustment, judge the spatial distribution change trend, summarize the changes in maximum interval, minimum interval and dispersion, and obtain the spatial distribution change characteristics. S2: Based on the spatial distribution variation characteristics and focusing on the goal of balancing cognitive load, analyze the knowledge point paragraphs marked on the teaching PPT page, locate the start and end character range of each knowledge point in the text, calculate the character span between knowledge points in turn, and aggregate the span calculation results of each paragraph in order to obtain the knowledge span distribution parameters. S3: Based on the knowledge span distribution parameters and combined with the cognitive load balancing requirements, the number of knowledge points in each paragraph is counted, and the number of knowledge points in each paragraph is matched with the span data in proportion. The knowledge structure characteristics of the paragraph are analyzed, and the main, supplementary and extension levels are divided according to the matching results to obtain the knowledge hierarchical structure identifier. S4: Based on the knowledge hierarchical structure identifier, according to the automatic layout requirements, analyze the structural component information of each section of the page, integrate the information density and visual hierarchy order, compare the correspondence between knowledge hierarchy and parameter items, screen out areas where the density and visual hierarchy order do not match, and obtain the structure mapping feature group. S5: Based on the structural mapping feature group, for parameter conflicts in the page layout area, the parameters are compared one by one according to the hierarchical priority rule of structural component type. The parameter combinations with lower ranking are eliminated, and the optimal layout configuration is retained to obtain the page layout configuration parameters.

[0018] Spatial distribution variation characteristics include distribution trend type, change range, and clustering characteristics; knowledge span distribution parameters include span type, span statistical results, and association patterns; knowledge hierarchical structure identifiers include main hierarchy, supplementary hierarchy, and extended hierarchy; structural mapping feature groups include partition category, density ratio, and visual sorting; and page layout configuration parameters include layout method, priority sequence, and content matching results.

[0019] In S1, an information unit refers to a basic content block within a teaching PPT page, typically including text segments, images, charts, tables, and other content components presented independently on the page. Each information unit has independent boundaries, position, and content. Spatial layout characteristics refer to the spatial distribution relationship and arrangement of information units within the PPT page, including spatial attributes such as the horizontal and vertical arrangement of information units and overall distribution density. Adjacent interval refers to the shortest distance measured between information units of the same category in the page coordinate system, such as the shortest distance between two text blocks or two images, reflecting the closeness between units. Interval change refers to the numerical change in the adjacent interval between information units before and after changes to the PPT page content, which is usually the focus of analysis when adding or removing content or adjusting the layout. Spatial distribution change trend refers to the overall trend of spatial distribution pattern becoming concentrated, dispersed, or balanced after comparing the interval data of all information units before and after content adjustment. Dispersion change refers to the change in the dispersion of information unit interval data, mainly judged by statistically analyzing the distribution, range, or standard deviation of each group of intervals to determine the uniformity of information distribution.

[0020] In S2, the cognitive load balancing goal refers to the reasonable allocation of page content during PPT layout, avoiding excessive concentration of information in a certain area, reducing the information processing pressure on the audience, and optimizing the learning experience; knowledge point paragraphs refer to paragraphs or text segments in the PPT text that have been clearly marked or divided into sections, with each paragraph generally carrying one or more knowledge points; start and end character range refers to the specific location range of a knowledge point in the paragraph text, that is, the start and end subscripts of each knowledge point in the string; character span refers to the character distance between two knowledge points in the text, reflecting the tightness of the content between knowledge points or the degree of information jump; sequential aggregation refers to organizing and classifying the span data between knowledge points according to the order in which the knowledge points appear in the text, which is convenient for subsequent statistics and analysis.

[0021] In S3, the cognitive load balancing requirement refers to applying the load balancing principle from cognitive psychology to PPT layout, requiring the reasonable division and display of knowledge points to ensure that the audience can smoothly receive and understand the information; proportional matching refers to pairing and associating the number of knowledge points with their corresponding span data according to a certain proportional relationship, which is used for subsequent knowledge level judgment; structural features refer to the knowledge organization structure attributes composed of the number, span, and distribution of knowledge points within each knowledge paragraph, which are used to identify paragraph types; the core, supplementary, and extension levels refer to classifying each knowledge paragraph into three different levels—core (core), auxiliary explanation (supplementary), and extension (extension)—based on the importance and density of the knowledge content, in order to facilitate structured presentation.

[0022] In S4, automatic layout requirements refer to the need to automatically generate aesthetically pleasing and scientifically sound teaching PPTs, involving multiple aspects such as content layout, logical hierarchy, and cognitive balance; structural component information refers to the structural attributes of various content components involved in page partitions, such as the size, position, and arrangement of text boxes and image boxes; information density refers to the comprehensive reflection of the number and spatial distribution of information units within a certain page area, used to assess the crowding degree of the area's content; visual hierarchy order refers to the hierarchical structure formed by various content components on the page based on visual appeal or logical priority, determining the order in which the audience reads the content; correspondence refers to the mapping between knowledge hierarchy classification and page parameter items (such as structure, density, visual hierarchy, etc.), used to guide reasonable layout; misaligned areas refer to page blocks in the actual page layout where there is a contradiction or unreasonable correspondence between information density and visual hierarchy order.

