Method for constructing a risk map based on animal metaphor

Abstract

This invention discloses a risk graph construction method based on animal metaphors, belonging to the field of data intelligent analysis and risk management technology. The method includes the following steps: compressing multi-source semantic data onto a unified timeline in chronological order; organizing the risk information in the semantic data; establishing a correspondence between risk elements and animal behavioral characteristics using animal metaphors; and generating a node activation rhythm index. This invention, by introducing a node activation rhythm index, semantic overlap fingerprints, and an influence intensity quota grading mechanism, achieves hierarchical expression and rhythmic control of risk semantics in both temporal and semantic dimensions. This enables accurate identification of high-density trigger areas and prevents mis-associations of nodes, reducing risk diffusion caused by semantic resonance, allowing for segmented release of risk propagation paths, and improving the stability and interpretability of the graph.

Description

Technical Field

[0001] This invention relates to the field of data intelligence analysis and risk management technology, specifically to a method for constructing risk maps based on animal metaphors. Background Technology

[0002] Risk mapping based on animal metaphors utilizes the typical behavioral characteristics, ecological relationships, and survival strategies exhibited by animals in nature as analogical vehicles for understanding and expressing risk characteristics. It semantically maps and structurally organizes abstract and complex risk elements using concrete animal metaphors, thereby forming a visualized and reasonable risk knowledge graph. This method extracts behavioral patterns such as predation, escape, herding, camouflage, and migration from animals and establishes correspondences with risk sources, transmission paths, receptor vulnerabilities, and defense mechanisms in real-world systems. Then, it uses semantic networks and knowledge graph technologies to structurally express these metaphorical relationships, visually representing the chains of action, evolutionary trends, and potential impacts between different types of risks. Ultimately, risk managers can use the cognitive framework of animal metaphors to more intuitively understand the dynamic balance and conflict patterns of the risk ecosystem, enabling the logical deduction of risk identification, classification, and transmission paths.

[0003] Existing technologies suffer from the following shortcomings: In existing technologies, risk maps based on animal metaphors typically simulate the dynamic evolutionary relationships between different risk factors by real-time parsing and node updates of multi-source semantic data. However, when the system operates under high-frequency dynamic input environments, the semantic frequencies of multiple animal metaphor nodes may overlap on the time axis, forming a collective resonance region within the semantic space. At this point, originally independent metaphor nodes, due to the convergence of semantic amplitudes, are misidentified by the system as associated risk entities with synchronously changing characteristics, leading to abnormal coupling of risk propagation paths. Furthermore, this semantic resonance triggers chain responses at the risk inference level, causing the risk model to erroneously amplify the influence intensity between some metaphor nodes, thereby triggering nonlinear cross-diffusion of multi-level risk chains. When the resonance region continues to expand and the feedback mechanism fails to suppress it in time, the overall structure of the risk propagation model becomes unstable, causing the risk diffusion range to spiral out of control. This makes it difficult to achieve dynamic constraints through conventional semantic separation strategies, severely impacting the interpretability and reliability of the risk map.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a risk mapping method based on animal metaphors to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a risk map construction method based on animal metaphors, comprising the following steps: Multi-source semantic data is compressed onto a unified time axis in chronological order. Risk information in the semantic data is organized and a correspondence between risk factors and animal behavioral characteristics is established using animal metaphors. A node activation rhythm index is generated to characterize the triggering rhythm characteristics of each animal metaphor node in the time dimension. Based on the node activation rhythm index, the triggering frequency, duration and variation of each animal metaphor node in a continuous time window are statistically calculated to identify time segments with a sudden increase in triggering density, and semantic overlap fingerprints are generated based on the identification results to characterize the triggering features of potential semantic resonance regions. The semantic boundaries of the trigger density anomaly segment are split using semantic overlap fingerprints, the trigger positions are distinguished and marked, and semantic correspondence is established based on the context relationship to generate a list of boundary anchor points to limit the semantic scope of each animal metaphor node. Based on the list of boundary anchor points, the node update time sequence in the density anomaly zone is adjusted in a gated manner. The centralized triggering behavior is dispersed by shifting the node update time forward and backward in rotation, and a staggered update table is generated to constrain the trigger synchronization of nodes in the time dimension. Based on the staggered update table, the triggering order between nodes is controlled in a hierarchical manner. An influence intensity quota sequence is introduced, and adjacent animal metaphor nodes are divided into different quota levels. The triggering conditions are reset according to the quota level. The triggering order in the abnormal segment is periodically reversed and controlled. A short cooling time is set during the triggering process so that the risk propagation path is released in segments, thereby avoiding the risk spread out of control caused by trigger resonance.

[0007] Preferably, the steps for generating the node activation rhythm index are as follows: Multi-source semantic data is compressed into a unified timeline in chronological order, and the time units of the data are unified and aligned to form a continuous semantic stream; Extract risk sources, propagation paths, carriers, defensive behaviors, feedback results, intervention measures and post-effect states from semantic data on the timeline to generate a set of risk semantic units; Establish a correspondence between the risk semantic unit set and the animal metaphor category, and generate a metaphor correspondence table containing time, semantic risk type, metaphor category, behavioral tendency and contextual information; Based on the metaphor correspondence table, the trigger interval, occurrence frequency and duration of animal metaphor nodes are statistically analyzed to distinguish periodic, intermittent and random trigger types, and the trigger frequency, time interval and semantic association weight are recorded to generate a node activation rhythm index.

[0008] Preferably, the semantically overlapping fingerprint generation steps are as follows: Based on the timestamp information in the node activation rhythm index, the triggering events of the animal metaphor nodes are arranged in chronological order. A combination of fixed length and sliding overlap is used to divide the continuous time window, and an overlap area is introduced at the window boundary to maintain time continuity. The number of triggers, intervals, and durations of animal metaphor nodes within each time window are statistically analyzed. The trigger start point, duration, and end point are recorded, and a weighted average is calculated to reflect the level of trigger activity. By comparing the trigger frequency and duration results of adjacent time windows, nodes with increased trigger frequency, extended duration, or shortened interval in continuous windows are identified to determine the trigger density surge segment. Extract the time range and participating nodes of the trigger density surge segment, record the trigger time, duration, semantic direction and metaphor category, and generate semantic overlap fingerprint based on semantic direction consistency to characterize the features of the semantic resonance region.

[0009] Preferably, in the process of generating semantic overlap fingerprints, the animal metaphor nodes in the trigger density surge segment are grouped according to semantic direction consistency and behavioral convergence to form semantic overlap groups. A hierarchical time series structure is established based on the trigger time, duration and semantic fusion degree of the nodes in the group, so that the semantic overlap fingerprint accurately reflects the temporal relationship and semantic interaction features between nodes in the semantic resonance segment.

