A community consciousness education effectiveness intelligent evaluation method based on multi-source data perception

Abstract

The present application relates to the technical field of teaching management, in particular to a kind of community consciousness education effectiveness intelligent evaluation method based on multi-source data perception, through the collection multi-source data in the education process, extract cognitive characteristics and construct time sequence, according to community consciousness element, construct time sequence knowledge graph containing cognitive state node and cognitive state transition edge, calculate cognitive depth score by time period time sequence comparison, to match cognitive state and generate cognitive state transition path, locate cognitive state change time point, get trigger event, combined with the integration degree and influence score calculated by graph neural network, complete education effectiveness evaluation.The present application realizes the dynamic tracking, traceable attribution and quantitative evaluation of community consciousness education effectiveness, improves the accuracy of evaluation results, and provides a reliable basis for the improvement of community consciousness education.

Description

Technical Field

[0001] This invention relates to the field of teaching management technology, and in particular to an intelligent evaluation method for the effectiveness of community awareness education based on multi-source data perception. Background Technology

[0002] The cognition of community consciousness education is a dynamic psychological process that changes as education progresses. It gradually improves or declines during classroom discussions, collaborative learning, and thematic education. In this process, there are clear patterns of cognitive state transition, such as from vague cognition to clear cognition, and from passive identification to active identification. Moreover, this transition of cognitive state is usually triggered by specific educational events and interactive behaviors. Existing technologies, by constructing static knowledge graphs and evaluation models, use single data from a single period for evaluation. This cannot reflect the true development trajectory of cognitive state, let alone determine the path of cognitive transition from one state to another. In other words, the evaluation results are disconnected from the related events in the educational process, and the key educational links that trigger cognitive improvement or decline cannot be identified, resulting in inaccurate evaluation results of educational effectiveness.

[0003] Therefore, it is necessary to propose an intelligent evaluation method for the effectiveness of community awareness education based on multi-source data perception to solve the above problems. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent evaluation method for the effectiveness of community awareness education based on multi-source data perception, in order to solve the problem that existing technologies cannot reflect the true development trajectory of cognitive states, let alone determine the path of cognition from one state to another.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A smart evaluation method for the effectiveness of community awareness education based on multi-source data perception includes the following steps: Identify multi-source data in the educational process, and extract time-stamped cognitive features from the multi-source data based on the timeline of the educational process to obtain a temporal sequence of cognitive features; Based on the elements of community consciousness, a knowledge graph is constructed as a standard for cognitive assessment. Cognitive state nodes and cognitive state transition edges are set in the knowledge graph, and the same timeline as the education process is added to obtain a temporal knowledge graph. Based on the timeline, the time sequence is compared with the time sequence knowledge graph for each time period to calculate the cognitive depth score for the corresponding time period. The score is then compared with the cognitive state node in real time to obtain the cognitive state for the current time period. Combined with the cognitive state transition edge and the continuous change of the cognitive state on the timeline, a cognitive state transition path is generated. Traverse the time sequence, combine the cognitive state transition path, locate the time point of cognitive state change, match the time point with multi-source data, and obtain the corresponding triggering event. Based on interactive network data of student collaboration, a graph neural network is used to calculate the integration and influence scores. Combined with the cognitive depth score, cognitive state transition path and triggering event, the evaluation results are obtained.

[0006] Preferably, the step of constructing a knowledge graph based on community consciousness elements includes: Transform consciousness elements into entity nodes of a knowledge graph and establish semantic relationships between entity nodes; Assign attributes to entity nodes and semantic relationships for cognitive evaluation; By integrating all entity nodes, semantic relationships, and attributes, a knowledge graph is obtained.

[0007] Preferably, the step of setting cognitive state nodes and cognitive state transition edges in the knowledge graph, and adding a timeline identical to the educational process to obtain a temporal knowledge graph, includes: Based on the cognitive development law of community consciousness, cognitive state nodes are added to the knowledge graph, wherein the cognitive state nodes include the degree of cognitive clarity; Based on the evolution logic of cognitive states, cognitive state transition edges are established between adjacent cognitive state nodes. These cognitive state transition edges are used to represent the transferable directions and progressive relationships of cognitive states. By adding time attributes to entity nodes, cognitive state nodes, and cognitive state transition edges within the knowledge graph, a temporal knowledge graph corresponding to the educational process and time nodes is obtained.