[0023] In S5, parameter conflict refers to a situation where there are contradictions among various parameters (such as content density, structural priority, visual hierarchy, etc.) within the page layout area, and it is difficult to satisfy them simultaneously. Hierarchical priority rule refers to the principle of prioritizing parameters in automatic layout. When parameter conflicts occur, the higher priority structural or content parameters are used to adjust the page. Optimal layout configuration refers to the final selection of the layout parameter combination that best suits the current page structure and the need for cognitive load balance after conflict screening and priority sorting.

[0024] Please see Figure 2 The specific steps for obtaining the spatial distribution variation characteristics are as follows: S111: Based on the teaching PPT page scenario, detect the page coordinates and boundaries of text segments, images, charts and tables within each page, collect the horizontal and vertical axis distribution parameters, calculate the shortest interval distance of similar information units in the page coordinate system, integrate various interval data, and obtain the information interval distribution within the page. The process reads each page sequentially, performing categorization on all text segments, images, charts, tables, and other information units. For each type of information unit, its positional boundary information on the page is extracted. By obtaining its top-left and bottom-right corners, its center position is deduced. Then, all units of the same category are arranged sequentially. Next, the shortest distance between pairs of units is calculated horizontally and vertically. The difference between the horizontal and vertical center points of two information units is compared, and the smaller difference is selected as the shortest interval distance for that pair. This process is repeated for all similar information units on a page, and the smallest result is selected as the shortest interval data for that type of information unit. For example, if a page contains three text segments, with their centers located in the left, center, and right areas respectively, the horizontal interval between them is calculated. If the minimum value is found to be 60 pixels, it is recorded as the shortest interval for text. Simultaneously, for each type of information unit, the horizontal and vertical axis values ​​at its center are collected, and the maximum, minimum, and average values ​​in this set are recorded to express its horizontal and vertical distribution boundaries and concentration. For example, if the horizontal axis coordinates of the center of image units on a page are 100, 300, and 800, then the maximum value is 800, the minimum value is 100, and the average value is 400, indicating a relatively dispersed distribution. The distribution boundaries and shortest interval values ​​for the four types of information units—text, images, charts, and tables—are obtained in this way, and integrated into four sets of distribution parameters. These serve as the interval data set for the page. When analyzing changes in the PPT page structure, this data can be used as the basis for determining the density and layout of page space, thus completing the extraction of the information interval distribution within the page.

[0025] S112: Based on the information interval distribution within the page, compare the shortest interval data of the same type of information unit on the same page before and after the page content adjustment, statistically analyze the distribution of interval increase or decrease, and classify it into three categories: interval shortening, interval widening, and no change, to obtain the type of page interval change. Before and after layout, the content of each page was compared. The shortest interval data for each type of information unit was extracted before and after the adjustment, and the difference between the two sets of data was compared. The positive or negative value was used to determine whether the information units became more concentrated or sparse after layout. If the change value was negative and greater than 5 pixels, it was classified as a shortened interval; if the change value was positive and greater than 5 pixels, it was classified as a widened interval; if the change value fluctuated between plus and minus 5 pixels, it was judged as no change. Here, 5 pixels is a manually set threshold, based on the line height and layout sensitivity of paragraphs in common PPTs. Usually, when the line height is 20 pixels, 5 pixels is sufficient to constitute a significant change in visual interval. This method was used to classify the interval change types of all information unit categories sequentially, and the frequency of occurrence was counted for each of the three categories: shortened, widened, and no change. For example, in a teaching page, the shortest interval of a text segment changed from 60 pixels to 48 pixels, a change of... If the interval is 12 pixels, it is classified as a shortened interval. If the interval of an image unit changes from 80 pixels to 120 pixels, which is a change of +40 pixels, it is classified as a widened interval. If the interval of a chart unit changes by +3 pixels, it is recorded as no change. After classifying in this way, the type of interval change on the page is recorded, and the output is the change trend of four types of information units. For example, the interval type of the page can be represented as shortened text, widened image, no change in chart, and no change in table, which can be used for subsequent horizontal and vertical axis analysis to complete the classification of the types of interval changes between pages.