[0010] Preferably, the steps for generating the boundary anchor point list are as follows: Based on the semantic overlap fingerprint parsing time segment identifier, node category, semantic fusion degree, trigger frequency distribution and duration characteristics, semantic clustering is performed on the nodes within the time segment to determine the semantic core nodes and auxiliary semantic intervals; Analyze the semantic behavior type, semantic object, semantic transmission direction and semantic emotional color of each animal metaphor node within the semantic interval, identify semantic overlap area, semantic conflict area and semantic transition area and record the hierarchical relationship; On the timeline, semantic overlay areas, semantic conflict areas, and semantic transition areas are located, a time boundary marker sequence is generated, and semantic continuity, semantic transformation, and semantic differentiation correspondences are established based on the behavior type, semantic direction, and target of the preceding and following semantic nodes. The time boundary markers are anchored to generate a list of boundary anchors containing information such as time coordinates, node type, semantic behavior attributes, semantic direction, and corresponding context, which is used to limit the semantic scope of the animal metaphor nodes.

[0011] Preferably, when generating the boundary anchor list, the anchor information is recorded sequentially according to the time sequence, the anchor corresponding to the semantic core node is arranged first, and the anchor hierarchy order is determined according to the semantic priority. The semantic interval is divided by the anchor time coordinate, and the semantic function of the animal metaphor node is limited to the adjacent anchors to form a continuous or segmented semantic function boundary.

[0012] Preferably, the steps for generating the staggered update table are as follows: Extract time anchor information within density anomaly segments from the boundary anchor list, divide continuous time segments based on time coordinates, and divide animal metaphor nodes into independent node groups and dependent node groups according to semantic dependency relationships to form a semantic dependency matrix. Time windows are divided according to time distribution characteristics, time overlap areas are set to ensure time continuity, node trigger concentration is calculated and gating scheduling conditions are established, trigger interval constraints are set for high-density windows and trigger filling constraints are set for low-density windows. Based on the gating scheduling conditions, the animal metaphor nodes are adjusted by alternating forward and backward movements, prioritizing the adjustment of the update time order of semantically independent nodes and dependent nodes to avoid concentrated triggering of behaviors; The results of the update time adjustment are recorded in a structured manner, generating a staggered update table based on time, and a trigger order constraint mechanism is established to ensure the time consistency of the node triggering logic.

[0013] Preferably, the trigger sequence hierarchical control steps are as follows: Extract the update time, semantic category, semantic direction, trigger duration, semantic intensity and semantic relationship information of animal metaphor nodes from the staggered update table, divide the continuous time layer according to time and establish a semantic dependency chain to construct the trigger hierarchy structure; An influence intensity quota sequence is established based on temporal layer and semantic dependency relationship. Animal metaphor nodes are divided into high quota level, medium quota level and low quota level according to the triggering frequency, duration and propagation ability of the nodes. Triggering conditions are set according to quota levels. Time interval constraints and semantic exclusion conditions are set for high quota nodes, follow conditions and connection conditions are set for medium quota nodes, and compensation conditions and delay conditions are set for low quota nodes. Implement periodic reverse rotation control to adjust the triggering order of nodes of different quota levels periodically to maintain semantic balance; Set a short cooldown period between node triggers, and determine the cooldown duration according to the quota level, so as to realize the segmented release of the risk propagation path and maintain the stability of propagation.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention introduces node activation rhythm indexes and semantic overlap fingerprints during the risk graph construction process, enabling synchronous compression and rhythmic expression of multi-source semantic data in the temporal dimension. This allows for the accurate identification of high-density segments triggered by semantics, and the semantic boundaries are split and defined before and after triggering using a list of boundary anchor points, effectively avoiding mis-associations of nodes caused by semantic overlap. This approach provides a precise hierarchical expression of risk semantics in both temporal and semantic dimensions, making the semantic structure of the risk graph clearer and its logical relationships more traceable, fundamentally improving the accuracy of risk identification and semantic reasoning.

[0015] This invention employs a gated update time control and influence intensity quota grading mechanism to hierarchically manage the triggering order of animal metaphor nodes within abnormal segments. It also introduces periodic reverse rotation and short cooldown periods during the triggering process, achieving rhythmic release of risk semantic propagation. This approach effectively reduces the probability of synchronous triggering between nodes, suppresses the cumulative effect of semantic resonance, and allows the risk propagation path to unfold in a segmented manner. Even under high-frequency input conditions, the system maintains structural stability and controllable propagation, thereby improving the dynamic evolutionary stability and interpretability of the risk map. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0017] Figure 1 This is a flowchart of the risk mapping method based on animal metaphors of the present invention. Detailed Implementation

[0018] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0019] This invention provides, for example Figure 1 The risk mapping method based on animal metaphors shown includes the following steps: Multi-source semantic data is compressed onto a unified time axis in chronological order. Risk information in the semantic data is organized and a correspondence between risk factors and animal behavioral characteristics is established using animal metaphors. A node activation rhythm index is generated to characterize the triggering rhythm characteristics of each animal metaphor node in the time dimension. To enable multi-source semantic data to form a continuous semantic stream along a unified time dimension, and to establish a correspondence between risk factors and animal behavioral characteristics using animal metaphors, thereby generating a node activation rhythm index that can characterize time-triggered rhythm features, the specific implementation steps are as follows: During implementation, semantic data from different collection channels needs to be integrated according to time signatures. All data must have a clear timestamp or a deducible time sequence identifier. Data sources can include public opinion information streams, financial transaction records, natural disaster reports, online public opinion texts, policy document content, geographic location change information, and environmental monitoring logs. Each type of data uses a different time scale during collection; for example, some data is recorded at the second level, while others are updated at the hour or day level. To ensure consistency of the timeline, the time unit of all data is first unified, using the smallest time resolution as the time step, mapping all types of data content to a continuous interval under a unified time scale. In this process, for datasets with long time spans, linear interpolation is used to fill time gaps to maintain the continuity of the timeline; for data with excessively fine time granularity, integration is performed according to a set time window to ensure that only one set of semantic feature data exists within the same time interval. After time alignment, the data content is semantically segmented, and semantic fragments in the text information are reorganized according to four dimensions: event subject, behavior, result attribute, and context. This ensures that each semantic record corresponds to a complete risk situation unit. After this process, the asynchronous semantic data, which was originally distributed across different time periods, is uniformly compressed onto the same timeline. All risk-related semantic content is arranged continuously in the time dimension, thus establishing a temporal reference basis for subsequent semantic behavior mapping.