[0008] Preferably, a cognitive depth is set for the level of cognitive clarity, and a continuous and non-overlapping score threshold range is set for the cognitive depth; The level of cognitive clarity is linked to a score threshold range for cognitive depth; Store the binding relationships in a time-series knowledge graph.

[0009] Preferably, the step of comparing the time-series sequence with the time-series knowledge graph according to the time axis and calculating the cognitive depth score for the corresponding time period includes: Based on the timeline, the evaluation periods are extracted, and the temporal sequence of cognitive features is divided using the evaluation periods; The identified cognitive features are compared with the entity nodes in the temporal knowledge graph corresponding to the evaluation period. Based on the comparison results, the cognitive depth score for the current assessment period is calculated.

[0010] Preferably, the step of generating a cognitive state transition path by combining the cognitive state transition edges and the continuous changes of the cognitive state on the time axis includes: The cognitive states obtained from comparing different time periods are then linked together in chronological order along the timeline. Based on the cognitive state transition edge, for the series of cognitive states, the two adjacent cognitive states are checked to determine whether the change of cognitive state satisfies the preset transition direction and progressive relationship of the cognitive state transition edge. Those that pass the verification are integrated in chronological order to generate cognitive state transition paths.

[0011] Preferably, the steps of traversing the time sequence, combining the cognitive state transition path, locating the time point of cognitive state change, matching the time point with multi-source data, and obtaining the corresponding triggering event include: By traversing the temporal sequence of cognitive features and combining the cognitive state transition path, the time periods in which cognitive states change are selected to obtain the time points of cognitive state changes. Define matching time windows centered on time points, and extract multi-source data within the time windows; The extracted multi-source data is processed to identify the corresponding educational behaviors and obtain the triggering events that cause changes in cognitive state.

[0012] Preferably, the step of using a graph neural network to calculate the output integration and influence scores includes: Based on interactive network data, an interactive relationship graph is constructed with students as graph nodes and the frequency and duration of interactions between students as edges. The interaction graph is input into a graph neural network to calculate the integration and influence scores.

[0013] Preferably, the student's community integration score is calculated by using the number of node interaction connections, total interaction frequency, and total interaction duration extracted by the graph neural network.

[0014] Preferably, the community influence score of students is calculated by using the node weighted interaction degree, total frequency of interaction and total duration of interaction extracted by graph neural network.

[0015] The technical effects and advantages of the present invention in the above technical solution are as follows: This invention collects data from the educational process and constructs a temporal sequence of cognitive features based on a unified timeline. It then combines this with elements of community consciousness to construct a temporal knowledge graph containing cognitive states and cognitive state transition edges. This enables the calculation of cognitive depth scores for each time period, the determination of cognitive states, and the generation of dynamic cognitive state transition paths. It also identifies triggering events that cause changes in cognitive states. Furthermore, based on a student interaction network, it constructs an interaction relationship graph and utilizes a graph neural network to calculate integration and influence scores. Finally, it integrates cognitive depth scores, cognitive states, and cognitive state transition paths to complete the evaluation of educational effectiveness. This effectively solves the problems of traditional static evaluation, which fails to reflect the true trajectory of cognitive development, struggles to determine cognitive transition paths, and cannot connect with educational stages, leading to inaccurate evaluation results. It achieves a precise evaluation of the effectiveness of community consciousness education and enhances the application value of educational effectiveness evaluation. Attached Figure Description