[0026] S113: Based on the type of variation in the intervals between pages, statistically analyze the types of interval changes of information units on the horizontal and vertical axes of each page, determine whether the information units on the entire page are clustered, diffused, or fluctuating, and obtain the spatial distribution variation characteristics. On each page, the changes in the horizontal and vertical axes of similar information units are statistically analyzed. The difference between the horizontal center position change and the vertical position change before and after content adjustment is calculated for each type of information unit, and it is determined whether it is shortened, enlarged, or unchanged. For example, comparing the horizontal center point distance of the chart module before and after, if it decreases by 10 pixels, it is marked as horizontal shortening; if it increases by 12 pixels, it is marked as horizontal enlargement; if the change is within... 5 pixels to +5 pixels are considered unchanged. Similarly, the vertical interval state is determined. For each type of information unit, the frequency of its interval change type in the entire page is recorded along both the horizontal and vertical axes, and the total number is counted. The results are used to determine if a certain state in a particular direction accounts for more than 60%. For example, if four out of five image areas on the page show vertical enlargement, the vertical trend is considered the dominant direction as diffusion. If three horizontal areas show shortening and two show no change, the dominant direction is clustering. If the dominant types in both directions are the same, the information unit is marked as either clustered or diffused. If the directional trends are inconsistent, it is judged as a fluctuating distribution. If the frequency of all directional change types is balanced and the dominant direction is not established, it is recorded as having no significant distribution trend. For example, if there are six pairs in the horizontal axis of a text area on a page, five of which show shortening intervals and the remaining one remains unchanged, then the horizontal direction is the dominant aggregation direction. If four of the six pairs in the vertical axis show diffusion, then the vertical direction is determined to be the diffusion direction. The text units in this page are typical fluctuating distributions. Thus, the dominant trends in the horizontal and vertical directions and the overall spatial distribution variation characteristics of each type of information unit are generated page by page, completing the classification output of the page spatial distribution variation characteristics.

[0027] Please see Figure 3 The specific steps for obtaining the knowledge span distribution parameters are as follows: S211: Based on the spatial distribution variation characteristics, analyze the knowledge point paragraphs marked on the teaching PPT page, extract the start and end positions of the characters of the knowledge points in each paragraph, analyze the coordinates of each knowledge point in the character sequence according to the character order, calculate the character span between adjacent knowledge points, and obtain the knowledge point span interval sequence. Based on the knowledge point paragraphs marked on the teaching PPT slides, the text content is read paragraph by paragraph, and a character index is established. Each character in the paragraph is numbered, for example, starting from the first character and numbered sequentially as position 1, position 2, and so on until the end of the paragraph. Then, according to the pre-set knowledge point marking information, the start and end positions of each knowledge point in the character sequence are extracted. For example, in a paragraph of text "Voltage is the reason for driving current", the start position of "voltage" is position 1 and the end position is position 2, and the start position of "current" is position 6 and the end position is position 7. After extraction, the knowledge point intervals are formed as "positions 1-2" and "positions 6-7". Then, the start positions of all knowledge points are arranged according to the character numbering order to ensure that the processing order is consistent with the actual reading order in the text. Then, the relationships between adjacent knowledge points are calculated one by one. The span, or number of characters between the end position of the current knowledge point and the start position of the next knowledge point, is the number of characters between the 6th and 2nd positions of the adjacent knowledge points "voltage" and "current" in the example above. This results in a 4-character interval. This process is repeated to iterate through the intervals of all adjacent knowledge points in the entire paragraph, forming the knowledge point span interval sequence for that paragraph. If the paragraph contains five knowledge points, four span values ​​can be obtained. In practice, for the same text, there may be multiple phrases or terms as knowledge points, separated by other ordinary characters, spaces, or punctuation. During processing, it is necessary to exclude the interference of non-knowledge point characters on the interval values ​​to ensure that only the span between knowledge points is analyzed and is not affected by non-critical content. The set of adjacent interval values ​​obtained after this process is completed is the knowledge point span interval sequence.

[0028] S212: Based on the knowledge point span interval sequence, sort the character span intervals according to the order of knowledge points, perform summation and difference calculation of maximum and minimum intervals on the span intervals respectively, and summarize the central tendency of the data to obtain the paragraph span interval statistical parameters. Maintaining the original arrangement order, the interval values ​​are organized according to the order in which the knowledge points appear in the text. For example, if a paragraph contains knowledge point intervals of 3, 5, 2, 6, and 4, resulting in five sets of values, the following statistical operations are performed on these values. The first step is summation: all interval values ​​are added together, i.e., each set of interval values ​​is added from beginning to end, forming the cumulative character span between all knowledge points in the paragraph. The second step is to identify the largest and smallest values ​​in this set of interval values ​​and calculate the difference between them. For example, in the above interval sequence, the maximum value is 6, the minimum value is 2, and the difference is 4. The third step is to determine the central tendency of this set of interval values, i.e., whether the values ​​are concentrated in one area or not. The distribution can be determined by judging whether most of the values ​​in each group are clustered in the middle range. For example, if most interval values ​​are concentrated between 3 and 5, and only a very few are higher or lower than this range, then the concentration trend is considered to be obvious. Conversely, if all values ​​are evenly distributed between 2 and 6, the concentration trend is weak. In addition, the deviation between each interval value and the average level can be estimated. For example, using the average span of a paragraph as a reference standard, we can see how much each interval value is higher or lower than the standard, so as to make a preliminary judgment on the magnitude of the fluctuation in the interval distribution. Through the statistical process, the span characteristics between knowledge points in the paragraph at the text level can be comprehensively described, and the span interval statistical parameters of the paragraph can be obtained.