[0020] In this stage, semantic parsing is performed on the semantic segments corresponding to each time scale, extracting risk-related semantic components from the text or event descriptions. The extracted risk elements mainly include seven dimensions: risk source, risk propagation path, risk bearer, risk defense behavior, risk feedback results, risk intervention measures, and risk aftereffects. This structured approach reorganizes the mixed descriptive information in the original semantic content into identifiable risk semantic units. When performing risk semantic identification, it is necessary to maintain the semantic contextual coherence according to the time sequence. If multiple risk descriptions appear within a certain time period, they are arranged in order of appearance, and the semantic connection relationship is recorded. For example, when both market fluctuations and policy adjustments exist in a time segment, a causal sequence relationship should be established between them at the semantic level, recording policy adjustments as a precursor to market fluctuations. Subsequently, the organized semantic risk units are clustered and summarized, categorized into six types based on the differences in semantic content characteristics: aggressive risk, evasive risk, collaborative risk, defensive risk, latent risk, and migrating dynamic risk. Each type corresponds to different behavioral tendencies. Through this stage of processing, the semantic risk content within each time period was clearly classified and semantic relationships were established, forming a set of risk elements within the semantic time window, providing a clear semantic structure for the next step of animal metaphor mapping.

[0021] At this stage, each set of risk elements sorted semantically is corresponded to the corresponding animal metaphor category. In specific implementation, predator animal metaphors are used to represent the active intrusion behaviors of risks in the external environment, such as financial attacks, public opinion guidance or information penetration; escape animal metaphors are used to represent risk avoidance, retreat or information dissipation behaviors, such as the outflow of market funds or the transfer of public attention; clustering animal metaphors are used to represent the concentrated responses of risks in similar events, such as mass reactions, linkages of similar events or emotional aggregation; camouflage animal metaphors are used to represent the concealment and misleading behaviors of risks at the semantic level, such as the spread of false information or the manifestation of risks disguised as safety signals; migration animal metaphors are used to represent the diffusion and transfer of risks in time or space, such as the diffusion from local risks to other fields. When performing metaphor mapping, according to the characteristics of risk semantic units in each time period, a one-to-one or many-to-one relationship is established between them and the animal metaphor category to form a metaphor correspondence table. The metaphor correspondence table records five dimensions: time, semantic risk type, metaphor category, behavior tendency and context information, and is used to describe the behavioral characteristics and semantic background of each metaphor node in time. In the case of strong time continuity, for example, the same type of risk repeatedly appears in consecutive time periods, it is mapped as a continuous triggering event of the same animal metaphor node. After the processing of this stage, all risk semantic units on the time axis obtain clear metaphorical representations, and the behavioral attributes and semantic sources of each metaphor node are detailedly recorded, laying a foundation for constructing a rhythmical time triggering relationship.

[0022] After completing the animal metaphor mapping, the triggering characteristics of all animal metaphor nodes on the time axis need to be statistically analyzed. First, the temporal distribution of each animal metaphor node is scanned, calculating the time interval, frequency, and duration of its appearance across the entire time axis, forming a temporal trigger sequence for the node. Next, based on the uniformity and continuity of the time intervals in the trigger sequence, the temporal behavior of the metaphor nodes is categorized into three forms: periodic triggering, intermittent triggering, and random triggering. Periodically triggered nodes typically represent systemic risk sources, intermittently triggered nodes represent local disturbance risks, and randomly triggered nodes represent sudden event risks. Subsequently, the triggering intervals and durations of each metaphor node in the time dimension are normalized to ensure comparability of different categories of metaphor nodes on a temporal scale. For metaphor nodes that overlap across multiple time windows, slight shifts are made by moving the time forward or backward to maintain the continuity and non-conflictual nature of the rhythm index. Finally, the triggering frequency, time interval, duration, and semantic association weight of the nodes are comprehensively recorded to form a node activation rhythm index. This index, with time as its main thread, comprehensively records the triggering rhythm information of each animal metaphor node in the temporal dimension, used to describe the dynamic patterns of risk interaction in semantic space and temporal dimension. The establishment of this index clearly expresses the rhythmic characteristics of risk elements in the time domain, providing a continuous and traceable framework for subsequent semantic overlap detection, temporal peak-shifting updates, and propagation path regulation.

[0023] Based on the node activation rhythm index, the triggering frequency, duration and variation of each animal metaphor node in a continuous time window are statistically calculated to identify time segments with a sudden increase in triggering density, and semantic overlap fingerprints are generated based on the identification results to characterize the triggering features of potential semantic resonance regions. Based on the node activation rhythm index, a comprehensive statistical analysis of the triggering features of animal metaphor nodes within a continuous time window is performed, and time intervals with sudden increases in trigger density are identified to generate semantic overlap fingerprints for characterizing semantic resonance states. The specific steps are as follows: During implementation, the trigger events of all animal metaphor nodes are first arranged chronologically based on the timestamp information recorded in the node activation rhythm index, ensuring the continuity and completeness of the time distribution. The division of time windows requires consideration of the trigger interval distribution of different nodes in the rhythm index, employing a combination of fixed length and sliding overlap. Fixed-length time windows ensure consistent analysis granularity, while sliding overlap ensures smooth transitions in time changes. The length of each time window can be set based on the average value of previous node trigger cycles; for example, shorter time intervals are chosen in scenarios with rapid risk evolution, while longer time intervals are used in scenarios with gradual risk changes, to maintain a match between time granularity and the rate of change in risk behavior. Subsequently, the trigger events of each animal metaphor node are assigned to the corresponding time windows, and overlapping areas are introduced at the window boundaries, allowing adjacent windows to share a portion of the time segment to capture the continuity of node trigger behavior within the boundary time. After the time window division is completed, the data within each window undergoes time alignment processing. For trigger records of multiple animal metaphor nodes within the same time period, they need to be arranged in order of timestamp from earliest to latest, and each trigger needs to be labeled with its precise position on the timeline. In this way, a complete temporal structure containing the node triggering order, duration, and triggering interval can be constructed in a continuous time dimension, providing a rigorous temporal basis for subsequent frequency analysis.