[0016] Figure 1 This is a flowchart of an intelligent evaluation method for the effectiveness of community awareness education based on multi-source data perception, according to the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] like Figure 1 As shown, this embodiment provides an intelligent evaluation method for the effectiveness of community awareness education based on multi-source data perception, including the following steps: S1: Determine the multi-source data in the education process, and extract the time-stamped cognitive features from the multi-source data based on the timeline of the education process to obtain the temporal sequence of the cognitive features; S2: Based on the elements of community consciousness, construct a knowledge graph as a standard for cognitive assessment. Set cognitive state nodes and cognitive state transition edges in the knowledge graph, and add the same timeline as the education process to obtain a temporal knowledge graph. S3: Based on the timeline, compare the time sequence with the time sequence knowledge graph for each time period, calculate the cognitive depth score for the corresponding time period, and compare it with the cognitive state node in real time to obtain the cognitive state of the current time period. Combine the cognitive state transition edge and the continuous change of cognitive state on the timeline to generate the cognitive state transition path. S4: Traverse the time sequence, combine the cognitive state transition path, locate the time point of cognitive state change, match the time point with multi-source data, and obtain the corresponding triggering event; S5: Based on interactive network data of student collaboration, a graph neural network is used to calculate the integration and influence scores. Combined with cognitive depth scores, cognitive state transition paths and triggering events, the evaluation results are obtained.

[0019] As shown in steps S1 to S5 above, this invention extracts cognitive features with time attributes from multi-source data in the educational process and according to a unified timeline of the educational process, thereby constructing a temporal sequence of cognitive features. Based on the elements of community consciousness, a basic knowledge graph is constructed. On this basis, cognitive state nodes and cognitive state transition edges are added to the knowledge graph. After adding the timeline, a temporal knowledge graph with tracking capabilities is formed to assess students' cognitive levels. Using the temporal knowledge graph, the temporal sequence of cognitive features is compared time-by-time according to the timeline to calculate and match cognitive depth scores to obtain the current cognitive state. Combining the cognitive state transition edges with the continuous changes of the state on the timeline, a cognitive state transition path is generated to obtain its development trajectory. Then, using the temporal sequence and the cognitive state transition path, the time point when the cognitive state changes is located, thereby obtaining the educational event that caused the change in cognitive state. Combining the integration score and influence score calculated by the graph neural network, the evaluation result is obtained. This solves the technical problem that existing evaluation methods cannot reflect the true development trajectory and transition path of cognition, nor can they determine which educational link caused the cognitive change. For multi-source data, it mainly includes text data, voice data, behavioral interaction data and teaching process data generated during the education process. Text data includes classroom discussion texts, group communication records, etc.; voice data includes classroom speaking audio, group collaborative discussion audio, etc.; behavioral interaction data includes collaborative interaction logs, speaking frequency, mutual assistance behavior, etc.; teaching process data includes teaching activity data such as classroom collaborative discussion. After acquiring multi-source data, time-stamped cognitive features are extracted from the multi-source data. Specifically, based on a unified timeline of the education process, the timestamps of all multi-source data are normalized. Based on the assessment dimensions of community awareness education, features related to cognitive level and identification tendency are extracted from the processed data. For example, semantic cognitive features include the expression of understanding of the community; cognitive clarity features include the degree of cognitive ambiguity and clarity; identification tendency features include the strength of identification with collective values; and interactive participation features include the degree of collaborative responsiveness. The extracted features are bound to time points and arranged in chronological order along the timeline to obtain a temporal sequence of cognitive features.

[0020] In one embodiment of the present invention, the step of constructing a knowledge graph based on community consciousness elements includes: S21: Transform consciousness elements into entity nodes of a knowledge graph and establish semantic relationships between entity nodes; S22: Assign attributes to entity nodes and semantic relationships for cognitive evaluation; S23: Integrate all entity nodes, semantic relationships, and attributes to obtain a knowledge graph; The steps of setting cognitive state nodes and cognitive state transition edges in the knowledge graph, and adding a timeline identical to that of the educational process to obtain a temporal knowledge graph, include: S24: Based on the cognitive development law of community consciousness, cognitive state nodes are added to the knowledge graph, wherein the cognitive state nodes include the degree of cognitive clarity; S25: Based on the evolution logic of cognitive states, establish cognitive state transition edges between adjacent cognitive state nodes. The cognitive state transition edges are used to represent the transferable directions and progressive relationships of cognitive states. S26: For entity nodes, cognitive state nodes, and cognitive state transition edges in the knowledge graph, add time attributes using the time axis to obtain a temporal knowledge graph corresponding to the educational process and time nodes. To determine the level of cognitive clarity, a cognitive depth is set, and a continuous and non-overlapping score threshold range is defined for the cognitive depth. The level of cognitive clarity is linked to a score threshold range for cognitive depth; Store the binding relationships in a time-series knowledge graph.