[0029] S213: Based on the paragraph span interval statistical parameters, combined with the number of knowledge points, character density diffusion rate and character span fluctuation rate of each paragraph, calculate the distribution dispersion of each paragraph under the character arrangement structure, and obtain the knowledge span distribution parameters; After completing the statistical parameter analysis of knowledge point intervals, it is necessary to combine it with the number of knowledge points, text density, and span fluctuation of the paragraph itself for comprehensive evaluation. First, count the number of knowledge points in the paragraph, that is, the total number of content marked as knowledge points. Then, obtain the total number of characters in the paragraph, that is, the length from the first character to the last character, including spaces and punctuation. Then, compare the number of knowledge points with the total number of characters to obtain a density ratio value, reflecting the density of knowledge points in the paragraph. For example, if a paragraph has 150 characters and 6 knowledge points are marked, it means that there is one knowledge point for every 25 characters. Next, it is necessary to evaluate the fluctuation of the knowledge point span interval, review the previously obtained span sequence, and observe the values. Whether they are close or not, if the sequence is 5, 5, 5, 5, 5, it indicates that the span is stable with small fluctuations. If it is 2, 7, 3, 8, 4, it indicates that the span fluctuates significantly and the difference is large. Further, the difference range or quantity ratio of the interval value can be used for measurement. For example, a significant fluctuation can be defined when the difference in the span interval exceeds a certain value. Then, the overall fluctuation of the paragraph is judged. Finally, the three indicators of knowledge point quantity, character density and span fluctuation are combined and represented in a unified data structure form. They are compared according to a unified judgment standard. For example, these three data are compared in five paragraphs at the same time to determine which paragraphs have a uniform distribution of knowledge points, which are too concentrated or have large jumps. The knowledge span distribution parameters of each paragraph are output in the form of numerical quantification.

[0030] Please see Figure 4 The specific steps for obtaining the knowledge hierarchy structure identifier are as follows: S311: Based on the knowledge span distribution parameters, sequentially identify the start and end positions of characters of knowledge points in paragraph text, calculate the character distance between each pair of adjacent knowledge points, aggregate character span data within paragraph, and count the cumulative span and quantity to obtain the paragraph span aggregation quantity. The process involves sequentially identifying the start and end positions of characters for knowledge points within a paragraph. First, the paragraph text is read, and each character is indexed sequentially (e.g., the first character is numbered 1, and so on). Then, based on the annotation information, the start and end numbers of each knowledge point within the paragraph are extracted. All knowledge points are sorted from smallest to largest according to their start positions to ensure the analysis order matches the text's order of appearance. Next, for adjacent knowledge points, the number of characters between the end position of one knowledge point and the start position of the next is calculated as the character span for that pair of knowledge points. For example, the number of characters between the end position of one knowledge point and the start position of the next is calculated. If knowledge point A starts and ends at positions 5 to 10, and knowledge point B starts and ends at positions 15 to 18, then the span is 15 minus 10, which equals 5 characters. After processing the span data between all adjacent knowledge points in the entire paragraph, the character spans are aggregated, that is, all span values ​​are added together to obtain the total span of the paragraph. The number of spans participating in the aggregation is also counted. For example, if the paragraph contains 6 knowledge points, then there are 5 groups of adjacent pairs. If the spans of each group are 3, 6, 5, 4, and 7, then the total is 25, and the number is 5. All span data and statistical results together constitute the span aggregation of the paragraph.

[0031] S312: Based on the paragraph span aggregation, retrieve the number of knowledge points in each paragraph, calculate the knowledge point density offset, characterize the proportional matching features of each paragraph, compare the structural difference distribution trend, and obtain the paragraph matching trend quantity. First, based on the completed cumulative span statistics, for each paragraph, the total number of knowledge points contained in that paragraph is retrieved. For example, paragraph A contains 5 knowledge points, paragraph B contains 3, and paragraph C contains 8. These are recorded for subsequent proportional analysis. Then, the span aggregation value of each paragraph is paired with the number of knowledge points to calculate the average distribution span of knowledge points in each paragraph, which serves as the basis for assessing the density. In specific processing, a judgment standard value for character density offset is set. For example, setting no more than 1 knowledge point per 10 characters is considered low density, and more than 3 is considered high density. If the average span in a paragraph is 4 characters... If it is 12 characters, it is considered a high-density area; if it is 12 characters, it is considered a low-density area. Then, based on the judgment results, the density status of each paragraph is marked. Then, based on the density status of different paragraphs, the proportional matching relationship of the distribution characteristics between paragraphs is identified. For example, paragraph A has a high density, paragraph B has a moderate density, and paragraph C has a low density. Finally, the distribution status of all paragraphs is compared and processed to identify the overall trend of structural difference distribution. If the density has abrupt changes or gradual changes in multiple paragraphs, it is recorded as a significant trend. If the density difference is not obvious, it is classified as a stable trend. The trend type, combined with the density judgment of each paragraph and the order of distribution difference classification, forms the paragraph matching trend quantity.

