A method for intelligent courseware generation based on knowledge graph

Abstract

The present invention relates to the technical field of courseware generation, and in particular to a method for intelligently generating courseware based on a knowledge graph, comprising: determining a graph selection method to generate a target graph according to a dynamic connection index and a graph interactivity; determining a generation state according to a graph reference value and a learning characteristic value of the target graph, and determining a knowledge point selection method according to the generation state to obtain selected knowledge points; in node analysis selection, determining a knowledge point supplementation method according to a node evaluation value; in path analysis selection, determining a node state according to node complexity and a node distribution coefficient, and determining a screening method according to the node state; under a selection completion condition, determining an optimization method to obtain supplementary knowledge points according to a pre-replacement coefficient of an optimization range; and filling in each selected knowledge point and each supplementary knowledge point to generate courseware. The present invention can improve the accuracy of courseware generation.

Description

Technical Field

[0001] The present invention relates to the technical field of courseware generation, and in particular to a method for intelligently generating courseware based on a knowledge graph. Background Art

[0002] The traditional way of producing courseware mainly relies on teachers to manually collect information, organize knowledge points, and use office software for typesetting and design. This process is not only time-consuming and labor-intensive, but also easily limited by teachers' personal knowledge reserves and teaching experience, resulting in uneven quality of courseware content. In contrast, knowledge graph technology can connect scattered and isolated knowledge points into an organic whole, forming a knowledge network with rich semantic associations, but it still faces problems such as the integrity of knowledge graph construction and the effective selection of knowledge points. Therefore, how to effectively screen knowledge points in the knowledge graph to improve the accuracy of courseware generation is a technical problem that needs to be urgently solved by technical personnel in this field.

[0003] Chinese patent publication number CN116932488A discloses a method, device, system and storage medium for generating courseware based on knowledge graph, the method comprising: responding to a courseware creation request initiated by a user terminal and selecting a corresponding knowledge graph; obtaining a first mapping relationship between class hours and knowledge points based on the knowledge graph, and sending the first mapping relationship to the user terminal; receiving a second mapping relationship sent by the user terminal, using the knowledge graph to obtain a third mapping relationship between class hours, knowledge points and courseware resources in the second mapping relationship, and sending the third mapping relationship to the user terminal, the second mapping relationship being obtained by the user terminal after editing the first mapping relationship; generating courseware based on a fourth mapping relationship sent by the user terminal, the fourth mapping relationship being obtained by the user terminal after screening the third mapping relationship. It can be seen that the above technical solution has the following problems: generating courseware based only on mapping relationships, lacking effective screening of knowledge points, and being unable to dynamically adjust knowledge points according to actual conditions, resulting in poor accuracy of generated courseware. Summary of the Invention

[0004] To this end, the present invention provides a method for intelligently generating courseware based on a knowledge graph, so as to overcome the problem in the prior art of lacking effective screening of knowledge points and being unable to dynamically adjust the knowledge points according to the actual status, resulting in poor accuracy of generated courseware.

[0005] To achieve the above objectives, the present invention provides a method for intelligently generating courseware based on a knowledge graph, comprising:

[0006] Determine a graph selection method based on the dynamic connection index and the graph interaction degree to generate a target graph, wherein the graph selection method is to select based on the feature interaction value alone or to select based on the explicit interaction value and the implicit coupling degree;

[0007] Determine the generation state according to the graph reference value and the learning characteristic value of the target graph, and determine the knowledge point selection method as node analysis selection or path analysis selection according to the generation state;

[0008] During node analysis and selection, the knowledge point supplementation method is determined based on the node evaluation value, which is node expansion supplementation or association enhancement supplementation;

[0009] In path analysis selection, the node status is determined based on the node complexity and node distribution coefficient, and the screening method is determined based on the node status, such as paragraph analysis selection or node analysis selection;

[0010] Under the selected completion conditions, the optimization method is determined based on the pre-replacement coefficient of the optimization range. The optimization method is to supplement knowledge points based on the missing coefficient or replace knowledge points based on the replacement efficiency threshold.

[0011] Fill in each knowledge point to generate courseware.

[0012] Furthermore, the graph selection method is determined based on the dynamic connection index and the graph interaction, including:

[0013] If the dynamic connection index is greater than or equal to the preset dynamic connection index and the graph interaction degree is greater than or equal to the preset graph interaction degree, the graph selection method is to select separately according to the feature interaction value;

[0014] If the dynamic connection index is less than the preset dynamic connection index or the graph interaction is less than the preset graph interaction, the graph selection method is to select based on the explicit interaction value and the implicit coupling degree.

[0015] Furthermore, if the generation status is that the graph reference value is greater than or equal to the preset graph reference value or the learning characteristic value is greater than or equal to the preset learning characteristic value, the knowledge point selection method is node analysis selection.

[0016] Furthermore, if the generated state is that the map reference value is less than the preset map reference value and the learning characteristic value is less than the preset learning characteristic value, the knowledge point selection method is path analysis selection.

[0017] Furthermore, if the node evaluation value is greater than or equal to the preset node evaluation value, the knowledge point supplement method is node expansion supplement;

[0018] In the node expansion and supplement, each related node and each node within the expansion depth corresponding to each related node are selected as knowledge points;

[0019] The expansion depth corresponding to a single associated node is positively correlated with the hierarchical quantitative index corresponding to the associated node.

[0020] Furthermore, if the node evaluation value is less than the preset node evaluation value, the knowledge point supplementation method is association enhancement supplementation;

[0021] In the association enhancement and supplementation, the node combination is determined based on the node association degree and semantic distance coefficient, the characteristic paragraph is determined according to the combination influence and feature coupling strength, and the association enhancement method of each characteristic paragraph is determined according to the path dependency coefficient to obtain several supplementary nodes. The supplementary nodes corresponding to each association node and each characteristic paragraph are used as selected knowledge points;

[0022] If the path dependency coefficient is greater than or equal to the preset path dependency coefficient, the association enhancement method is interval node enhancement supplement;

[0023] If the path dependence coefficient is less than the preset path dependence coefficient, the association enhancement method is the characteristic path enhancement supplement.

[0024] Furthermore, if the node status is that the node complexity is greater than or equal to the preset node complexity or the node distribution coefficient is greater than or equal to the preset node distribution coefficient, the screening method is paragraph analysis selection;

[0025] In paragraph analysis and selection, the associated paragraphs are determined based on the associated reference value, and the number of nodes to be selected is determined based on the paragraph intersection degree and paragraph feature coefficient;

[0026] The number of node selections corresponding to a single associated paragraph is positively correlated with the paragraph intersection degree and paragraph feature coefficient corresponding to the associated paragraph.

