An intelligent evaluation method and system for the value of historical and cultural resources in urban renewal

Abstract

The application belongs to the technical field of urban renewal, and discloses a historical and cultural resource value intelligent evaluation method and system in urban renewal, which comprises the following steps: acquiring static attribute data and dynamic sensing data of a plurality of historical and cultural resources in a target area; constructing a topological correlation model based on the correlation among the historical and cultural resources; taking the historical and cultural resources as nodes and taking the static attribute data and the dynamic sensing data as feature vectors of the nodes; performing feature aggregation processing on the topological correlation model by using a graph learning algorithm, identifying and quantifying the interactive relationship attributes between adjacent nodes; calculating individual evaluation values of the historical and cultural resources, and determining corresponding value correction components according to the interactive relationship attributes; and comprehensively calculating the individual evaluation values and the value correction components to obtain a whole value evaluation result of the target area, so that the real whole value of a historical area can be accurately judged.

Description

Technical Field

[0001] This application relates to the field of urban renewal technology, and more specifically, to a method and system for intelligent assessment of the value of historical and cultural resources in urban renewal. Background Technology

[0002] In the context of urban renewal, the scientific assessment of historical and cultural resources is a crucial step for planning departments in formulating protection and development strategies. While existing intelligent assessment systems can evaluate the value of individual historical and cultural resources, such as an ancient building or a historical archway, from multiple perspectives and identify areas where resources are clustered based on geographical proximity, these systems fall short in understanding the interactions between different resources within these clusters. They cannot distinguish whether these resources are mutually reinforcing, mutually antagonistic, or simply exist independently. This lack of understanding of the deep connections between resources makes it difficult for the system to accurately determine the overall value of a historical area, thus affecting the scientific validity and effectiveness of macro-level protection and renewal strategies.

[0003] Specifically, existing intelligent assessment systems, when processing historical and cultural resources, often treat each resource point and its sensor data as an independent entity for analysis. They can accurately report "peak visitor traffic at ancient building A on weekends" or "abnormal air quality in a section of historical district B," but struggle to understand the deeper connections behind these data. For example, the system might detect a surge in visitors to an ancient stage (resource X) in a historical district during a specific period, while simultaneously, a traditional handicraft workshop (resource Y) at the other end of the district also experiences a significant increase in visitors during the same time. The system can identify these two independent phenomena, but it cannot automatically establish a causal or correlational relationship between them. This blind spot regarding the potential interactions between resource points prevents the system from understanding, at a macro level, how different resources within a region influence each other and collectively constitute a holistic cultural ecosystem.

[0004] To address the limitations of this isolated analysis, the system development team decided that they couldn't just look at the data of individual resource points; they also needed to consider their spatial distribution. The system was further upgraded to include spatial clustering analysis. It can automatically identify "value clusters" composed of multiple historical and cultural resource points based on geographical proximity. When the system detects that within a relatively small geographical area, there are multiple resource points with high individual value scores, and these resource points exhibit a close spatial clustering, it marks this area as a potential "overall value region."

[0005] However, this clustering method based on spatial proximity remains too crude in practical applications and encounters deeper bottlenecks. While the system can identify "a resource cluster," it cannot further distinguish and characterize the complex functional and cultural relationships between different resources within this cluster. For example, an area identified as a "value cluster" might contain a well-preserved Ming and Qing dynasty house, a modern creative café, and a historic stone-paved road. The system can assess the historical value of the house, the social activity of the café, and the cultural significance of the stone-paved road, but it cannot determine whether the relationship between these three elements is mutually reinforcing (e.g., the house and the road together create a historical atmosphere, while the café serves as a source of modern vitality with a harmonious style), mutually conflicting (e.g., the modern style of the café clashes with the house, disrupting the overall aesthetic), or simply independent and unaffected. In other words, the system cannot distinguish whether this cluster is a "harmoniously coexisting cultural ecosystem" or a "conflicting architectural hodgepodge." This lack of identification of the intrinsic connections between resources makes it impossible for the system to accurately measure the overall added value or detrimental effect of resource combinations, ultimately making it difficult to accurately judge the true overall value of historical areas. This leaves planning departments without the most crucial, in-depth basis when formulating larger-scale block-level or district-level protection and renewal strategies, potentially leading to improper resource allocation or even damage to the overall cultural fabric of the area.