[0032] S313: Based on paragraph matching trend data, combine paragraph distribution features, and divide the trend data into main level, supplementary level and extended level according to distribution type to obtain knowledge hierarchical structure identifier; The trend identifiers for each paragraph are read, including their density and relative change trends. Based on this, the trend results of all paragraphs are classified. The main level is defined as having high knowledge point density, stable span, and relatively concentrated trend. The supplementary level is defined as having moderate density, limited span variation, and no significant trend fluctuation. The extended level is defined as having low knowledge point density, large span fluctuation, and unstable trend. For example, paragraph A has high knowledge point density, an average span of 4 characters, a fluctuation value of less than 3 characters, and a concentrated trend, so it is classified as the main level. Paragraph B has a density of 7 characters per knowledge point, a moderate fluctuation value, and a stable trend distribution, so it is classified as the supplementary level. Paragraph C has uneven knowledge point spans, with intervals between 3 and 20 characters, and the trend distribution changes continuously as the paragraph progresses, so it is classified as the extended level. In the above judgment process, if multiple paragraphs are under the boundary conditions, the matching trend difference between paragraphs is used for priority sorting to ensure that the level division has continuity and structural differentiation. The knowledge content level type of each paragraph is output, forming a knowledge hierarchical structure identifier.

[0033] Please see Figure 5 The specific steps for obtaining the structure mapping feature set are as follows: S411: Based on the knowledge hierarchical structure identifier, according to the structural component information in each section of the page, detect the number of information units in each section, obtain the actual page area occupied by the section, calculate the number of information units per unit area of ​​each section, record the information unit density measure of the corresponding section, and obtain the information density set. The layout area division of each page is read, and the structural component attribute information of each area is extracted. Each structural component corresponds to an information unit, and its specific type needs to be identified, such as text box, image box, chart area, table area, etc. Then, the number of information units contained in each partition is counted as the basis for counting the content of that area. At the same time, the actual area occupied by each partition on the page is obtained. The area is calculated by the length and width boundary values ​​of the area in the page coordinate system. For example, if a partition is 12 cm long and 8 cm wide, the area is 96 square centimeters. Combining the 6 information units contained in the partition, the number of area information units is 6 divided by 96, which is 0.0625. The calculation is performed on all partitions of the page in this way to obtain the number of area information units, and this value is used as the information unit density measure of the partition. After all page area density measures are uniformly sorted, a complete information density set is formed.

[0034] S412: Based on the information density set, the page coordinates and dimensions of each structural component are statistically analyzed. A visual sorting sequence for each partition is established based on the coordinates and dimensions. The visual sorting sequence of each partition is compared with the information density set in turn. Partitions with inconsistent visual sorting and information density sorting are filtered out to obtain the density and sorting difference group. For each structural component, its coordinate information and dimension data on the page are extracted. The coordinate information is the component's positioning coordinates in the upper left corner reference frame of the page, and the dimension data is the component's width and height. Then, based on the coordinate positions and dimension information of all structural components, the visual presentation order of the components on the page from top to bottom and from left to right is identified. A visual sorting sequence is formed according to the standard eye flow reading order. For example, the component at the top left is numbered 1, the component in the center is numbered 2, and the component at the bottom right is numbered 3, forming the sequence [1, 2, 3]. At this time, the visual sorting sequence is mapped to each section and the information already recorded in each section. The unit density measure is compared, and the units are reordered from high to low to form another sequence. For example, the density order is {2, 3, 1}. If there is a discrepancy between this order and the visual order, for example, number 1 is ranked first in the visual order but third in the density order, then the partition is judged to be inconsistent. All partitions with inconsistent order positions are included in the difference group. The selection criteria need to be clearly defined. For example, if the difference in the ranking of the same partition in the two order sequences exceeds 1 unit, it is considered to be inconsistent. The set of partitions with inconsistent density and order is output, and the density and order difference group is obtained.

[0035] S413: Based on the density and sorting difference groups, extract the knowledge level type, visual sorting sequence and information density measure of each partition, refer to the knowledge level structure parameters, perform partitioning calculation according to the degree of sorting deviation, and obtain the structure mapping feature group. The knowledge level type corresponding to each difference partition is extracted. This type comes from the classification results of the previous knowledge hierarchical structure identification, including three types: backbone level, supplementary level, and extended level. Then, the visual ranking position and density metric value of each partition are read one by one and classified in order. Based on this, the standard matching rules set by the knowledge hierarchy structure parameters are applied to classify the degree of ranking deviation of each partition. In specific operation, the content of the backbone level is set to be at the top of the visual ranking and have a high density. If the actual ranking position is more than two places lower than the density ranking, it is judged as a deviation. The supplementary level is set to be in the middle. If the difference between the density ranking and the visual ranking is within plus or minus one place, it is judged as conforming. The extended level is allowed to have larger fluctuations, and the difference within three places is acceptable. According to the above standards, the ranking deviation of each difference partition is compared one by one. Partitions that do not meet the expected level are marked and their knowledge level type, visual ranking position, density metric value, and ranking deviation rank are recorded to form a structural mapping feature group, which serves as the basis for subsequent layout structure adjustment.