[0027] Furthermore, if the node status is that the node complexity is less than the preset node complexity and the node distribution coefficient is less than the preset node distribution coefficient, the screening method is node analysis selection;

[0028] In the node analysis selection, the nodes to be selected and the associated nodes whose analysis characteristic values ​​are greater than the preset analysis characteristic values ​​are selected as knowledge points.

[0029] Furthermore, the optimization method is determined according to the pre-replacement coefficient of the optimization range, including:

[0030] If the pre-replacement coefficient is less than the preset pre-replacement coefficient, the optimization method is to supplement the knowledge points according to the missing coefficient;

[0031] If the pre-replacement coefficient is greater than or equal to the preset pre-replacement coefficient, the optimization method is selected to replace the knowledge points according to the replacement efficiency threshold.

[0032] Furthermore, the method for confirming the optimization range includes:

[0033] Determine the initial optimization range based on the radiation impact coefficient and the courseware deviation threshold, and increase the optimization range based on the comparison range deviation value;

[0034] The initial optimization range after the increase adjustment is recorded as the optimization range.

[0035] Compared with the existing technology, the beneficial effect of the present invention lies in that, in the technical solution of the present invention, the content adaptability of the course and each knowledge graph is effectively reflected through the dynamic connection index and graph interactivity, and then different graph selection methods are adaptively selected according to the dynamic connection index and graph interactivity, so that the selected graph selection method can accurately adapt to the course requirements, avoid the omission of key knowledge or logical faults, and dynamically match the knowledge graph that best meets the course requirements, thereby improving the accuracy of courseware generation.

[0036] Furthermore, the present invention determines the generation state according to the graph reference value and learning characteristic value of the target graph, and effectively reflects the knowledge density and structural integrity of the target graph through the graph reference value and the learning characteristic value, and then adaptively selects different knowledge point selection methods according to the generation state. The node analysis selection can focus on the core knowledge points and reduce the cognitive load; the path analysis selection can build a complete knowledge chain to strengthen logical deduction, effectively avoiding the waste of resources caused by forced path analysis of low-quality graphs, and preventing the loose logic problem caused by only using node analysis of high-value graphs, which significantly improves the accuracy and adaptability of courseware generation and makes the knowledge point selection method more in line with actual application scenarios.

[0037] Furthermore, the present invention effectively reflects the actual status of valid nodes through node evaluation values, and then adaptively selects different knowledge point supplementation methods according to the node evaluation values, so that the selection of knowledge point supplementation methods is more in line with actual application scenarios and can expand key knowledge points. At the same time, it enhances the association of weak nodes, ensures the logical coherence of the knowledge graph, and thus improves the content quality and teaching adaptability of the courseware generation.

[0038] Furthermore, the present invention determines the node status according to the node complexity and the node distribution coefficient, and effectively reflects the complexity and distribution of the knowledge points through the node complexity and the node distribution coefficient, and then adaptively selects different screening methods according to the node status. When the node status is complex, selection through paragraph analysis can ensure that the path covers the core knowledge points and avoids logical faults. When the node status is simple, selection through endpoint analysis can strengthen the logical connection between simple nodes, ensuring that the generated courseware not only covers the core knowledge points but also maintains structural coherence.

[0039] Furthermore, the present invention effectively reflects the degree of replaceability of knowledge points in the optimization range through the pre-replacement coefficient, and then adaptively selects different optimization methods according to the pre-replacement coefficient, so that the selection of optimization method is more in line with the actual application scenario, and the optimal solution is selected according to the status of the knowledge point, thereby improving the optimization efficiency and thus improving the accuracy of courseware generation. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 Schematic diagram of the method for intelligently generating courseware based on knowledge graph of the present invention;

[0041] Figure 2 This is a flow chart of the present invention for determining a graph selection method based on a dynamic connection index and graph interactivity;

[0042] Figure 3 This is a flow chart of the present invention for determining a method for selecting knowledge points based on a generation state;

[0043] Figure 4 This is a flow chart of the present invention for determining a knowledge point supplement method based on node evaluation values. DETAILED DESCRIPTION

[0044] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention.

[0045] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0046] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside", and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0047] Furthermore, it should be noted that, in the description of the present invention, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integral connections; mechanical connections or electrical connections; direct connections or indirect connections through an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0048] See also Figures 1 to 4 As shown, the present invention provides a method for intelligently generating courseware based on a knowledge graph, comprising:

[0049] Determine a graph selection method based on the dynamic connection index and the graph interaction degree to generate a target graph, wherein the graph selection method is to select based on the feature interaction value alone or to select based on the explicit interaction value and the implicit coupling degree;

[0050] Determine the generation state according to the graph reference value and the learning characteristic value of the target graph, and determine the knowledge point selection method as node analysis selection or path analysis selection according to the generation state;

[0051] During node analysis and selection, the knowledge point supplementation method is determined based on the node evaluation value, which is node expansion supplementation or association enhancement supplementation;

[0052] In path analysis selection, the node status is determined based on the node complexity and node distribution coefficient, and the screening method is determined based on the node status, such as paragraph analysis selection or node analysis selection;

[0053] Under the selected completion conditions, the optimization method is determined based on the pre-replacement coefficient of the optimization range. The optimization method is to supplement knowledge points based on the missing coefficient or replace knowledge points based on the replacement efficiency threshold.

[0054] Fill in each knowledge point to generate courseware.

[0055] The application scenario of the present invention is to generate courseware using knowledge graphs. The present invention includes several knowledge graphs to be selected, and a single knowledge graph to be selected includes several entity keywords. Each entity keyword represents a node. A single node is connected to several nodes through an edge. The edge is used to describe the semantic logic between two nodes. The semantic logic includes but is not limited to belonging to, having, containing, and having. This is content that is easy to understand for those skilled in the art and will not be described in detail.

[0056] When generating courseware in the present invention, each courseware corresponds to courseware information. The courseware information includes, but is not limited to, the courseware theme, applicable objects, and teaching objectives. The courseware themes include, but are not limited to, "Basics of Python Functions," "Graphs and Properties of Quadratic Functions," and "Derivation and Application of the Sum Formula of Arithmetic Sequences." The applicable objects include grades and majors. The teaching objectives are texts that describe the knowledge, skills, and attitude goals that students should achieve by studying the courseware. This is content that is easily understood by those skilled in the art and will not be described in detail.