[0006] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0007] This invention discloses an intelligent assessment method and system for the value of historical and cultural resources in urban renewal, aiming to solve the shortcomings of existing assessment systems for historical and cultural resources in urban renewal in understanding the deep interactive relationships between resources, and the resulting inaccurate judgment of the overall value of the region.

[0008] In a first aspect, the present invention provides an intelligent assessment method for the value of historical and cultural resources in urban renewal, used to assess the overall value of a target area containing multiple historical and cultural resources; characterized in that the method includes the following steps: A1. Obtain static attribute data and dynamic sensing data of multiple historical and cultural resources within the target area; A2. Construct a topological association model based on the relationships between the historical and cultural resources; the topological association model uses the historical and cultural resources as nodes, and uses the static attribute data and the dynamic perception data as feature vectors of the nodes; A3. Use graph learning algorithms to perform feature aggregation processing on the topological association model to identify and quantify the interaction relationship attributes between adjacent nodes; the interaction relationship attributes include relationship type and relationship strength. A4. Calculate the individual evaluation value of each of the historical and cultural resources, and determine the corresponding value correction component based on the interaction relationship attributes; A5. By combining the individual assessment values ​​and the value correction components, the overall value assessment result of the target area is calculated.

[0009] Secondly, this application provides an intelligent assessment system for the value of historical and cultural resources in urban renewal, used to assess the overall value of a target area containing multiple historical and cultural resources; the system includes: The data acquisition module is used to acquire static attribute data and dynamic sensing data of multiple historical and cultural resources within the target area; The model building module is used to construct a topological association model based on the relationship between the historical and cultural resources; the topological association model uses the historical and cultural resources as nodes, and uses the static attribute data and the dynamic perception data as the feature vectors of the nodes. The aggregation processing module is used to perform feature aggregation processing on the topological association model using graph learning algorithms, and to identify and quantify the interaction relationship attributes between adjacent nodes; the interaction relationship attributes include relationship type and relationship strength. The component calculation module is used to calculate the individual evaluation value of each of the historical and cultural resources, and determine the corresponding value correction component based on the interaction relationship attribute. The result generation module is used to combine the individual assessment values ​​and the value correction components to calculate the overall value assessment result of the target area.

[0010] Beneficial Effects: This application provides an intelligent assessment method and system for the value of historical and cultural resources in urban renewal. It acquires static attribute data and dynamic perception data of historical and cultural resources, constructs a topological association model based on the relationships between resources, and uses graph learning algorithms to perform feature aggregation processing on the model, identifying and quantifying the interaction attributes between adjacent nodes. Based on this, it calculates the individual assessment value of each historical and cultural resource and determines the corresponding value correction component according to the interaction attributes. Finally, it combines the individual assessment values ​​and the value correction component to calculate the overall value assessment result of the target area. This effectively solves the technical problems of existing technologies that treat historical and cultural resources as independent individuals for analysis, failing to deeply understand the interactions between resources and accurately judging the overall value of historical areas. By introducing a topological association model and graph learning algorithms, this application can understand from a macro perspective how different resources within a region influence each other and jointly constitute a holistic cultural ecology, overcoming the blind spots of traditional assessment systems regarding potential interactions between resources. Furthermore, by identifying and quantifying the enhancing or conflicting relationships between resources and determining the value correction component accordingly, this application can accurately measure the overall value-added or diminishing effect of resource combination, thereby making an accurate judgment on the true overall value of the historical area. This provides a crucial and in-depth basis for planning departments to formulate larger-scale block-level or area-level protection and renewal strategies, avoiding the risk of improper resource allocation or even damage to the overall cultural fabric of the area. Attached Figure Description

[0011] Figure 1 A flowchart of an intelligent assessment method for the value of historical and cultural resources in urban renewal provided for this application.