[0024] After completing the time window division and time alignment, detailed trigger characteristic statistics are needed for the animal metaphor nodes contained in each time window. First, the number of triggers for each node within the time window is counted, and the interval between two adjacent triggers is calculated. The statistical results of the trigger count and time interval together reflect the node's activity frequency. Then, the trigger duration for each node within the time window is calculated. Duration refers to the complete time span experienced by the node from the start to the end of the trigger. When a node triggers continuously within consecutive time windows, its time intervals spanning across windows need to be concatenated to ensure statistical continuity. To obtain accurate time distribution information, the trigger records for each node are refined, dividing a single trigger behavior into three stages: the trigger start point, the duration, and the end point. Each stage is marked with a timestamp and duration length. This method not only clearly identifies the number of triggers and time characteristics of nodes but also reflects the distribution of node activity intensity throughout the entire time window. During the statistical process, for multiple animal metaphor nodes of different categories within the same time window, their triggering characteristics need to be recorded separately, and the weighted average of the total number of triggers for all nodes within that time window needs to be calculated to reflect the overall triggering activity level during that period. This step allows for a clear understanding of the triggering frequency distribution and duration variation patterns of animal metaphor nodes within each time window, laying the foundation for calculating the magnitude of changes and identifying density spikes in the next stage.

[0025] Based on the statistical results of trigger frequency and duration, the magnitude of trigger changes is calculated, and segments with sudden increases in trigger density are identified. In this stage, the trigger frequency and duration results of adjacent time windows are compared one by one to analyze the changing trend of node activity behavior in the time series. For the same animal metaphor node, the difference in the number of triggers and the difference in duration between two consecutive time windows are first calculated to obtain the direction and intensity of the node's trigger change. When a node shows a trend of continuously increasing trigger count, significantly prolonged duration, or continuously shortening trigger interval in consecutive time windows, it can be determined that the node has entered a high-activity state. After identifying high-activity nodes, the persistence of their changes in the time series is observed. If the trigger density of a node remains in an increasing state in multiple consecutive windows, the node is defined as a density-surge node. Subsequently, the density-surge time segments of multiple nodes are cross-compared. When two or more animal metaphor nodes simultaneously exhibit density-surge characteristics in the same time window or adjacent windows, this time period can be marked as a trigger density-surge segment. In the process of identifying trigger density-surge segments, it is also necessary to combine the metaphor type characteristics of the node for comprehensive judgment. For example, when both predatory metaphor nodes and clustering metaphor nodes exhibit high-frequency triggering within the same time period, and the semantic expression directions of the two types of nodes are consistent, it indicates the existence of a high-intensity semantic concentration phenomenon within that time period. This comprehensive identification method can distinguish between individual high-frequency triggering of a single node and collaborative high-frequency triggering among multiple nodes, thereby accurately identifying potential semantic resonance segments.

[0026] Semantic overlap fingerprints are generated based on identified trigger density surge segments. In this stage, the time range of all trigger density surge segments is first extracted, and a set of animal metaphor nodes that trigger frequently within these time ranges is collected. Then, the node trigger events are arranged chronologically to form a temporal overlay sequence. Within the same time period, the trigger start time, duration, semantic direction, and metaphor category of each node are recorded. To reflect the semantic overlay relationship between different nodes, the similarity and behavioral tendencies between node semantic attributes need to be compared. When multiple nodes have consistent semantic directions or mutually reinforcing features within the same time period, they are marked as semantic overlap groups. Next, the temporal distribution and semantic relationships of each semantic overlap group are integrated, arranging trigger time, node type, duration, and semantic relationship in a hierarchical structure. Semantic overlap fingerprints are generated by encoding the overlay relationships of all groups. A semantic overlap fingerprint is a structured semantic description based on temporal order and node behavior, used to reflect specific segments where semantic resonance occurs in the temporal dimension. Each semantic overlap fingerprint contains five elements: time segment identifier, participating node type, node trigger frequency distribution, duration statistics, and semantic fusion degree. This fingerprint can intuitively present the formation process and semantic interaction characteristics of the semantic resonance region. Through this step, the temporal characteristics, node characteristics, and semantic relationships of the potential semantic resonance region are fully and structurally expressed, providing a data foundation and logical basis for subsequent semantic boundary decomposition, trigger order adjustment, and risk propagation suppression.

[0027] The semantic boundaries of the trigger density anomaly segment are split using semantic overlap fingerprints, the trigger positions are distinguished and marked, and semantic correspondence is established based on the context relationship to generate a list of boundary anchor points to limit the semantic scope of each animal metaphor node. The semantic boundaries of trigger density anomaly segments are precisely decomposed using semantic overlap fingerprints, and the trigger positions are meticulously distinguished and marked in the semantic space. At the same time, a clear semantic correspondence is established based on the context, thereby generating a list of boundary anchor points to limit the semantic scope of each animal metaphor node. The specific steps are as follows: Based on the generated semantic overlap fingerprint, a detailed analysis is performed on its included time segment identifiers, node categories, semantic fusion degree, trigger frequency distribution, and duration characteristics. The semantic overlap fingerprint, as a comprehensive record of multiple nodes triggering and overlapping over time, contains semantic resonance relationships between different animal metaphor nodes within the same time segment. In implementation, time segment information needs to be extracted from the semantic overlap fingerprint one by one, and a corresponding semantic record table needs to be established for each time segment. This record table contains five types of information: segment start time, segment end time, participating node type, semantic direction identifier, and semantic fusion weight. Next, semantic clustering is performed on the participating nodes within each time segment, grouping nodes with similar semantic directions and semantic strengths into a semantic cluster unit. This method identifies regions with a high degree of semantic overlap within the same time range. Subsequently, within each semantic cluster unit, the duration and participation degree of the dominant semantic node are calculated, and the node with the longest trigger duration and the highest semantic fusion weight is determined as the semantic core node. The semantic core node represents the main semantic behavior within that time segment, and its corresponding time range is defined as the core semantic interval. Other nodes are defined as auxiliary semantic intervals based on their semantic direction and differences from the core nodes. In this way, each trigger density anomaly segment is subdivided into several semantic intervals, each with a clear start time, end time, and semantic attributes, providing a temporal and semantic reference basis for subsequent boundary identification.