[0021] In this embodiment of the invention, as shown in steps S21 to S23 above, consciousness elements are transformed into entity nodes of a knowledge graph. These consciousness elements refer to the consciousness content of the community, mainly including conceptual elements, subject object elements, educational behavior elements, and value orientation elements. These serve as nodes in the knowledge graph, aiming to structurally express the abstract community consciousness content and form the basic units of the knowledge graph for use as cognitive benchmarks for evaluation. Semantic relationships between entity nodes are used to express the association between two nodes. For example, in an inclusion relationship, community consciousness includes collective identity, a sense of belonging, and a sense of responsibility; in a promotion relationship, collaborative learning promotes collaborative spirit, and thematic education enhances collective identity. After establishing semantic relationships between nodes, attributes are assigned to them. Entity node attributes include concept definition, cognitive level, cognitive standards, and educational goals, while semantic relationship attributes include relationship strength, direction of influence, and effect. By assigning attributes, the knowledge graph can be used for quantitative calculations of cognitive evaluation. To accurately reflect the trajectory of cognitive state changes as the educational process progresses, as shown in steps S24 to S26 above, cognitive state nodes and cognitive state transition edges are set in the constructed knowledge graph to connect these nodes. Specifically, based on the natural cognitive development law of community consciousness from vague to clear and from shallow to deep, cognitive state nodes of varying clarity are set in the graph. Following the progressive development logic of cognitive states, cognitive state transition edges are set between adjacent cognitive state nodes. These transition edges limit changes in cognitive state to only occur along the direction from vague to relatively clear to clear, thus ensuring a positive progression of cognitive states. This makes the transition path of cognitive state changes in the graph reasonable and legal, conforming to educational laws and cognitive logic. In this process, the same time attribute is added to all entity nodes, cognitive state nodes, and cognitive state transition edges in the graph, based on the timeline used in the educational process, so that graph elements correspond to educational periods or stages. For the level of cognitive clarity, a cognitive depth is set, and continuous and non-overlapping score threshold intervals are divided for the cognitive depth. The level of cognitive clarity is uniquely bound to the corresponding score threshold interval, and the binding relationship is stored as an attribute in the temporal knowledge graph. This is to transform cognitive states such as vague, relatively clear, and clear into score indicators that can be quantified and calculated, and to obtain standardized cognitive state judgment rules.

[0022] In one embodiment of the present invention, the step of comparing the time-series sequence with the time-series knowledge graph according to the time axis and calculating the cognitive depth score for the corresponding time period includes: S31: Based on the timeline, extract the evaluation period and use the evaluation period to divide the temporal sequence of cognitive features; S32: Compare the identified cognitive features with the entity nodes in the temporal knowledge graph corresponding to the evaluation period; S33: Calculate the cognitive depth score for the current assessment period based on the comparison results; The step of generating a cognitive state transition path by combining the cognitive state transition edges and the continuous changes of the cognitive state on the time axis includes: S34: Connect the cognitive states obtained from comparing each time period in chronological order according to the timeline; S35: Based on the cognitive state transition edge, for the series of cognitive states, verify the two adjacent cognitive states to determine whether the cognitive state change satisfies the preset transition direction and progressive relationship of the cognitive state transition edge. S36: For those that pass the verification, integrate them in chronological order to generate cognitive state transition paths.