[0012] Figure 2 This application provides a structural diagram of an intelligent assessment system for the value of historical and cultural resources in urban renewal.

[0013] Labeling Explanation: 1. Data Acquisition Module; 2. Model Building Module; 3. Aggregation Processing Module; 4. Component Calculation Module; 5. Result Generation Module. Detailed Implementation

[0014] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0015] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0016] Reference Figure 1 This invention proposes an intelligent assessment method for the value of historical and cultural resources in urban renewal, used to assess the overall value of a target area containing multiple historical and cultural resources; characterized in that the method includes the following steps: A1. Obtain static attribute data and dynamic sensing data of multiple historical and cultural resources within the target area; A2. Construct a topological association model based on the relationships between the historical and cultural resources; the topological association model uses the historical and cultural resources as nodes, and uses the static attribute data and the dynamic perception data as feature vectors of the nodes; A3. Use graph learning algorithms to perform feature aggregation processing on the topological association model to identify and quantify the interaction relationship attributes between adjacent nodes; the interaction relationship attributes include relationship type and relationship strength. A4. Calculate the individual evaluation value of each of the historical and cultural resources, and determine the corresponding value correction component based on the interaction relationship attributes; A5. By combining the individual assessment values ​​and the value correction components, the overall value assessment result of the target area is calculated.

[0017] This application, by introducing a topological association model and graph learning algorithm, can deeply explore the interactive relationships between historical and cultural resources, thereby more comprehensively and accurately assessing the overall value of the target area and overcoming the limitations of existing technologies in understanding the deep connections between resources.

[0018] "Historical and cultural resources" refer to tangible and intangible heritage with historical, cultural, artistic, or scientific value within urban renewal areas, such as ancient buildings, historical blocks, traditional handicraft workshops, and cultural sites.

[0019] "Static attribute data" refers to the inherent and relatively stable characteristic data of historical and cultural resources, such as the type of resources, architectural style, and historical period.

[0020] "Dynamic sensing data" refers to sensing data that reflects the real-time operational status and environmental changes of historical and cultural resources, such as real-time pedestrian flow, ambient temperature and humidity, and noise levels.

[0021] The "topological association model" is a graph structure model in which historical and cultural resources are abstracted as nodes, and the relationships between resources are abstracted as edges.

[0022] "Graph learning algorithms" are a class of machine learning algorithms used to process graph-structured data. They can learn the features of nodes and edges and perform feature aggregation.

[0023] "Interactive relationship attributes" refer to the nature of the interaction between historical and cultural resources, including the type of relationship (such as enhancement and conflict) and the strength of the relationship.

[0024] "Individual assessment value" refers to the value assessment result of a single historical and cultural resource.

[0025] "Value adjustment component" is the amount of adjustment made to the overall value based on the interaction between historical and cultural resources.

[0026] The “overall value assessment result” is the total value after the combined effect of all historical and cultural resources within the target area.

[0027] This application provides an intelligent assessment method for the value of historical and cultural resources in urban renewal. This method achieves the assessment of the overall value of the target area through a series of steps.

[0028] In step A1, it is necessary to acquire static attribute data and dynamic sensing data of multiple historical and cultural resources within the target area. Static attribute data reflects the inherent characteristics of historical and cultural resources. For example, information such as the construction date, architectural style, and historical events of ancient buildings can be obtained through manual input, consulting historical documents, or utilizing professional databases. Dynamic sensing data reflects the real-time operational status of historical and cultural resources. For example, a sensor network can be deployed in the area where the historical and cultural resources are located to collect data such as pedestrian traffic, ambient temperature and humidity, and noise levels in real time. For instance, in a historical district, the architectural style of one ancient house can be manually entered as "Huizhou style," and its construction date as "Qing Dynasty." Simultaneously, infrared sensors installed at the entrance of the district can acquire real-time pedestrian traffic data for that district.