[0028] For a given set of semantic intervals, the internal semantic structure is analyzed one by one. Each semantic interval contains multiple animal metaphor nodes, which overlap or are adjacent in time, and may exhibit convergent, contradictory, or progressive characteristics at the semantic level. First, the semantic expression of each node in the interval is dissected, extracting its semantic behavior type, semantic object, semantic transmission direction, and semantic emotional tone. For example, predator-type metaphor nodes exhibit aggressive behavior in risk semantics, and their semantic object is usually the target individual or group; escape-type metaphor nodes exhibit avoidance behavior, and their semantic object is usually the risk source or constraint; clustering-type metaphor nodes exhibit aggregation behavior, and their semantic object is often a group of similar nodes; camouflage-type metaphor nodes exhibit hiding behavior, and their semantic object is mostly the external observer; migration-type metaphor nodes exhibit displacement behavior, and their semantic object is usually environmental elements or spatial areas. After identifying these semantic behavior characteristics, the semantic relationships between the nodes within each interval are compared one by one. When two nodes overlap in time and have the same semantic object and consistent semantic direction, they are considered to have a semantic superposition relationship. When they overlap in time but have opposite semantic directions, they are considered to have a semantic conflict relationship. When they are adjacent in time but have interconnected semantic directions, they are considered to have a semantic transition relationship. To ensure the accuracy of semantic division, these semantic relationships need to be recorded hierarchically, clearly distinguishing between semantic superposition areas, semantic conflict areas, and semantic transition areas. The semantic superposition area represents the region where the semantics of multiple nodes merge in time; the semantic conflict area represents the region where the semantics of different nodes contradict each other; and the semantic transition area represents the continuous region where semantics gradually shift from one state to another. Through this stage of analysis, the semantic structure within the semantic interval set is fully dissected, and the ambiguous areas of semantic boundaries are clearly revealed, providing a basis for trigger position marking.

[0029] The identified semantic overlap, semantic conflict, and semantic transition regions are located on the timeline. For each semantic overlap region, the point with the highest degree of temporal overlap is selected as the boundary center point, and its temporal coordinates, corresponding node type, and semantic behavior features are recorded. For semantic conflict regions, the time node with the most drastic semantic direction shift is selected as the boundary center point; for semantic transition regions, the midpoint of semantic change is selected as the boundary center point. All boundary center points are arranged chronologically to form a temporal boundary marker sequence. After the temporal boundary markers are completed, semantic features are assigned to these marker points. Each marker point is assigned the association information between the preceding and following semantic nodes, including the semantic type, semantic direction, object of action, and time span of the preceding semantic node, as well as the corresponding information of the following semantic node. In this way, a complete semantic context chain is established. The correspondence between the preceding and following contexts is divided into three types: semantic continuity, semantic transformation, and semantic differentiation. Semantic continuity indicates that the preceding and following semantic nodes have consistent behavior and the same direction; semantic transformation indicates that the preceding and following semantic nodes have different behavior but the same semantic object; semantic differentiation indicates that the behavior and object of the preceding and following semantic nodes both change. In recording these relationships, it is necessary to ensure the accuracy of the temporal sequence and the continuity of the semantic logic, so that each triggering position is clearly identified and related to adjacent semantic behaviors, thereby constructing a semantic flow structure with temporal coherence and semantic logical consistency.

[0030] Based on the established semantic boundary marker sequence, each marker point is anchored. The core purpose of anchoring is to transform semantic boundary points into traceable semantic location identifiers. Each anchor point contains six elements: time coordinates, associated node type, semantic behavior attribute, semantic direction, semantic object, and contextual information. Anchor points are generated sequentially according to time, ensuring a continuous and unique arrangement on the timeline. For multiple anchor points within the same time period, their hierarchical order is determined based on semantic priority, typically placing the anchor point corresponding to the semantic core node first, with the remaining anchor points arranged sequentially. After generating the boundary anchor point list, the list is used to limit the semantic scope of animal metaphor nodes. Specifically, semantic intervals are divided based on the anchor point's time coordinates, restricting the semantic function of each animal metaphor node to the intervals between two adjacent anchor points. For nodes triggered consecutively across multiple time intervals, extended semantic intervals are formed by connecting them with anchor points; for nodes exhibiting semantic jumps across different time intervals, multiple independent semantic function fragments are formed by separating them with anchor points. In this way, each animal metaphor node forms a clear boundary of function in the semantic space, avoiding semantic diffusion or cross-influence. The final list of boundary anchors not only records the specific time and semantic content of the semantic boundary, but also provides a precise semantic positioning basis for subsequent trigger order adjustment and risk path reasoning.

[0031] Based on the list of boundary anchor points, the node update time sequence in the density anomaly zone is adjusted in a gated manner. The centralized triggering behavior is dispersed by shifting the node update time forward and backward in rotation, and a staggered update table is generated to constrain the trigger synchronization of nodes in the time dimension. Based on the boundary anchor point list, the update time order of animal metaphor nodes in the density anomaly zone is adjusted in a gated manner. The centralized triggering behavior is distributed by alternating the forward and backward shifts of node update time, thereby generating a staggered update table to constrain the synchronization of node triggering in the time dimension. The specific steps are as follows: Extract all temporal anchor information located within density anomaly zones from the generated list of boundary anchors. Each boundary anchor includes a temporal coordinate, the corresponding animal metaphor node category, semantic behavioral features, semantic direction, trigger duration, and preceding and following semantic relationships. First, based on the temporal coordinates, the density anomaly zones are divided into several continuous time segments, each consisting of the time intervals of two adjacent anchors. Then, the animal metaphor nodes within each time segment are arranged according to their trigger times, forming a temporal distribution sequence. After the temporal distribution sequence is formed, further analysis of the semantic dependencies between nodes is required. For predatory nodes, it is necessary to identify their dependent preceding behavioral nodes, such as escape or camouflage nodes; for clustered nodes, it is necessary to identify their semantically triggered similar response nodes; for migrating nodes, it is necessary to identify their semantic connection nodes between preceding and following time segments. Through this analysis, a semantic dependency matrix is ​​obtained, which describes the temporal dependencies of each node. Based on this matrix, nodes are divided into independent node groups and dependent node groups. Independent node groups refer to a set of nodes that have complete semantic expression within the current segment and do not require prior semantic support; dependent node groups are a set of nodes that can only be triggered after the previous node has been triggered. This classification clarifies the semantic triggering logic of each node, laying the foundation for subsequent gating time adjustments.