[0023] In this embodiment of the invention, as shown in steps S31 to S33 above, before matching cognitive features to a temporal knowledge graph, it is necessary to divide the temporal sequence of cognitive features. Based on the time axis, evaluation periods are extracted to obtain continuous and non-overlapping independent evaluation units. Specifically, evaluation periods are extracted according to a fixed duration or the duration of a teaching segment. The periods are continuous and non-overlapping and cover the entire education process. Each evaluation period is marked with a start and end timestamp to ensure complete alignment with the corresponding time slice of the temporal knowledge graph. After identifying cognitive characteristics by dividing the assessment period, these characteristics are compared with the corresponding entity nodes in the temporal knowledge graph. Specifically, based on the start and end timestamps of the assessment period, corresponding entity nodes are selected from the temporal knowledge graph. These include elements such as community consciousness, collective identity, and collaborative spirit in classroom discussions and thematic education sessions. Cognitive characteristics are then broken down into semantic expression characteristics, cognitive accuracy characteristics, element coverage characteristics, and identification tendency characteristics, ensuring a one-to-one correspondence between these characteristics and the attributes of the entity nodes in the graph. Specifically, for semantic fit comparison, the students' cognitive expressions need to be compared with the conceptual definition attributes of the entity nodes. The process involves several steps: First, comparing the cognitive features to determine if the expressions are standardized and consistent in meaning. Second, comparing the cognitive features to check if they cover the entity nodes in the graph for that specific time period, thus determining the completeness of cognition. Third, comparing the cognitive accuracy to compare the students' cognitive results with the standard cognitive attributes in the graph, thus determining if the cognition is correct. Fourth, comparing the cognitive features to verify if they conform to the semantic relationships between entity nodes in the graph, such as inclusion or subordination, to determine the rationality of the cognitive logic. Finally, combining the results, the matching fit, coverage completeness, and cognitive accuracy are calculated to obtain a cognitive depth score. Based on the cognitive depth score, it is compared with the cognitive state node to obtain the cognitive state of the current evaluation period. Specifically, based on the temporal knowledge graph, the binding relationship of the pre-stored cognitive depth score threshold interval is retrieved to obtain the score interval corresponding to each cognitive state node such as fuzzy cognition, relatively clear cognition, and clear cognition. The cognitive depth score of the current evaluation period is matched with each score threshold interval to determine the corresponding cognitive state, and the cognitive state of the period is associated with the timestamp and recorded. After obtaining the cognitive state for each evaluation period, a cognitive state transition path is generated using the cognitive state and its continuous changes on the time axis, as shown in steps S34 to S36 above. Using the cognitive state and the timestamps of the associated records, the cognitive states are sequentially linked according to the time axis. Then, based on the cognitive state transition edge, the legality of the cognitive states in two adjacent time periods in the linked cognitive states is checked to determine whether the state change conforms to the preset cognitive state transition direction and hierarchical progression relationship from vague cognition to clearer cognition to clear cognition. The cognitive state transition path is obtained using the cognitive states that pass the check. It is important to understand that the preset cognitive state transition direction does not only include a unidirectional positive progression from vague cognition to clearer cognition and then to clear cognition. Instead, it combines the actual laws of dynamic improvement and decline in cognition to verify the transition path. The transition path includes: a positive progression from vague cognition to clearer cognition and then to clear cognition; a smooth transition of the same cognitive state, i.e., from vague cognition to vague cognition; and a reasonable decline from clear cognition to clearer cognition and then from clearer cognition to vague cognition. Therefore, during cognitive state verification, only illogical abnormal jumps are excluded, such as a direct jump from vague cognition to clear cognition. The above three types of transitions that conform to the logic of cognitive evolution are all legal, to ensure that the generated cognitive state transition path truly reflects the trajectory of cognitive change.

[0024] In one embodiment of the present invention, the step of traversing the time sequence, combining the cognitive state transition path, locating the time point of cognitive state change, matching the time point with multi-source data, and obtaining the corresponding triggering event includes: S41: By traversing the temporal sequence of cognitive features and combining the cognitive state transition path, the time periods in which cognitive states change are selected to obtain the time points of cognitive state changes. S42: Define a matching time window centered on a time point and extract multi-source data within the time window; S43: Process the extracted multi-source data, identify the corresponding educational behaviors, and obtain the triggering events that cause changes in cognitive state.