[0029] In step A2, a topological association model is constructed based on the relationships between historical and cultural resources. This model uses historical and cultural resources as nodes, and static attribute data and dynamic perception data as feature vectors for each node. For example, each ancient building, each site, and each traditional shop in a historical district can be abstracted as a node. The feature vectors of these nodes can be obtained by splicing or fusing the static attribute data (such as architectural style and historical period) and dynamic perception data (such as real-time pedestrian flow and environmental temperature and humidity) obtained in step A1. Relationships can include geographical distance. For example, if the geographical distance between two historical and cultural resources is less than a preset threshold, an edge can be established between them, and an initial weight can be assigned based on the reciprocal of the distance; the closer the distance, the greater the weight.

[0030] In step A3, graph learning algorithms are used to perform feature aggregation on the topological association model, identifying and quantifying the interaction relationship attributes between adjacent nodes. These interaction relationship attributes include relationship type and relationship strength. For example, a graph neural network (GNN) can be used to process the topological association model. A GNN determines complex relationships between nodes by transmitting and aggregating feature information between them. Specifically, each layer of the GNN aggregates the feature information of adjacent nodes to the current node, thereby updating the feature representation of the current node. In this way, the model can determine whether there are mutually reinforcing or mutually inhibiting relationships between adjacent historical and cultural resources and quantify the strength of such relationships. For example, through graph learning algorithms, the model might identify a "reinforcing relationship" between an ancient opera stage and an adjacent traditional handicraft workshop, with a high relationship strength, indicating that the activities of the opera stage significantly enhance the attractiveness of the workshop.

[0031] In step A4, the individual assessment value of each historical and cultural resource is calculated, and the corresponding value correction component is determined based on the interaction relationship attributes. The individual assessment value can be obtained by weighted summation of multiple dimensions of the historical and cultural resource, such as historical grade, artistic value rating, average visitor flow, and average visitor dwell time. These indicators can be normalized before weighted summation. The value correction component is determined based on the interaction relationship attributes identified in step A3. If the relationship type is an enhancing relationship, a value gain component will be generated; if the relationship type is a conflicting relationship, a value loss component will be generated. For example, if the relationship type is an enhancing relationship, the value gain component can be determined based on the relationship strength, the sum of the individual assessment values ​​of the two corresponding nodes, and the product of a preset gain coefficient, serving as the value correction component. If the relationship type is a conflicting relationship, the value loss component can be determined based on the relationship strength, the sum of the individual assessment values ​​of the two corresponding nodes, and the product of a preset loss coefficient, serving as the value correction component. This correction mechanism ensures that the assessment results reflect not only the value of individual resources but also the overall added value or reduced value effect produced by the combination of resources.

[0032] In step A5, the overall value assessment result of the target area is calculated by combining the individual assessment values ​​and the value correction components. This can be obtained by summing the individual assessment values ​​of all historical and cultural resources, adding all value gain components, and then subtracting all value loss components. For example, if there are three historical and cultural resources in a target area, with individual assessment values ​​of 80, 75, and 90 respectively, and if there is an enhancing relationship between resource 1 and resource 2 that generates a +10-point gain, and a conflicting relationship between resource 2 and resource 3 that generates a -5-point loss, then the overall value assessment result of the target area will be (80+75+90)+10-5=250 points.