[0036] Please see Figure 6 The specific steps to obtain the page layout configuration parameters are as follows: S511: Based on the structural mapping feature group, call the partition category of the page layout area, analyze the density ratio and visual sorting parameters in each area, classify the structural component types item by item, establish a sorting sequence using the hierarchical priority rule, compare the priority order of each parameter, filter out combinations that do not meet the priority rule, and obtain the parameter priority structure sequence. The specific type of each content section in the page layout area is retrieved, and parameter comparison is performed section by section. The type identifier of each section is read item by item from the page structure information. This identifier is usually classified according to the content structure components, including text areas, image areas, chart areas, table areas, etc. After reading, two key parameters are obtained for each area: density ratio and visual order. The density ratio is the ratio of the number of information units to the area of ​​the section, and the visual order is the order in which the section appears on the page reading path. Next, the structural component types are classified. For example, all text areas are uniformly labeled as type T, image areas as type I, chart areas as type C, and table areas as type G. After classification, the areas are sorted according to a preset hierarchical priority rule. The rule is set as follows: the type sorting weight of the main knowledge priority display area is text > T-type charts > C-type images > I-type tables > G-type. According to this rule, each area type is mapped and replaced with the corresponding priority number. Then, according to the priority and parameter value, the density ratio and visual sorting number of each area are sorted and compared one by one. If an area has a density ratio of 1st and a visual sorting of 3rd, with a priority of 2, then the combination does not meet the priority structure requirements. Such sorting results are eliminated according to the preset logic. Combinations that do not meet the priority order are recorded in the exclusion set. All area combinations that do not meet the type priority rule in both density ratio and visual sorting are eliminated. The structural parameter arrangement order formed under the parameter sorting priority constraint is obtained as the parameter priority structure sequence.

[0037] S512: Based on the parameter-priority structural sequence, perform order comparison on inconsistent arrangement items in each region, determine the positions of contradictions, adjust the sorting position of associated parameter items in turn, and combine and update the order of structural components and arrangement parameters to obtain the visual structural rearrangement sequence. Each region in the structural parameters is compared with the parameter sequence in the final arrangement according to its visual order. The region number, layout coordinates, visual order, and type identifier of each structural component are extracted according to the order in the sequence. Then, the layout position matching the region number is found in the original page layout structure. Its actual position on the page is compared with the expected order in the parameter priority structural sequence. If a discrepancy is found, the region is marked as a conflict, and the difference between its position number and parameter value is recorded. For example, if a region is ranked 2nd in the parameter priority sequence but 4th in the visual display sequence (a two-digit offset), its order is considered to have a structural contradiction. Subsequently, all... Inconsistent items undergo an order adjustment operation, which adjusts their arrangement order according to the order position recorded in the parameter priority structure sequence. During this process, the arrangement coordinates need to be called to revise the X and Y coordinates of the corresponding regions and align them with the preceding and following regions to form a rearranged structure. At the same time, the order number of each region in the arrangement is updated to make it completely match the structure priority sequence. For example, if the original visual order is [1, 3, 2, 4] and the parameter priority structure order is [1, 2, 3, 4], then the original 2nd and 3rd region needs to be swapped in the coordinates. After the adjustment is completed, a new arrangement order is obtained, and the region order identifier is regenerated to obtain the visual structure rearranged sequence after the structural parameter alignment and visual sorting are integrated.

[0038] S513: Based on the visual structural rearrangement sequence, identify the density ratio and visual sorting of the structural component types, screen the combinations with visual conflicts or structural repetitions one by one, retain the parameter combinations with priority in the arrangement order, and obtain the page layout configuration parameters. The density ratio and visual ranking information corresponding to each structural component are identified, and the combination is screened item by item. First, the density ratio value and visual ranking position of each region in the rearrangement sequence are read sequentially to construct the arrangement combination item for each region. Then, two types of screening operations are performed on all combinations. The first type of screening is to identify combinations with visual conflicts, that is, regions with high visual ranking but extremely low density ratios are judged as ranking anomalies. The second type of screening is to screen for structural repetition combinations, that is, to identify repetitive modules with the same type of identifier, similar ranking values, but very small differences in density ratio. Two judgment thresholds are set during screening. The density ratio anomaly judgment threshold is less than 0.00002, and the ranking position deviation exceeds 0.00002. If the difference in density ratio between three positions is less than 0.00001 and the difference in ranking is less than two positions, it is considered a duplicate combination. For the above screening results, combinations with poor visual ranking and density ratio are eliminated one by one, and combinations with high ranking are retained. For example, if the density ratios of two image regions are 0.00003 and 0.000012, and the rankings are 3 and 2, the combination with ranking 2 is retained, and the combination with ranking 3 is eliminated. Finally, all the retained combinations are summarized to generate the final set of layout parameters used on the page. Each parameter combination in this set is an effective configuration combination retained through visual structural rearrangement and priority screening, and is used as the page layout configuration parameter.