[0025] In this embodiment of the invention, as shown in steps S41 to S43 above, the cognitive feature time sequence is traversed, and the cognitive state transition path is compared with the cognitive state of adjacent time periods before and after each time period. The time periods in which the cognitive state changes, shows positive progression, stable maintenance, or reasonable decline are selected, and the time point of cognitive state change is located. With this time point as the center, a matching time window of preset duration is defined forward and backward. The time window can completely cover the educational process before and after the state change. All text, voice, behavioral interaction, and teaching link data within the time window are extracted from multi-source data and processed to obtain teaching activities and interactive behaviors related to the cognitive state change, which are used as trigger events for the cognitive state change.

[0026] In one embodiment of the present invention, the step of using a graph neural network to calculate the output integration degree and influence score includes: S51: Based on interactive network data, construct an interactive relationship graph with students as graph nodes and the frequency and duration of interaction between students as edges; S52: Input the interaction graph into a graph neural network to calculate the integration and influence scores; The student's community integration score is calculated by using the number of node interaction connections, total interaction frequency, and total interaction duration extracted by the graph neural network. The community influence score of students is calculated by using the node weighted interaction degree, total frequency of interaction and total duration of interaction extracted by graph neural network.

[0027] In this embodiment of the invention, as shown in steps S51 to S52 above, interactive network data refers to the behavioral interaction data generated by students during community collaboration. It includes student interaction object identifiers, single interaction duration, cumulative interaction frequency, interaction initiator, interaction response results, group collaboration participation records, classroom speaking interaction records, and mutual assistance and communication behavior data, etc. Using interactive network data, an interactive relationship graph is constructed. Specifically, each student participating in collaborative learning is used as a graph node. If there is collaborative interaction behavior between two students, a directed or undirected edge is established. The weight of the edge is obtained by weighting the interaction frequency and interaction duration. The higher the weight value, the closer the interaction between the two students. All nodes and edges are structurally integrated to generate an interactive relationship graph. In a static collaborative scenario with fixed interactions in classroom groups, a graph convolutional neural network is used. The interaction relationship graph is input into the model, and the model aggregates node features through an attention mechanism to extract node interaction features. Based on the number of node interaction connections, total interaction frequency, and total interaction duration output by the model, the student's community integration score is calculated. The higher the score, the more extensive the student's participation in the interaction. The student's community influence score is calculated through the node weighted interaction degree, total frequency of interaction, and total duration of interaction output by the model. The higher the score, the higher the degree of interaction the student receives. Based on the integration and influence scores calculated above, combined with the cognitive depth score, cognitive state transition path and triggering events, the evaluation is conducted from three perspectives: cognitive level, developmental trajectory and behavioral participation. The evaluation results are generated based on the weighted fusion and comprehensive judgment results of the above data.

[0028] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A method for intelligently evaluating the effectiveness of community awareness education based on multi-source data perception, characterized in that, Includes the following steps: Identify multi-source data in the educational process, and extract time-stamped cognitive features from the multi-source data based on the timeline of the educational process to obtain a temporal sequence of cognitive features; Based on the elements of community consciousness, a knowledge graph is constructed as a standard for cognitive assessment. Cognitive state nodes and cognitive state transition edges are set in the knowledge graph, and the same timeline as the education process is added to obtain a temporal knowledge graph. Based on the timeline, the time sequence is compared with the time sequence knowledge graph for each time period to calculate the cognitive depth score for the corresponding time period. The score is then compared with the cognitive state node in real time to obtain the cognitive state for the current time period. Combined with the cognitive state transition edge and the continuous change of the cognitive state on the timeline, a cognitive state transition path is generated. Traverse the time sequence, combine the cognitive state transition path, locate the time point of cognitive state change, match the time point with multi-source data, and obtain the corresponding triggering event. Based on interactive network data of student collaboration, a graph neural network is used to calculate the integration and influence scores. Combined with the cognitive depth score, cognitive state transition path and triggering event, the evaluation results are obtained.

2. The intelligent evaluation method for the effectiveness of community consciousness education based on multi-source data perception as described in claim 1, characterized in that, The steps for constructing a knowledge graph based on community consciousness elements include: Transform consciousness elements into entity nodes of a knowledge graph and establish semantic relationships between entity nodes; Assign attributes to entity nodes and semantic relationships for cognitive evaluation; By integrating all entity nodes, semantic relationships, and attributes, a knowledge graph is obtained.