[0057] In the present invention, several historical records are correspondingly provided, and any historical record records the characteristic interaction value, explicit interaction value, implicit coupling degree, graph reference value and learning characteristic value, etc. in the historical process of generating courseware by using knowledge graph at least once, and each historical record corresponds to a qualified mark, which records whether the process of generating courseware by using knowledge graph meets the user requirements. The qualified mark can be recorded manually. It can be understood that the user can determine whether the process of generating courseware by using knowledge graph meets the requirements according to the self-set indicators. The self-set indicators can be but not limited to the error index, which will not be described in detail here. The error index is the number of courseware generated that does not conform to the course content;

[0058] The selection completion condition is that the knowledge points selected through the knowledge point selection method are selected;

[0059] Fill in each knowledge point to generate courseware, including: exporting the nodes corresponding to each selected courseware knowledge point and the edges corresponding to each node from the knowledge graph, exporting to CSV or JSON format, using Python scripts to batch generate courseware content templates, and importing the generated content into the PPT template to generate courseware. This is content that is easy for technical personnel in this field to understand and will not be described in detail.

[0060] Specifically, the graph selection method is determined based on the dynamic connection index and graph interactivity, including:

[0061] If the dynamic connection index is greater than or equal to the preset dynamic connection index and the graph interaction degree is greater than or equal to the preset graph interaction degree, the graph selection method is to select separately according to the feature interaction value;

[0062] If the dynamic connection index is less than the preset dynamic connection index or the graph interaction is less than the preset graph interaction, the graph selection method is to select based on the explicit interaction value and the implicit coupling degree.

[0063] The courseware information corresponding to the courseware currently being generated is recorded as the target courseware information, and each keyword in the target courseware information is recorded as the courseware keyword;

[0064] The dynamic connection index is the maximum value among the sub-connection indices corresponding to each knowledge graph to be selected. The sub-connection index is confirmed by recording a single knowledge graph to be selected as the target knowledge graph to be selected, and recording the average value of the similarity reference values ​​corresponding to each courseware keyword as the sub-connection index corresponding to the target knowledge graph to be selected; the similarity reference value corresponding to a single courseware keyword is the maximum value among the similarity coefficients corresponding to the courseware keyword and each entity keyword in the target knowledge graph to be selected;

[0065] The similarity coefficient is confirmed by converting any two keywords into word vectors. The cosine value of the angle between the two word vectors is recorded as the similarity coefficient corresponding to the two keywords. The keywords are converted into word vectors using a vector model. The vector model includes but is not limited to Word2Vec, GloVe, and BERT. This is easy to understand for those skilled in the art and will not be described in detail here.

[0066] The graph interaction degree is the average of the sub-graph interaction degrees corresponding to each knowledge graph to be selected. The sub-graph interaction degree corresponding to a single knowledge graph to be selected is the average of the graph association degrees corresponding to the knowledge graph to be selected and other knowledge graphs to be selected. The graph association degree corresponding to any two knowledge graphs to be selected = the larger value of the number of entity keywords that appear simultaneously in the two knowledge graphs to be selected / the number of entity keywords corresponding to the two knowledge graphs to be selected;

[0067] The values ​​of the preset dynamic connection index and the preset graph interactivity can be determined by the user based on the actual application scenario. The smaller the values ​​of the preset dynamic connection index and the preset graph interactivity, the greater the user's demand for selecting based on the feature interaction value. The preset dynamic connection index and the preset graph interactivity are provided as follows: the preset dynamic connection index is 0.7, and the preset graph interactivity is 70%;

[0068] Selecting separately according to the feature interaction value includes: recording the to-be-selected atlases whose feature interaction value is greater than the preset feature interaction value as selected atlases, and fusing each selected atlas to obtain the target atlas;

[0069] The tools that can be used to fuse each selected atlas include but are not limited to Falcon-AO and LiMES, which are well understood by those skilled in the art and will not be described in detail here.

[0070] Selection is performed based on the explicit interaction value and the implicit coupling degree, including: recording the knowledge graph to be selected with the largest feature interaction value as the target selection graph, recording other knowledge graphs to be selected outside the target selection graph as reference knowledge graphs to be selected, recording the reference knowledge graph to be selected whose explicit interaction value with the target selection graph is greater than a preset explicit interaction value or whose implicit coupling degree is greater than a preset implicit coupling degree as a reference selection graph, and fusing each reference selection graph and the target selection graph to obtain the target graph;

[0071] The entity keywords in the knowledge graph to be selected whose similarity coefficient with any courseware keyword is greater than 0.8 are recorded as associated keywords. It can be understood that each associated keyword corresponds to a courseware keyword, and the courseware keyword corresponding to a single associated keyword is the courseware keyword with a similarity coefficient greater than 0.8 with the associated keyword;

[0072] Explicit interaction value = the number of different courseware keywords in all courseware keywords corresponding to the associated keywords contained in the two to-be-selected knowledge graphs / the total number of courseware keywords in the target courseware information;

[0073] The implicit coupling degree is confirmed by, for any two knowledge graphs to be selected, respectively recorded as the first graph and the second graph, recording the same associated keywords in the two knowledge graphs to be selected as the same keywords, recording other entity keywords other than the same keywords as the interval keywords, and detecting the shortest paths in the two knowledge graphs to be selected that can contain the nodes corresponding to the same keywords, respectively recorded as the first path and the second path, the implicit coupling degree corresponding to the two knowledge graphs to be selected = the total number of the same interval keywords in the first path and the second path / the larger value of the path lengths corresponding to the first path and the second path, the path length corresponding to a single path is the total number of nodes contained in the path;

[0074] Feature interaction value = subgraph interaction degree × sub-connectivity index;

[0075] The values ​​of the preset feature interaction value, the preset explicit interaction value and the preset implicit coupling degree can be determined by the user according to the actual application scenario. The smaller the user's demand for improving the accuracy of the generated courseware, the larger the values ​​of the preset feature interaction value, the preset explicit interaction value and the preset implicit coupling degree. A value of a preset feature interaction value, a preset explicit interaction value and a preset implicit coupling degree is provided. The historical records selected by the user based on the feature interaction value are detected, and the average value of the feature interaction values ​​corresponding to the historical records that can meet the user's needs is recorded as the preset feature interaction value. The historical records selected by the user based on the explicit interaction value and the implicit coupling degree are detected, and the average value of the explicit interaction values ​​corresponding to the historical records that can meet the user's needs is recorded as the preset explicit interaction value, and the average value of the implicit coupling degrees corresponding to the historical records that can meet the user's needs is recorded as the preset implicit coupling degree.