[0032] Based on the temporal distribution characteristics of density anomaly zones, the time axis is divided into windows, and gating scheduling conditions are established within each time window. First, based on the time segmentation results obtained in the previous step, the entire density anomaly zone is subdivided into several time windows. The length of each time window depends on the average time interval between node triggers within that segment. To ensure temporal continuity, overlapping segments are set between adjacent time windows to smooth time transitions. Within each time window, the concentration of node triggers is calculated. Concentration refers to the proportion of nodes triggering simultaneously within that time window to the total number of nodes in that window. When the concentration exceeds a set threshold, the window is considered a high-density window; when the concentration is below the threshold, it is considered a low-density window. Subsequently, gating time scheduling conditions are established for different types of windows. For high-density windows, it is necessary to reduce the phenomenon of simultaneous node triggering. Therefore, a trigger interval constraint is set in the gating conditions, requiring that the trigger time interval between adjacent nodes must be greater than twice the average trigger interval to ensure temporal spacing. For low-density windows, a trigger filling constraint is set, allowing the update time of some nodes to be advanced to fill time gaps, thereby making the time distribution more uniform. When setting gating conditions, semantic dependencies must also be considered. For example, if an escape node is a trigger condition for a predator node, the gating condition should ensure that the escape node's update time is earlier than the predator node's update time. For nodes with opposite semantic directions, such as camouflage nodes and revelation nodes, their trigger times must be separated into different time windows to prevent semantic cancellation. In this way, the gating time scheduling conditions establish a strict constraint framework on the timeline, providing precise control standards for subsequent time adjustments.

[0033] Based on the established gating and scheduling conditions, the update times of each animal metaphor node are adjusted by alternating between forward and backward shifts. First, a time diffusion operation is performed on nodes in high-density windows, dispersing centrally triggered nodes through forward and backward shifts. A forward shift adjusts the node update time relative to the original time, advancing its trigger time; a backward shift delays the node update time relative to the original time, postponing its trigger time. When performing a forward shift, semantically independent nodes are prioritized. For example, with predator-type nodes, when multiple predator-type nodes trigger simultaneously within the same time window, the node with lower semantic strength or weaker semantic dependency is selected, and its trigger time is advanced by one time interval unit to ensure triggering within the previous window, reducing semantic overlap. For backward shifts, semantically dependent nodes are prioritized. For instance, when escape and predator-type nodes trigger within the same window, the trigger time of the predator-type nodes should be delayed to ensure that escape logically precedes predation. During the time adjustment process, overlapping areas of time windows must also be considered. When a node's update time moves near the window boundary, it's necessary to check if there are any nodes of the same type in adjacent windows. If so, the time needs to be fine-tuned to avoid cross-window triggering leading to a new concentration of time density. Through multiple rounds of forward and backward shifting operations, the triggering order of each node on the timeline is redistributed, forming a uniform and orderly time triggering structure. This ensures that the update time of each node conforms to semantic logic while maintaining the time difference of triggering.

[0034] The update times of all nodes after time-rotation adjustments are summarized and structured to generate the final staggered update table. The staggered update table is time-based, recording the update time, semantic type, semantic direction, semantic dependency relationship, time difference before and after adjustment, and the time window number for each animal metaphor node. Each node's update time record is based on the actual trigger time and annotates the node's semantic level and semantic context. After generating the staggered update table, a trigger order constraint mechanism needs to be established. This mechanism ensures logical consistency in the trigger order of nodes after time adjustment. Specifically, for nodes with causal semantic relationships, such as escape and predator-prey nodes, the trigger order constraint requires that the update time of escape nodes must be earlier than that of predator-prey nodes. For nodes with antagonistic semantic relationships, such as camouflage and revelation nodes, the trigger order constraint requires that the update interval between them must not be less than the length of two time windows. For nodes with cooperative semantic relationships, such as clustering nodes, staggered triggering within consecutive windows is allowed, but only one clustering node must be active within each time window. Once all constraints are established, the resulting staggered update table becomes the core basis for subsequent time series management. This table not only defines the triggering order of each animal metaphor node on the timeline, but also, through a combination of gating logic and rotation strategies, enables flexible adjustment of the time distribution, preventing semantic resonance or excessive risk amplification caused by the synchronous triggering of high-frequency nodes. Ultimately, the staggered update table ensures the stability of the risk map at different time levels and the rationality of semantic propagation during dynamic updates, laying a solid temporal foundation for subsequent trigger quota allocation and risk path optimization.

[0035] Based on the staggered update table, the triggering order between nodes is controlled in a hierarchical manner. An influence intensity quota sequence is introduced, and adjacent animal metaphor nodes are divided into different quota levels. The triggering conditions are reset according to the quota level. The triggering order in the abnormal segment is periodically reversed and controlled. A short cooling time is set during the triggering process so that the risk propagation path is released in segments, thereby avoiding the risk spread out of control caused by trigger resonance. Based on the staggered update table, the triggering order between animal metaphor nodes is hierarchically regulated. By introducing an influence intensity quota sequence, adjacent animal metaphor nodes are divided into different quota levels. The triggering conditions are then reset according to the quota level. Periodic reverse rotation control is implemented for the triggering order within abnormal segments. A short cooldown time is set during the triggering process to allow the risk propagation path to be released in segments, thereby effectively preventing the risk spread out of control due to trigger resonance. The specific steps are as follows: Detailed information about animal metaphor nodes, adjusted for time rotation, is extracted from the staggered update table. This includes update time, semantic category, semantic direction, trigger duration, semantic intensity, and semantic relationship with preceding and following nodes. Density anomaly segments are divided into several consecutive time layers, each representing a triggering phase. The number of nodes, semantic type, and trigger interval within each time layer are fully recorded to construct the trigger hierarchy. Subsequently, semantic dependency chains are determined based on the semantic dependencies between nodes. Semantic dependency chains are node connections formed based on semantic logic, reflecting the triggering order and dependency characteristics between nodes. For example, when the semantic expression of an escape node depends on the behavioral semantics of a predator node in the previous time layer, the escape node is marked as a lower-level node in the semantic dependency chain, while the predator node is an upper-level node. When the semantic behavior of a clustering node depends on the continuous responses of nodes of the same type in adjacent time layers, these nodes form parallel dependency relationships. By establishing this semantic dependency chain, the node triggering sequence of the entire triggering segment can be divided into a multi-level structure: upper-level nodes are mainly semantic triggering sources, middle-level nodes are semantic transition links, and lower-level nodes are semantic release or feedback behaviors. Through hierarchical division, nodes with different semantic roles have relatively independent triggering rhythms on the timeline, providing a clear hierarchical framework for subsequent allocation of influence intensity quotas.