3. The intelligent evaluation method for the effectiveness of community awareness education based on multi-source data perception as described in claim 1, characterized in that, The steps of setting cognitive state nodes and cognitive state transition edges in the knowledge graph, and adding a timeline identical to that of the educational process to obtain a temporal knowledge graph, include: Based on the cognitive development law of community consciousness, cognitive state nodes are added to the knowledge graph, wherein the cognitive state nodes include the degree of cognitive clarity; Based on the evolution logic of cognitive states, cognitive state transition edges are established between adjacent cognitive state nodes. These cognitive state transition edges are used to represent the transferable directions and progressive relationships of cognitive states. By adding time attributes to entity nodes, cognitive state nodes, and cognitive state transition edges within the knowledge graph, a temporal knowledge graph corresponding to the educational process and time nodes is obtained.

4. The intelligent evaluation method for the effectiveness of community awareness education based on multi-source data perception as described in claim 3, characterized in that: To determine the level of cognitive clarity, a cognitive depth is set, and a continuous and non-overlapping score threshold range is defined for the cognitive depth. The level of cognitive clarity is linked to a score threshold range for cognitive depth; Store the binding relationships in a time-series knowledge graph.

5. The intelligent evaluation method for the effectiveness of community consciousness education based on multi-source data perception as described in claim 1, characterized in that, The step of comparing the time-series sequence with the time-series knowledge graph according to the time axis and calculating the cognitive depth score for the corresponding time period includes: Based on the timeline, the evaluation periods are extracted, and the temporal sequence of cognitive features is divided using the evaluation periods; The identified cognitive features are compared with the entity nodes in the temporal knowledge graph corresponding to the evaluation period. Based on the comparison results, the cognitive depth score for the current assessment period is calculated.

6. The intelligent evaluation method for the effectiveness of community awareness education based on multi-source data perception as described in claim 1, characterized in that, The step of generating a cognitive state transition path by combining the cognitive state transition edges and the continuous changes of the cognitive state on the time axis includes: The cognitive states obtained from comparing different time periods are then linked together in chronological order along the timeline. Based on the cognitive state transition edge, for the series of cognitive states, the two adjacent cognitive states are checked to determine whether the change of cognitive state satisfies the preset transition direction and progressive relationship of the cognitive state transition edge. Those that pass the verification are integrated in chronological order to generate cognitive state transition paths.

7. The intelligent evaluation method for the effectiveness of community consciousness education based on multi-source data perception as described in claim 1, characterized in that, The steps of traversing the time sequence, combining the cognitive state transition path, locating the time point of cognitive state change, matching the time point with multi-source data, and obtaining the corresponding triggering event include: By traversing the temporal sequence of cognitive features and combining the cognitive state transition path, the time periods in which cognitive states change are selected to obtain the time points of cognitive state changes. Define matching time windows centered on time points, and extract multi-source data within the time windows; The extracted multi-source data is processed to identify the corresponding educational behaviors and obtain the triggering events that cause changes in cognitive state.

8. The intelligent evaluation method for the effectiveness of community consciousness education based on multi-source data perception as described in claim 1, characterized in that, The step of using a graph neural network to calculate the output integration and influence scores includes: Based on interactive network data, an interactive relationship graph is constructed with students as graph nodes and the frequency and duration of interactions between students as edges. The interaction graph is input into a graph neural network to calculate the integration and influence scores.

9. The intelligent evaluation method for the effectiveness of community awareness education based on multi-source data perception as described in claim 8, characterized in that: The community integration score of students is calculated by using the number of node interaction connections, total interaction frequency, and total interaction duration extracted by graph neural network.

10. The intelligent evaluation method for the effectiveness of community awareness education based on multi-source data perception as described in claim 8, characterized in that: The community influence score of students is calculated by using the node weighted interaction degree, total frequency of interaction and total duration of interaction extracted by graph neural network.

Images