[0076] Specifically, if the generation status is that the graph reference value is greater than or equal to the preset graph reference value or the learning characteristic value is greater than or equal to the preset learning characteristic value, the knowledge point selection method is node analysis selection.

[0077] The generation state includes a first generation state and a second generation state. The first generation state is that the map reference value is greater than or equal to the preset map reference value or the learning characteristic value is greater than or equal to the preset learning characteristic value. The second generation state is that the map reference value is less than the preset map reference value and the learning characteristic value is less than the preset learning characteristic value.

[0078] The graph reference value is the average of the node reference values ​​corresponding to each node in the target graph. The node reference value corresponding to a single node is the number of nodes directly connected to the node.

[0079] The nodes corresponding to the associated keywords in the target graph are recorded as associated nodes, and the learning feature value = the sum of the node reference values ​​corresponding to each associated node / (the number of associated nodes + 1);

[0080] The values ​​of the preset graph reference value and the preset learning characteristic value can be determined by the user according to the actual application scenario. The smaller the values ​​of the preset graph reference value and the preset learning characteristic value, the greater the user's demand for node analysis and selection. A preset graph reference value and a preset learning characteristic value are provided to detect the historical records of the user's node analysis and selection, and the average value of the graph reference value corresponding to the historical records that can meet the user's needs is recorded as the preset graph reference value, and the average value of the learning characteristic value corresponding to the historical records that can meet the user's needs is recorded as the preset learning characteristic value.

[0081] Specifically, if the generated status is that the map reference value is less than the preset map reference value and the learning characteristic value is less than the preset learning characteristic value, the knowledge point selection method is path analysis selection.

[0082] Specifically, if the node evaluation value is greater than or equal to the preset node evaluation value, the knowledge point supplement method is node expansion supplement;

[0083] In the node expansion and supplement, each related node and each node within the expansion depth corresponding to each related node are selected as knowledge points;

[0084] The expansion depth corresponding to a single associated node is positively correlated with the hierarchical quantitative index corresponding to the associated node.

[0085] The node evaluation value is the average of the sub-evaluation values ​​corresponding to each associated node, and the sub-evaluation value corresponding to a single associated node is the number of associated nodes directly connected to the associated node;

[0086] The value of the preset node evaluation value can be determined by the user according to the actual application scenario. The smaller the value of the preset node evaluation value, the greater the user's demand for node expansion and supplementation. A value of the preset node evaluation value is provided, and the historical records of the user's node expansion and supplementation are detected. The average value of the node evaluation values ​​corresponding to the historical records that can meet the user's needs is recorded as the preset node evaluation value;

[0087] The method for confirming the hierarchical quantitative index is as follows: for a single associated node, the associated node is recorded as the first target associated node, and the other nodes in the target graph except the first target associated node are recorded as analysis nodes. The hierarchical quantitative index corresponding to the first target associated node = the number of analysis nodes directly connected to the first target associated node / the average value of the semantic distance coefficients corresponding to each analysis node and the first target associated node;

[0088] The expansion depth corresponding to the first target associated node is the range of nodes that can be reached by moving n times along the edge starting from the first target associated node, and n is positively correlated with the hierarchical quantitative index corresponding to the first target associated node.

[0089] Specifically, if the node evaluation value is less than the preset node evaluation value, the knowledge point supplement method is association enhancement supplement;

[0090] In the association enhancement and supplementation, the node combination is determined based on the node association degree and semantic distance coefficient, the characteristic paragraph is determined according to the combination influence and feature coupling strength, and the association enhancement method of each characteristic paragraph is determined according to the path dependency coefficient to obtain several supplementary nodes. The supplementary nodes corresponding to each association node and each characteristic paragraph are used as selected knowledge points;

[0091] If the path dependency coefficient is greater than or equal to the preset path dependency coefficient, the association enhancement method is interval node enhancement supplement;

[0092] If the path dependence coefficient is less than the preset path dependence coefficient, the association enhancement method is the characteristic path enhancement supplement.

[0093] Wherein, determining the node combination based on the node association degree and the semantic distance coefficient includes: performing a combination analysis on each associated node, when performing a combination analysis on a single associated node, recording the associated node as a target node, recording other associated nodes other than the target node that are not recorded in the node combination as reference nodes, recording a combination of the reference node and the target node whose node association degree with the target node is greater than a preset node association degree and whose semantic distance coefficient is less than a preset semantic distance coefficient as a node combination, and continuing to perform a combination analysis on the associated nodes that are not recorded in the node combination until all associated nodes are recorded in the node combination, and then stopping the combination analysis;

[0094] The method of confirming the node association degree is that, for any two nodes, the number of paths with the two nodes as the end points is recorded as the node association degree corresponding to the two nodes;

[0095] The semantic distance coefficient corresponding to any two nodes is the total number of nodes contained in the shortest path with the two nodes as the end points;

[0096] The values ​​of the preset node association degree and the preset semantic distance coefficient can be determined by the user according to the actual application scenario. The greater the user's demand for improving the accuracy of courseware generation, the greater the value of the preset node association degree and the smaller the value of the preset semantic distance coefficient. The values ​​of the preset node association degree and the preset semantic distance coefficient are provided, and the historical records of the user determining the node combination based on the node association degree and the semantic distance coefficient are detected. The average value of the reference node association degree corresponding to each node combination in the historical records that can meet the user's needs is recorded as the preset node association degree, and the average value of the reference semantic distance coefficient corresponding to each node combination in the historical records that can meet the user's needs is recorded as the preset semantic distance coefficient. The reference node association degree corresponding to a single node combination is the node association degree corresponding to any two nodes in the node combination, and the reference semantic distance coefficient corresponding to a single node combination is the semantic distance coefficient corresponding to any two nodes in the node combination;

[0097] Determining a characteristic segment according to the combined influence and the characteristic coupling strength includes: recording the associated node with the largest combined influence in each node combination as a characteristic node, performing matching analysis on each characteristic node, recording the characteristic node as a target characteristic node when performing matching analysis on a single characteristic node, recording other characteristic nodes other than the target characteristic node that are not recorded as matching nodes as reference characteristic nodes, taking the reference characteristic node with the largest characteristic coupling strength with the target characteristic node as the matching node of the target characteristic node, and continuing to perform matching analysis on characteristic nodes where no matching nodes exist until a preset condition is met, then stopping the matching analysis. It should be noted that when the number of node combinations is an even number, the preset condition is that each characteristic node has a matching node, and when the number of node combinations is an odd number, the preset condition is that only one characteristic node remains without a matching node; each characteristic node where a matching node exists is recorded as a characteristic reference node;

[0098] Each feature reference node corresponds to a feature paragraph. The feature paragraph corresponding to a single feature reference node is the nodes included in the shortest path with the feature reference node and the matching node corresponding to the feature reference node as the two end points.