[0039] A cognitive load-balanced automatic PPT formatting system, comprising: The spatial feature extraction module is based on the teaching PPT page scenario. It analyzes the spatial layout characteristics of information units in each page, calculates the adjacent intervals of similar information units in the page coordinate system, compares the changes of various intervals before and after page content adjustment, judges the spatial distribution change trend, summarizes the changes in maximum interval, minimum interval and dispersion, and obtains the spatial distribution change characteristics. Based on the spatial distribution variation characteristics and with the goal of balancing cognitive load, the knowledge span aggregation module analyzes the knowledge point paragraphs marked on the teaching PPT page, locates the start and end character range of each knowledge point in the text, calculates the character span between knowledge points in turn, and aggregates the span calculation results of each paragraph in order to obtain the knowledge span distribution parameters. The hierarchical structure division module is based on the knowledge span distribution parameters and combined with the cognitive load balancing requirements. It counts the number of knowledge points in each paragraph, matches the number of knowledge points in each paragraph with the span data in proportion, analyzes the knowledge structure characteristics of the paragraph, and divides the main, supplementary and extended levels according to the matching results to obtain the knowledge hierarchical structure identifier. The structure mapping generation module is based on the knowledge hierarchical structure identifier. According to the automatic layout requirements, it analyzes the structural component information of each section of the page, integrates the information density and visual hierarchy order, compares the correspondence between knowledge hierarchy and parameter items, screens out areas where the density and visual hierarchy order do not match, and obtains the structure mapping feature group. The layout configuration optimization module is based on the structural mapping feature group. For parameter conflicts in the page layout area, it compares the parameter order one by one according to the hierarchical priority rule of structural component type, eliminates the parameter combination with the lower ranking, and retains the optimal layout configuration to obtain the page layout configuration parameters.

[0040] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for automatically arranging a teaching PPT with balanced cognitive load, characterized in that, Comprise the following steps: S1: based on the teaching PPT page scene, analyze the spatial distribution of each information unit in the coordinate system, calculate the adjacent interval of the same unit, compare the interval data before and after the page content adjustment, judge the distribution trend, get the space distribution change characteristics; S2: based on the space distribution change characteristics, combining the cognitive load balance target, analyzing the annotated knowledge point paragraph, positioning each knowledge point in the character interval in the text, sequentially calculating the span between knowledge points, obtaining the knowledge span distribution parameters; S3: based on the knowledge span distribution parameters, combining the cognitive load balance, counting the number of knowledge points in the paragraph, matching the number and span data, analyzing the knowledge structure characteristics, dividing the main, supplementary and expansion categories according to the proportion, obtaining the knowledge hierarchical structure identification; S4: based on the knowledge hierarchical structure identification, for the automatic layout requirements, analyze the page partition structure parameters, integrate the information density and visual level order, compare the corresponding relationship between knowledge level and structure parameters, screen out inconsistent blocks, get the structure mapping feature group; S5: based on the structure mapping feature group, for the parameter conflict of page arrangement area, according to the hierarchical priority rule of structure component type, compare the parameter sequence one by one, eliminate the parameter combination of late sorting, retain the optimal arrangement configuration, get the page arrangement configuration parameter.

2. The method of claim 1, wherein the method further comprises: The space distribution change characteristics include distribution trend type, change interval, aggregation characteristics, the knowledge span distribution parameters include span type, span statistical result, correlation mode, the knowledge hierarchical structure identification includes main level, supplementary level, expansion level, the structure mapping feature group includes partition category, dense matching, visual sorting, the page arrangement configuration parameter includes arrangement mode, priority sequence, content matching result.

3. The method of claim 1, wherein the method further comprises: The acquisition step of the space distribution change characteristics is specifically: S111: based on the teaching PPT page scene, detect the page coordinates and boundaries of text segments, images, charts and table information units in each page, collect the horizontal and vertical distribution parameters, calculate the shortest interval distance of the same information unit in the page coordinate system, integrate the interval data, get the information interval distribution in the page; S112: based on the information interval distribution in the page, comparing the shortest interval data of the same information unit before and after the page content adjustment, statistics the distribution state of interval increase and decrease, classified into three categories of interval shortening, expanding and no change, get the page interval change type; S113: according to the page interval change type, statistics the interval change category of information unit in horizontal and vertical axis in each page, judge the aggregation, diffusion or fluctuation distribution of information unit in the whole page, get the space distribution change characteristics.

4. The method of claim 1, wherein the method further comprises: The acquisition step of the knowledge span distribution parameters is specifically: S211: based on the space distribution change characteristics, analyze the annotated knowledge point paragraph of teaching PPT page, extract the character start and end position of knowledge point in each paragraph text, analyze the coordinates of each knowledge point in the character sequence according to the character order, calculate the character span between adjacent knowledge points, get the knowledge point span interval sequence; S212: based on the knowledge point span interval sequence, the character span interval is sorted according to the knowledge point arrangement order, the sum and the maximum-minimum interval difference calculation are respectively performed on the span interval, and the centralized trend of the data is summarized to obtain the paragraph span interval statistical parameter; S213: based on the paragraph span interval statistical parameter, the distribution discrete degree of each paragraph under the character arrangement structure is calculated by combining the number of knowledge points, the character density diffusion rate and the character span fluctuation rate of each paragraph, and a knowledge span distribution parameter is obtained.