[0036] To constrain and allocate the triggering influence range of nodes at different levels, an influence intensity quota sequence is established. The quota sequence is based on the node's position in the semantic layer, trigger duration, trigger frequency, and semantic propagation capability. First, the trigger frequency of each node on the timeline is quantified, calculating the triggering proportion of each node within the same time layer. Then, the propagation capability is determined by combining the node's semantic type: predatory nodes are defined as high-risk propagation nodes, escape nodes as medium-risk propagation nodes, swarming nodes as cooperative propagation nodes, camouflaged nodes as hidden propagation nodes, and migrating nodes as cross-temporal propagation nodes. Based on these characteristics, the influence intensity of all nodes is divided into high quota level, medium quota level, and low quota level. High quota level nodes mainly include predatory and swarming nodes, which play a core driving role in semantic propagation; medium quota level nodes include escape and migrating nodes, mainly serving the functions of semantic transmission and buffering; low quota level nodes are mainly camouflaged nodes, serving the functions of semantic delay and masking. After the allocation is completed, a quota distribution ratio is established: high-quota nodes should not exceed one-third of the total number of nodes, medium-quota nodes should not be less than one-third, and low-quota nodes should make up the remaining proportion. At this point, the quota level allocation forms a complete semantic strength distribution sequence, giving the node triggering order a hierarchical energy control basis and providing a clear strength basis for the next step of adjusting the triggering conditions.

[0037] Specific triggering conditions are set for nodes with different quota levels to ensure a controllable distribution of triggering behavior in both time and semantics. High-quota nodes, as the dominant element in semantic propagation, have stricter triggering conditions. For high-quota nodes, a minimum time interval condition is set: the triggering interval between adjacent high-quota nodes within the same time window must not be less than twice the average triggering interval, to prevent high-intensity nodes from triggering consecutively in a short period and causing semantic resonance. Furthermore, a semantic exclusivity condition is set: after a high-quota node is triggered, it is prohibited from triggering again in the same semantic direction during its triggering duration, thus maintaining semantic independence. For medium-quota nodes, follow-up and connection conditions are set. The follow-up condition requires that the triggering of a medium-quota node must occur within a short time frame after the triggering of a high-quota node to maintain the causal order of semantic propagation; the connection condition requires that a medium-quota node maintains semantic coherence with the previous node, thus forming a semantic transition. For example, when a predator-type node is triggered, an escape-type node should trigger within a short time, forming a semantic attack-defense connection. For low-quota nodes, compensation and delay conditions are set. Compensation conditions are used to fill time gaps, ensuring the continuity of the semantic propagation chain; delay conditions are used to limit the activation frequency of low-quota nodes during the triggering of high-quota nodes, preventing semantic interference. Through these condition settings, the entire node triggering sequence forms a hierarchical control system, enabling semantic triggering to cooperate and differentiate at different quota levels, thus providing a rule basis for subsequent reverse rotation control.

[0038] To prevent the semantic diffusion path from becoming monotonous and the temporal rhythm from becoming fixed in the long term, a periodic reverse rotation control of the triggering order is required. First, a fixed triggering cycle is defined, for example, a complete semantic propagation cycle. The triggering order of the current cycle is reversed in the next cycle. The principle of reverse rotation is to allow different nodes to take turns assuming the triggering order role in multiple cycles without disrupting the semantic logic. In practice, reverse rotation is implemented in layers according to quota levels. High-quota nodes rotate once every two cycles to maintain the stability of the semantic core; medium-quota nodes rotate once per cycle to balance the semantic propagation rhythm; low-quota nodes rotate within half a cycle to create temporal disturbances and prevent semantic propagation from becoming fixed. During the rotation process, when there are strong logical constraints in the semantic dependency chain, such as predator nodes depending on the triggering of escape nodes, the logical order must remain unchanged, and only nodes without dependency relationships are swapped in order. Through periodic reverse rotation, the trigger sequence of nodes changes continuously over multiple time periods, giving the semantic propagation path a dynamic evolutionary characteristic. This rotation mechanism prevents nodes from remaining fixed in specific trigger positions for extended periods, reducing the likelihood of continuous formation of semantic resonance zones, thereby maintaining the diversity of semantic propagation and the temporal balance of the system.

[0039] To prevent the accumulation of semantic energy and the uncontrolled spread of risk caused by continuous node triggering, a short cooldown period needs to be set between node triggers. The cooldown period refers to the time interval during which a node enters an inactive state after triggering, during which it does not participate in new triggering activities. The length of the cooldown period is determined based on the node's quota level: high-quota nodes have the longest cooldown period to avoid the superposition of continuous high-intensity semantics; medium-quota nodes have a medium cooldown period to maintain a stable propagation rhythm; and low-quota nodes have the shortest cooldown period to fill propagation gaps. In practice, when a high-quota node completes its trigger, it immediately enters a cooldown state and does not accept new semantic stimuli during this period. Subsequent medium-quota nodes begin triggering during the high-quota node's cooldown period, forming a semantic connection. After a medium-quota node finishes triggering, a low-quota node activates after a short cooldown interval, acting as a buffer and releasing energy. The entire triggering process exhibits a hierarchical rhythm of strong triggering-slow triggering-weak triggering, releasing the risk propagation path in segments. Furthermore, to avoid a complete interruption of semantic energy during the cooling-off period, a micro-semantic recovery mechanism is introduced at the end of the cooling interval. This allows nodes to recover some semantic activity for a short period before the cooling-off period ends, ensuring the continuity of propagation. Through this combination of segmented release and cooling control, the risk propagation path forms a rhythmic release pattern on the timeline, transforming semantic diffusion from continuous excitation to periodic excitation. This avoids energy superposition and uncontrolled diffusion caused by semantic resonance, thus maintaining the stability, controllability, and self-equilibrium of the dynamic evolution of the risk map.

[0040] This invention introduces node activation rhythm indexes and semantic overlap fingerprints during the risk graph construction process, enabling synchronous compression and rhythmic expression of multi-source semantic data in the temporal dimension. This allows for the accurate identification of high-density segments triggered by semantics, and the semantic boundaries are split and defined before and after triggering using a list of boundary anchor points, effectively avoiding mis-associations of nodes caused by semantic overlap. This approach provides a precise hierarchical expression of risk semantics in both temporal and semantic dimensions, making the semantic structure of the risk graph clearer and its logical relationships more traceable, fundamentally improving the accuracy of risk identification and semantic reasoning.

[0041] This invention employs a gated update time control and influence intensity quota grading mechanism to hierarchically manage the triggering order of animal metaphor nodes within abnormal segments. It also introduces periodic reverse rotation and short cooldown periods during the triggering process, achieving rhythmic release of risk semantic propagation. This approach effectively reduces the probability of synchronous triggering between nodes, suppresses the cumulative effect of semantic resonance, and allows the risk propagation path to unfold in a segmented manner. Even under high-frequency input conditions, the system maintains structural stability and controllable propagation, thereby improving the dynamic evolutionary stability and interpretability of the risk map.