[0099] The method for confirming the combined influence is to record a single associated node in a single node combination as the first target node, and record the other nodes in the node combination except the first target node as the first reference nodes. The combined influence corresponding to the first target node = the average value of the semantic distance coefficients corresponding to the first target node and each first reference node × the average value of the node association degrees corresponding to the first target node and each first reference node;

[0100] The feature coupling strength is confirmed as follows: for any two feature nodes, the feature coupling strength corresponding to the two feature nodes = the similarity coefficient corresponding to the entity keywords corresponding to the two feature nodes × the total number of nodes included in the shortest path with the two feature nodes as the two end points;

[0101] The path dependency coefficient is confirmed by recording the two feature nodes corresponding to a single feature paragraph as analysis feature nodes, and recording the path with the two analysis feature nodes as the two end points as the analysis path. The path dependency coefficient = the average value of the feature reference values ​​corresponding to each analysis path / the standard deviation of the feature reference values ​​corresponding to each analysis path;

[0102] The value of the preset path dependency coefficient can be determined by the user according to the actual application scenario. The smaller the value of the preset path dependency coefficient, the greater the user's demand for feature path enhancement and supplementation. A value of the preset path dependency coefficient is provided, and the historical records of the user's feature path enhancement and supplementation are detected. The average value of the path dependency coefficients corresponding to the historical records that can meet the user's needs is recorded as the preset path dependency coefficient;

[0103] Interval node enhancement and supplementation, including: taking the analysis path whose characteristic reference value is greater than the preset characteristic reference value as the path to be selected, selecting the nodes in the path to be selected according to the node selection interval, and taking the nodes selected in each path to be selected as supplementary nodes;

[0104] The node selection interval corresponding to a single path to be selected is negatively correlated with the characteristic reference value corresponding to the path to be selected;

[0105] Feature path enhancement and supplementation, including: taking the analysis path with the largest feature reference value as the feature path, and taking each node in the feature path as a supplementary node;

[0106] The node selection interval is the number of nodes between two adjacent supplementary nodes selected in a single path to be selected;

[0107] The characteristic reference value is confirmed in the following way: for a single analysis path, the analysis path is recorded as the target analysis path, the nodes in the target analysis path are recorded as target path nodes, and the nodes in other analysis paths outside the target analysis path are recorded as reference path nodes. The characteristic reference value corresponding to the target analysis path = the number of reference path nodes directly connected to the target path node / the number of target path nodes;

[0108] The value of the preset feature reference value can be determined by the user according to the actual application scenario. The greater the user's demand for improving the correlation between knowledge points, the smaller the value of the preset feature reference value is. A value of a preset feature reference value is provided, and the historical records of the user's interval node enhancement and supplementation are detected. The average value of the feature reference values ​​corresponding to the historical records that can meet the user's needs is recorded as the preset feature reference value.

[0109] Specifically, if the node status is that the node complexity is greater than or equal to the preset node complexity or the node distribution coefficient is greater than or equal to the preset node distribution coefficient, the screening method is paragraph analysis selection;

[0110] In paragraph analysis and selection, the associated paragraphs are determined based on the associated reference value, and the number of nodes to be selected is determined based on the paragraph intersection degree and paragraph feature coefficient;

[0111] The number of node selections corresponding to a single associated paragraph is positively correlated with the paragraph intersection degree and paragraph feature coefficient corresponding to the associated paragraph.

[0112] The node state includes a first node state and a second node state. The first node state is that the node complexity is greater than or equal to the preset node complexity or the node distribution coefficient is greater than or equal to the preset node distribution coefficient. The second node state is that the node complexity is less than the preset node complexity and the node distribution coefficient is less than the preset node distribution coefficient.

[0113] The node complexity is the number of associated nodes in the target graph, the node distribution coefficient is the average value of the node influence corresponding to each associated node, for a single associated node, the associated node is recorded as the target associated node, the other associated nodes other than the target associated node are recorded as reference associated nodes, and the minimum value of the semantic distance coefficient between the target associated node and each reference associated node is recorded as the node influence corresponding to the target associated node;

[0114] The values ​​of the preset node complexity and the preset node distribution coefficient can be determined by the user according to the actual application scenario. The smaller the values ​​of the preset node complexity and the preset node distribution coefficient, the greater the user's demand for paragraph analysis and selection. The values ​​of the preset node complexity and the preset node distribution coefficient are provided, and the historical records of the user's paragraph analysis and selection are detected. The average value of the node complexity corresponding to the historical records that can meet the user's needs is recorded as the preset node complexity, and the average value of the node distribution coefficient corresponding to the historical records that can meet the user's needs is recorded as the preset node distribution coefficient;

[0115] Determining an associated paragraph according to an associated reference value includes: recording other nodes other than the associated nodes that appear in the shortest path that can contain each associated node as to-be-selected nodes, performing an associated analysis on each to-be-selected node, when performing an associated analysis on a single to-be-selected node, recording the to-be-selected node as a target to-be-selected node, recording to-be-selected nodes other than the target to-be-selected node and not recorded in the associated paragraph as reference to-be-selected nodes, recording a set of reference to-be-selected nodes and the target to-be-selected node whose associated reference values ​​with the target to-be-selected node are greater than a preset associated reference value as an associated paragraph, and continuing to perform an associated analysis on the to-be-selected nodes that are not recorded in the associated paragraph until all to-be-selected nodes are recorded in the associated paragraph, then stopping the associated analysis;

[0116] The paragraph intersection degree corresponding to a single associated paragraph is the number of other nodes to be selected that are directly connected to each node to be selected in the associated paragraph;

[0117] The paragraph feature coefficient corresponding to a single associated paragraph = the number of nodes to be selected contained in the associated paragraph × the number of nodes contained in the shortest path that can contain each node to be selected in the associated paragraph;

[0118] The number of nodes selected for a single associated paragraph is the number of nodes to be selected in the associated paragraph. The number of nodes selected for a single associated paragraph is positively correlated with the paragraph evaluation value. The paragraph evaluation value = the paragraph intersection degree corresponding to the associated paragraph + the paragraph feature coefficient corresponding to the associated paragraph.

[0119] Each node to be selected and each associated node selected in each associated paragraph is used as a selected knowledge point. Each node to be selected in a single associated paragraph is a node randomly selected according to the node selection amount corresponding to the associated paragraph.