5. The method of claim 1, wherein the method further comprises: The acquisition step of the knowledge hierarchical structure identifier is specifically: S311: based on the knowledge span distribution parameter, the character start and end positions of the knowledge points in the paragraph text are sequentially identified, the character distance between each pair of adjacent knowledge points is calculated, the paragraph character span data is aggregated, the span cumulative amount and the number are counted, and a paragraph span aggregation amount is obtained; S312: based on the paragraph span aggregation amount, the number of knowledge points in each paragraph is retrieved, the knowledge point dense offset amount is calculated, the proportion matching feature of each paragraph is represented, the structure difference distribution trend is compared, and a paragraph matching trend amount is obtained; S313: based on the paragraph matching trend amount, the paragraph distribution features are combined, the trend amount is sequentially divided into a main level, a supplementary level and an expansion level according to the distribution type, and a knowledge hierarchical structure identifier is obtained.

6. The method of claim 1, wherein the method further comprises: The acquisition step of the structure mapping feature group is specifically: S411: based on the knowledge hierarchical structure identifier, the number of information units in each partition information unit is detected according to the structure component information in each partition of the page, the actual occupied page area of the partition is obtained, the number of information units per unit area in each partition is calculated, the information unit density amount of the corresponding partition is recorded, and an information density set is obtained; S412: based on the information density set, the page coordinates and the size of each structure component are counted, the visual sorting sequence of each partition is established according to the coordinates and the size, the partition visual sorting sequence and the information density set are sequentially compared, the partitions with inconsistent visual sorting and information density sorting are screened, and a dense and sorting difference zone group is obtained; S413: based on the dense and sorting difference zone group, the knowledge level type, the visual sorting sequence and the information density amount of each partition are extracted, the knowledge hierarchical structure parameter is referenced, the partition calculation is performed according to the sorting deviation degree, the visual dense sorting deviation amplitude is obtained, and a structure mapping feature group is obtained.

7. The method of claim 1, wherein the method further comprises: The acquisition step of the page arrangement configuration parameter is specifically: S511: based on the structure mapping feature group, the partition category of the page arrangement area is called, the dense matching and visual sorting parameters in each region are analyzed, the classification of the structure component type is performed item by item, the hierarchical priority rule is used to establish the sorting sequence, the priority order of each parameter is compared, the combination that does not satisfy the priority rule is screened out, and a parameter priority structure sequence is obtained; S512: based on the parameter priority structure sequence, the order comparison of the inconsistent arrangement items in each region is performed, the position with contradiction is judged, the sorting position of the associated parameter item is sequentially adjusted, the order of the structure component and the arrangement parameter is combined and updated, and a visual structure reordering sequence is obtained. S513: Based on the visual structure reordering column, identify the dense matching corresponding to the structure component type and the arrangement of visual ordering, screen each combination of visual conflict or structure repetition, retain the parameter combination with priority in arrangement order, and obtain the page arrangement configuration parameter.

8. A cognitive load balanced teaching PPT automatic layout system, characterized in that, The system is used to realize the teaching PPT automatic layout method of cognitive load balance according to any one of claims 1-7, and the system comprises: The spatial feature extraction module analyzes the spatial layout features of information units in each page based on the teaching PPT page scene, calculates the adjacent intervals of similar information units in the page coordinate system, compares the interval changes of various types before and after the page content adjustment, judges the spatial distribution change trend, summarizes the maximum interval, minimum interval and dispersion change, and obtains the spatial distribution change feature; The knowledge span aggregation module analyzes the annotated knowledge point paragraphs in the teaching PPT page based on the spatial distribution change feature, locates the start and end character intervals of each knowledge point in the text, measures the character span between knowledge points in turn, aggregates the span measurement results of each paragraph in order, and obtains the knowledge span distribution parameter; The hierarchical structure division module combines the cognitive load balance requirement based on the knowledge span distribution parameter, counts the number of knowledge points in each paragraph, matches the number of knowledge points and span data according to the proportion, analyzes the paragraph knowledge structure features, divides the main, supplementary and expansion levels according to the matching results, and obtains the knowledge hierarchical structure identification; The structure mapping generation module analyzes the structure component information of each partition of the page based on the knowledge hierarchical structure identification, integrates the information density and visual hierarchical order, compares the correspondence between the knowledge hierarchy and the parameter item, screens the areas where the density and visual hierarchical order do not match, and obtains the structure mapping feature group; The arrangement configuration optimization module compares the parameter order according to the hierarchical priority rules of the structure component type based on the structure mapping feature group, eliminates the parameter combination with late order, retains the optimal arrangement configuration, and obtains the page arrangement configuration parameter.

Images