[0042] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A risk mapping method based on animal metaphors, characterized in that, Includes the following steps: Multi-source semantic data is compressed onto a unified timeline in chronological order. Risk information in the semantic data is organized, and a correspondence between risk factors and animal behavioral characteristics is established using animal metaphors to generate a node activation rhythm index. The steps for generating the node activation rhythm index are as follows: Multi-source semantic data is compressed into a unified timeline in chronological order, and the time units of the data are unified and aligned to form a continuous semantic stream; Extract risk sources, propagation paths, carriers, defensive behaviors, feedback results, intervention measures and post-effect states from semantic data on the timeline to generate a set of risk semantic units; Establish a correspondence between the risk semantic unit set and the animal metaphor category, and generate a metaphor correspondence table containing time, semantic risk type, metaphor category, behavioral tendency and contextual information; Based on the metaphor correspondence table, the trigger interval, occurrence frequency and duration of animal metaphor nodes are statistically analyzed to distinguish periodic, intermittent and random trigger types, and the trigger frequency, time interval and semantic association weight are recorded to generate a node activation rhythm index. Based on the node activation rhythm index, the trigger frequency, duration and change amplitude of each animal metaphor node in a continuous time window are statistically calculated to identify the time segment of sudden increase in trigger density, and semantic overlap fingerprints are generated based on the identification results. The steps for generating semantically overlapping fingerprints are as follows: Based on the timestamp information in the node activation rhythm index, the triggering events of the animal metaphor nodes are arranged in chronological order. A combination of fixed length and sliding overlap is used to divide the continuous time window, and an overlap area is introduced at the window boundary to maintain time continuity. The number of triggers, intervals, and durations of animal metaphor nodes within each time window are statistically analyzed. The trigger start point, duration, and end point are recorded, and a weighted average is calculated to reflect the level of trigger activity. By comparing the trigger frequency and duration results of adjacent time windows, nodes with increased trigger frequency, extended duration, or shortened interval in continuous windows are identified to determine the trigger density surge segment. Extract the time range and participating nodes of the trigger density surge segment, record the trigger time, duration, semantic direction and metaphor category, and generate semantic overlap fingerprint based on semantic direction consistency; The semantic boundaries of the trigger density anomaly segment are split using semantic overlap fingerprints, the trigger positions are distinguished and marked, and semantic correspondence is established based on the context relationship to generate a list of boundary anchor points. Based on the list of boundary anchor points, the node update time sequence in the density anomaly zone is adjusted in a gated manner. The behavior of centralized triggering is distributed by shifting the node update time forward and backward in rotation, and a staggered update table is generated. Based on the staggered update table, the triggering order between nodes is controlled in a hierarchical manner. An influence intensity quota sequence is introduced to divide adjacent animal metaphor nodes into different quota levels. The triggering conditions are reset according to the quota level. The triggering order in the abnormal segment is controlled by periodic reverse rotation. A short cooldown time is set during the triggering process to release the risk propagation path in segments.

2. The risk mapping method based on animal metaphors according to claim 1, characterized in that, In the process of generating semantic overlap fingerprints, animal metaphor nodes in the trigger density surge segment are grouped according to semantic direction consistency and behavioral convergence to form semantic overlap groups, and a hierarchical time series structure is established based on the trigger time, duration and semantic fusion degree of the nodes in the group.

3. The risk mapping method based on animal metaphors according to claim 1, characterized in that, The steps to generate the boundary anchor point list are as follows: Based on the semantic overlap fingerprint parsing time segment identifier, node category, semantic fusion degree, trigger frequency distribution and duration characteristics, semantic clustering is performed on the nodes within the time segment to determine the semantic core nodes and auxiliary semantic intervals; Analyze the semantic behavior type, semantic object, semantic transmission direction and semantic emotional color of each animal metaphor node within the semantic interval, identify semantic overlap area, semantic conflict area and semantic transition area and record the hierarchical relationship; On the timeline, semantic overlay areas, semantic conflict areas, and semantic transition areas are located, a time boundary marker sequence is generated, and semantic continuity, semantic transformation, and semantic differentiation correspondences are established based on the behavior type, semantic direction, and target of the preceding and following semantic nodes. Anchoring is performed on the time boundary markers to generate a list of boundary anchors that includes time coordinates, node type, semantic behavior attributes, semantic direction, and corresponding information in the surrounding context.

4. The risk mapping method based on animal metaphors according to claim 3, characterized in that, When generating the list of boundary anchor points, the anchor point information is recorded in chronological order, with the anchor points corresponding to the semantic core nodes arranged first. The anchor point hierarchy is determined according to semantic priority, and the semantic interval is divided by the anchor point time coordinate. The semantic function of the animal metaphor node is limited to adjacent anchor points to form a continuous or segmented semantic function boundary.

5. The risk mapping method based on animal metaphors according to claim 3, characterized in that, The steps for generating the staggered update table are as follows: Extract time anchor information within density anomaly segments from the boundary anchor list, divide continuous time segments based on time coordinates, and divide animal metaphor nodes into independent node groups and dependent node groups according to semantic dependency relationships to form a semantic dependency matrix. Time windows are divided according to time distribution characteristics, time overlap areas are set to ensure time continuity, node trigger concentration is calculated and gating scheduling conditions are established, trigger interval constraints are set for high-density windows and trigger filling constraints are set for low-density windows. Based on the gating scheduling conditions, the animal metaphor nodes are adjusted by alternating forward and backward movements, prioritizing the adjustment of the update time order of semantically independent nodes and dependent nodes to avoid concentrated triggering of behaviors; The results of the update time adjustment are recorded in a structured manner, generating a staggered update table based on time, and a trigger order constraint mechanism is established to ensure the time consistency of the node triggering logic.

6. The risk mapping method based on animal metaphors according to claim 5, characterized in that, The trigger sequence hierarchical control steps are as follows: Extract the update time, semantic category, semantic direction, trigger duration, semantic intensity and semantic relationship information of animal metaphor nodes from the staggered update table, divide the continuous time layer according to time and establish a semantic dependency chain to construct the trigger hierarchy structure; An influence intensity quota sequence is established based on temporal layer and semantic dependency relationship. Animal metaphor nodes are divided into high quota level, medium quota level and low quota level according to the triggering frequency, duration and propagation ability of the nodes. Triggering conditions are set according to quota levels. Time interval constraints and semantic exclusion conditions are set for high quota nodes, follow conditions and connection conditions are set for medium quota nodes, and compensation conditions and delay conditions are set for low quota nodes. Implement periodic reverse rotation control to adjust the triggering order of nodes of different quota levels periodically to maintain semantic balance; Set a short cooldown period between node triggers, and determine the cooldown duration based on the quota level.

Images