[0120] The method for confirming the association reference value is as follows: for any two nodes, respectively recorded as the first node and the second node, the node directly connected to the first node is recorded as the first reference node, and the node directly connected to the second node is recorded as the second reference node. The association reference value = the number of nodes that are the same in the first reference node and the second reference node / the number of nodes included in the shortest path with the first node and the second node as the endpoints;

[0121] The value of the preset associated reference value can be determined by the user according to the actual application scenario. The greater the user's demand for improving the accuracy of courseware generation, the larger the value of the preset associated reference value is. A preset associated reference value is provided to detect the historical records of the user determining the associated paragraphs based on the associated reference value. The average value of the reference associated reference values ​​corresponding to each associated paragraph in the historical records that can meet the user's needs is recorded as the preset associated reference value. The reference associated reference value corresponding to a single associated paragraph is the associated reference value corresponding to any two nodes in the associated paragraph.

[0122] Specifically, if the node status is that the node complexity is less than the preset node complexity and the node distribution coefficient is less than the preset node distribution coefficient, the screening method is node analysis selection;

[0123] In the node analysis selection, the nodes to be selected and the associated nodes whose analysis characteristic values ​​are greater than the preset analysis characteristic values ​​are selected as knowledge points.

[0124] The method for confirming the analysis feature value is as follows: for a single node to be selected, the node to be selected is recorded as the target node to be selected, and the analysis feature value corresponding to the target node to be selected = the number of other nodes to be selected directly connected to the target node to be selected / the average value of the semantic distance coefficients corresponding to the target node to be selected and each associated node;

[0125] The value of the preset analysis characteristic value can be determined by the user according to the actual application scenario. The greater the user's demand for improving the accuracy of the courseware, the larger the value of the preset analysis characteristic value is. A value of a preset analysis characteristic value is provided to detect the historical records of the user's node analysis selection, and the average value of the analysis characteristic values ​​corresponding to the nodes as supplementary knowledge points in the historical records that can meet the user's needs is recorded as the preset analysis characteristic value.

[0126] Specifically, the optimization method is determined according to the pre-replacement coefficient of the optimization range, including:

[0127] If the pre-replacement coefficient is less than the preset pre-replacement coefficient, the optimization method is to supplement the knowledge points according to the missing coefficient;

[0128] If the pre-replacement coefficient is greater than or equal to the preset pre-replacement coefficient, the optimization method is selected to replace the knowledge points according to the replacement efficiency threshold.

[0129] The node corresponding to each selected knowledge point is recorded as a selection node, the pre-replacement coefficient is the average value of the sub-replacement coefficients corresponding to each selection node, and the sub-replacement coefficient is confirmed in the following way: for a single selection node, the selection node is recorded as the target selection node, and other selection nodes other than the target selection node are recorded as reference selection nodes. The historical records that can meet the user's needs and in which the target selection node appears are detected and recorded as reference historical records. The sub-replacement coefficient corresponding to the target selection node is the average value of the number of reference selection nodes that appear in each reference historical record. It should be noted that if there is no reference historical record for the target selection node, the sub-replacement coefficient corresponding to the target selection node is 0.

[0130] The value of the preset pre-replacement coefficient can be determined by the user according to the actual application scenario. The larger the value of the preset pre-replacement coefficient, the greater the user's demand for supplementing knowledge points based on the missing coefficient. A value of the preset pre-replacement coefficient is provided, and the historical records of users supplementing knowledge points based on the missing coefficient are detected. The average value of the pre-replacement coefficients corresponding to the historical records that can meet the user's needs is recorded as the preset pre-replacement coefficient;

[0131] Supplementing knowledge points based on the missing coefficient includes: supplementing knowledge points for selected nodes with missing coefficients greater than a preset missing coefficient; when supplementing knowledge points for a single selected node, the expansion depth corresponding to the selected node is positively correlated with the missing coefficient corresponding to the selected node; the expansion depth corresponding to the selected node is the range of nodes that can be reached by moving m times along the edge from the selected node, and m is positively correlated with the missing coefficient corresponding to the selected node;

[0132] Each node within the expansion depth corresponding to the selected node with a missing coefficient greater than the preset missing coefficient is selected as a supplementary selected knowledge point, and each supplementary selected knowledge point and each selected knowledge point is used as a courseware knowledge point;

[0133] Replacing knowledge points according to a replacement efficiency threshold includes: replacing knowledge points for selected nodes whose sub-replacement coefficient is greater than a preset sub-replacement coefficient; when replacing knowledge points for a single selected node, deleting the selected node and using the unselected node with the largest replacement efficiency threshold corresponding to the selected node as a supplementary selected knowledge point; and using the supplementary selected knowledge points and the selected knowledge points that have not been replaced as courseware knowledge points;

[0134] The missing coefficient is confirmed by recording a single selection node as the target selection node. The missing coefficient corresponding to the target selection node = the number of selection nodes directly connected to the target selection node / the total number of nodes connected to the target selection node.

[0135] The values ​​of the preset missing coefficient and the preset sub-replacement coefficient can be determined by the user according to the actual application scenario. The greater the user's demand for improving the accuracy of the generated courseware, the smaller the values ​​of the preset missing coefficient and the preset sub-replacement coefficient are. A value of the preset missing coefficient and the preset sub-replacement coefficient is provided. The preset missing coefficient is 50%. The historical records of the user supplementing knowledge points according to the missing coefficient are detected, and the average value of the sub-replacement coefficients corresponding to the selected nodes for supplementing knowledge points in the historical records that can meet the user's needs is recorded as the preset sub-replacement coefficient;

[0136] The replacement efficiency threshold is confirmed in the following way: other nodes except the selected node in the target graph are recorded as unselected nodes, and unselected nodes directly connected to the selected node are recorded as adjacent unselected nodes. For a single selected node, the selected node is recorded as the target selected node. For a single adjacent unselected node directly connected to the target selected node, the adjacent unselected node is recorded as the target adjacent unselected node. The replacement efficiency threshold corresponding to the target adjacent unselected node and the target selected node = the similarity coefficient between the entity keyword corresponding to the target adjacent unselected node and the entity keyword corresponding to the target selected node / the average value of the semantic distance coefficients corresponding to the target adjacent unselected node and each selected node.

[0137] Specifically, the method for confirming the optimization range includes:

[0138] Determine the initial optimization range based on the radiation impact coefficient and the courseware deviation threshold, and increase the optimization range based on the comparison range deviation value;

[0139] The initial optimization range after the increase adjustment is recorded as the optimization range.

[0140] The radiation influence coefficient is confirmed by recording the nodes corresponding to each selected knowledge point as selected nodes, and recording the other nodes in the selected nodes except the associated nodes as non-associated nodes. The radiation influence coefficient = the average value of the sub-radiation influence coefficients corresponding to each non-associated node / the standard deviation of the sub-radiation influence coefficients corresponding to each non-associated node. The sub-radiation influence coefficient corresponding to a single non-associated node is the average value of the semantic distance coefficients corresponding to the non-associated node and each associated node.

[0141] The courseware deviation threshold is confirmed in the following way: for a single non-associated node, the maximum value of the similarity coefficients between the entity keyword corresponding to the non-associated node and the entity keyword corresponding to each associated node is recorded as the courseware deviation threshold corresponding to the non-associated node;

[0142] When determining the initial optimization range based on the radiation influence coefficient and the courseware deviation threshold, the number of non-associated nodes in the initial optimization range is positively correlated with the radiation influence coefficient, and the initial optimization range is the non-associated nodes selected in descending order according to the courseware deviation threshold;

[0143] Comparison range deviation value = |Comparison range reference value corresponding to the courseware currently being generated - average comparison range reference values ​​corresponding to historical records that can meet user needs|, non-associated nodes outside the initial optimization range are recorded as analysis non-associated nodes, and associated nodes within the initial optimization range are recorded as analysis associated nodes. Comparison range reference value = average value of connection coefficients corresponding to each analysis non-associated node × number of analysis non-associated nodes. The connection coefficient corresponding to a single analysis non-associated node is the number of analysis associated nodes directly connected to the analysis non-associated node;

[0144] The increase in the optimization range is positively correlated with the deviation in the comparison range. The increase in the optimization range = the number of non-associated nodes included in the optimization range - the number of non-associated nodes included in the initial optimization range.

[0145] According to the courseware deviation threshold, non-correlated nodes are selected for analysis in descending order until the increase value of the optimization range is reached, and the selected non-correlated nodes for analysis and the related nodes for analysis in the initial optimization range are used as the optimization range.

[0146] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

[0147] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that the present invention is susceptible to various modifications and variations. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A method for intelligently generating courseware based on knowledge graph, characterized in that: include: Determine a graph selection method based on the dynamic connection index and the graph interaction degree to generate a target graph, wherein the graph selection method is to select based on the feature interaction value alone or to select based on the explicit interaction value and the implicit coupling degree; Determine the generation state according to the graph reference value and the learning characteristic value of the target graph, and determine the knowledge point selection method as node analysis selection or path analysis selection according to the generation state; During node analysis and selection, the knowledge point supplementation method is determined based on the node evaluation value, which is node expansion supplementation or association enhancement supplementation; In path analysis selection, the node status is determined based on the node complexity and node distribution coefficient, and the screening method is determined based on the node status, such as paragraph analysis selection or node analysis selection; Under the selected completion conditions, the optimization method is determined based on the pre-replacement coefficient of the optimization range. The optimization method is to supplement knowledge points based on the missing coefficient or replace knowledge points based on the replacement efficiency threshold. Fill in each knowledge point to generate courseware; The dynamic connection index is the maximum value among the sub-connection indices corresponding to each knowledge graph to be selected. The sub-connection index is confirmed by recording a single knowledge graph to be selected as the target knowledge graph to be selected, and recording the average value of the similarity reference values ​​corresponding to each courseware keyword as the sub-connection index corresponding to the target knowledge graph to be selected; the similarity reference value corresponding to a single courseware keyword is the maximum value among the similarity coefficients corresponding to the courseware keyword and each entity keyword in the target knowledge graph to be selected; The graph interaction degree is the average of the sub-graph interaction degrees corresponding to each knowledge graph to be selected. The sub-graph interaction degree corresponding to a single knowledge graph to be selected is the average of the graph association degrees corresponding to the knowledge graph to be selected and other knowledge graphs to be selected. The graph association degree corresponding to any two knowledge graphs to be selected = the larger value of the number of entity keywords that appear simultaneously in the two knowledge graphs to be selected / the number of entity keywords corresponding to the two knowledge graphs to be selected; Explicit interaction value = the number of different courseware keywords in all courseware keywords corresponding to the associated keywords contained in the two to-be-selected knowledge graphs / the total number of courseware keywords in the target courseware information; The implicit coupling degree is confirmed by, for any two knowledge graphs to be selected, respectively recorded as the first graph and the second graph, recording the same associated keywords in the two knowledge graphs to be selected as the same keywords, recording other entity keywords other than the same keywords as the interval keywords, and detecting the shortest paths in the two knowledge graphs to be selected that can contain the nodes corresponding to the same keywords, respectively recorded as the first path and the second path, the implicit coupling degree corresponding to the two knowledge graphs to be selected = the total number of the same interval keywords in the first path and the second path / the larger value of the path lengths corresponding to the first path and the second path, the path length corresponding to a single path is the total number of nodes contained in the path; The graph reference value is the average of the node reference values ​​corresponding to each node in the target graph. The node reference value corresponding to a single node is the number of nodes directly connected to the node. Learning feature value = sum of node reference values ​​corresponding to each associated node / (number of associated nodes + 1); The node complexity is the number of associated nodes in the target graph, the node distribution coefficient is the average value of the node influence corresponding to each associated node, for a single associated node, the associated node is recorded as the target associated node, the other associated nodes other than the target associated node are recorded as reference associated nodes, and the minimum value of the semantic distance coefficient between the target associated node and each reference associated node is recorded as the node influence corresponding to the target associated node; The replacement effectiveness threshold is determined by recording all nodes other than the selected node in the target graph as unselected nodes, recording the unselected nodes directly connected to the selected node as adjacent unselected nodes, and for a single selected node, recording that selected node as the target selected node. For a single adjacent unselected node directly connected to the target selected node, recording that adjacent unselected node as the target adjacent unselected node. The replacement effectiveness threshold corresponding to the target adjacent unselected node and the target selected node = the similarity coefficient between the entity keyword corresponding to the target adjacent unselected node and the entity keyword corresponding to the target selected node / the average of the semantic distance coefficients corresponding to the target adjacent unselected node and each selected node; The node evaluation value is the average of the sub-evaluation values ​​corresponding to each associated node. The sub-evaluation value corresponding to a single associated node is the number of associated nodes directly connected to the associated node. The pre-replacement coefficient is the average of the sub-replacement coefficients corresponding to each selection node. The sub-replacement coefficient is determined by recording a single selection node as the target selection node and other selection nodes other than the target selection node as reference selection nodes. Historical records that meet user requirements and in which the target selection node appears are detected and recorded as reference historical records. The sub-replacement coefficient corresponding to the target selection node is the average of the number of reference selection nodes that appear in each reference historical record. The missing coefficient is confirmed by recording a single selection node as the target selection node, and the missing coefficient corresponding to the target selection node = the number of selection nodes directly connected to the target selection node / the total number of nodes connected to the target selection node.

2. The method for intelligently generating courseware based on knowledge graph according to claim 1, characterized in that: The graph selection method is determined based on the dynamic connection index and graph interactivity, including: If the dynamic connection index is greater than or equal to the preset dynamic connection index and the graph interaction degree is greater than or equal to the preset graph interaction degree, the graph selection method is to select separately according to the feature interaction value; If the dynamic connection index is less than the preset dynamic connection index or the graph interaction is less than the preset graph interaction, the graph selection method is to select based on the explicit interaction value and the implicit coupling degree.

3. The method for intelligently generating courseware based on knowledge graph according to claim 2, characterized in that: If the generation status is that the graph reference value is greater than or equal to the preset graph reference value or the learning characteristic value is greater than or equal to the preset learning characteristic value, the knowledge point selection method is node analysis selection.

4. The method for intelligently generating courseware based on knowledge graph according to claim 3, characterized in that: If the generated status is that the map reference value is less than the preset map reference value and the learning characteristic value is less than the preset learning characteristic value, the knowledge point selection method is path analysis selection.

5. The method for intelligently generating courseware based on knowledge graph according to claim 3, characterized in that: If the node evaluation value is greater than or equal to the preset node evaluation value, the knowledge point supplement method is node expansion supplement; In the node expansion and supplement, each related node and each node within the expansion depth corresponding to each related node are selected as knowledge points; The expansion depth corresponding to a single associated node is positively correlated with the hierarchical quantitative index corresponding to the associated node.

6. The method for intelligently generating courseware based on knowledge graph according to claim 3, characterized in that: If the node evaluation value is less than the preset node evaluation value, the knowledge point supplement method is association enhancement supplement; In the association enhancement and supplementation, the node combination is determined based on the node association degree and semantic distance coefficient, the characteristic paragraph is determined according to the combination influence and feature coupling strength, and the association enhancement method of each characteristic paragraph is determined according to the path dependency coefficient to obtain several supplementary nodes. The supplementary nodes corresponding to each association node and each characteristic paragraph are used as selected knowledge points; If the path dependency coefficient is greater than or equal to the preset path dependency coefficient, the association enhancement method is interval node enhancement supplement; If the path dependence coefficient is less than the preset path dependence coefficient, the association enhancement method is the characteristic path enhancement supplement; The method for confirming the combined influence is to record a single associated node in a single node combination as the first target node, and record the other nodes in the node combination except the first target node as the first reference nodes. The combined influence corresponding to the first target node = the average value of the semantic distance coefficients corresponding to the first target node and each first reference node × the average value of the node association degrees corresponding to the first target node and each first reference node; The feature coupling strength is confirmed as follows: for any two feature nodes, the feature coupling strength corresponding to the two feature nodes = the similarity coefficient corresponding to the entity keywords corresponding to the two feature nodes × the total number of nodes included in the shortest path with the two feature nodes as the two end points; The path dependence coefficient is confirmed in the following way: for a single characteristic paragraph, the two characteristic nodes corresponding to the characteristic paragraph are recorded as analysis characteristic nodes, and the path with the two analysis characteristic nodes as the two end points is recorded as the analysis path. The path dependence coefficient = the average value of the characteristic reference values ​​corresponding to each analysis path / the standard deviation of the characteristic reference values ​​corresponding to each analysis path.

7. The method for intelligently generating courseware based on knowledge graph according to claim 4, characterized in that: If the node status is that the node complexity is greater than or equal to the preset node complexity or the node distribution coefficient is greater than or equal to the preset node distribution coefficient, the screening method is paragraph analysis selection; In paragraph analysis and selection, the associated paragraphs are determined based on the associated reference value, and the number of nodes to be selected is determined based on the paragraph intersection degree and paragraph feature coefficient; The number of nodes selected for a single associated paragraph is positively correlated with the paragraph intersection degree and paragraph feature coefficient corresponding to the associated paragraph; The paragraph intersection degree corresponding to a single associated paragraph is the number of other nodes to be selected that are directly connected to each node to be selected in the associated paragraph; The paragraph feature coefficient corresponding to a single associated paragraph = the number of nodes to be selected contained in the associated paragraph × the number of nodes contained in the shortest path that can contain each node to be selected in the associated paragraph.

8. The method for intelligently generating courseware based on knowledge graph according to claim 4, characterized in that: If the node status is that the node complexity is less than the preset node complexity and the node distribution coefficient is less than the preset node distribution coefficient, the screening method is node analysis selection; In the node analysis selection, the nodes to be selected and the associated nodes whose analysis characteristic values ​​are greater than the preset analysis characteristic values ​​are selected as knowledge points.

9. The method for intelligently generating courseware based on knowledge graph according to claim 8, characterized in that: The optimization method is determined based on the pre-replacement coefficient of the optimization range, including: If the pre-replacement coefficient is less than the preset pre-replacement coefficient, the optimization method is to supplement the knowledge points according to the missing coefficient; If the pre-replacement coefficient is greater than or equal to the preset pre-replacement coefficient, the optimization method is selected to replace the knowledge points according to the replacement efficiency threshold.

10. The method for intelligently generating courseware based on knowledge graph according to claim 9, characterized in that: Methods for confirming the optimization range include: Determine the initial optimization range based on the radiation impact coefficient and the courseware deviation threshold, and increase the optimization range based on the comparison range deviation value; The initial optimization range completed by increasing the adjustment is recorded as the optimization range; The radiation influence coefficient is confirmed by recording the nodes corresponding to each selected knowledge point as selected nodes, and recording the other nodes in the selected nodes except the associated nodes as non-associated nodes. The radiation influence coefficient = the average value of the sub-radiation influence coefficients corresponding to each non-associated node / the standard deviation of the sub-radiation influence coefficients corresponding to each non-associated node. The sub-radiation influence coefficient corresponding to a single non-associated node is the average value of the semantic distance coefficients corresponding to the non-associated node and each associated node. The courseware deviation threshold is confirmed by, for a single non-associated node, recording the maximum value of the similarity coefficients between the entity keyword corresponding to the non-associated node and the entity keyword corresponding to each associated node as the courseware deviation threshold corresponding to the non-associated node.

Images