[0033] The intelligent assessment method for the value of historical and cultural resources in urban renewal proposed in this application works by first acquiring multi-source data to comprehensively capture the static characteristics and dynamic states of historical and cultural resources. Then, these resources and their data are abstracted into nodes and feature vectors in a graph structure to construct a topological association model, thereby integrating discrete resource points into a unified association network. Crucially, graph learning algorithms are used to perform feature aggregation processing on this topological association model. This allows the system to go beyond simple spatial proximity judgments and identify complex and hidden interactive relationships between adjacent resources, including the type (e.g., enhancement or conflict) and intensity of these relationships. Based on this, the method not only calculates the individual value of each historical and cultural resource but, more importantly, introduces value correction components based on the identified interactive relationship attributes. These correction components quantify the value gain or loss resulting from mutual promotion or conflict between resources. Finally, by integrating individual assessment values ​​and value correction components, the method can calculate the overall value assessment result of the target area. These steps work together to enable the system to understand, from a macro perspective, how different resources within a region influence each other and together form a holistic cultural ecosystem. This overcomes the limitations of isolated analysis in existing technologies and achieves an accurate assessment of the overall value of historical areas.

[0034] The intelligent assessment method for the value of historical and cultural resources in urban renewal proposed in this application has significant advantages and innovations compared to existing technologies. Traditional methods often analyze historical and cultural resources as independent entities or simply cluster them based on geographical proximity, resulting in an inability to deeply understand the complex interactions between resources and thus making it difficult to accurately assess the overall value of a historical area. For example, existing systems may be able to identify the foot traffic data of an ancient stage and an adjacent traditional handicraft workshop in a historical district, but they cannot automatically establish a causal relationship between the two, where stage performances drive business in the workshops.

[0035] This application effectively solves the aforementioned problems by introducing a topological association model and graph learning algorithms. First, historical and cultural resources are abstracted as graph nodes, and their static attribute data and dynamic perception data are used as feature vectors to construct a topological association model that comprehensively reflects the characteristics of the resources. Second, graph learning algorithms are used to aggregate features of this model, enabling the system to identify the interactive relationship attributes between adjacent nodes, including relationship type (such as reinforcement and conflict) and relationship strength. This innovation allows this method to go beyond simple spatial proximity and delve deeper into the functional and cultural connections between resources. For example, through graph learning algorithms, this method can identify an "reinforcing relationship" between the ancient opera stage and traditional handicraft workshops and quantify its strength, thus determining that the activities of the opera stage do indeed significantly enhance the attractiveness of the workshops.

[0036] Furthermore, this method, in calculating the overall value, not only considers the individual assessment values ​​of each historical and cultural resource, but more importantly, introduces value correction components determined based on the attributes of their interactive relationships. These correction components can quantify the value gain or loss effect generated by the combination of resources, thus making the overall value assessment results more comprehensive and accurate. For example, in an area containing old houses, cafes, and cobblestone streets, this method can distinguish whether these three elements constitute a "harmonious and symbiotic cultural ecosystem" or a "conflicting architectural hodgepodge," and adjust the overall value accordingly. This identification and quantification of the intrinsic connections between resources enables this method to accurately judge the true overall value of historical areas, providing crucial in-depth evidence for planning departments to formulate larger-scale block-level or area-level protection and renewal strategies, avoiding the risk of improper resource allocation or even destruction of the overall cultural fabric of the region. Therefore, this application has significant technological advancements and practical value in the field of historical and cultural resource value assessment in urban renewal.

[0037] In some implementations, the static attribute data includes the type, architectural style, and historical period of the historical and cultural resources; the dynamic sensing data includes real-time pedestrian traffic, ambient temperature and humidity, and noise levels in the area where the historical and cultural resources are located.

[0038] Specifically, the static attribute data refers to relatively stable information that reflects the inherent characteristics of historical and cultural resources. Among these, the type of historical and cultural resource can refer to its classification as ancient architecture, historical sites, cultural heritage, museums, etc., which directly affects its cultural attributes and protection level; architectural style can refer to the regional characteristics, era style, or school characteristics it embodies, such as Ming and Qing architectural style or Western classical architectural style, which is crucial for assessing its artistic and historical value; historical period clarifies the historical period in which the resource exists, helping to trace its historical evolution and cultural accumulation.

[0039] Furthermore, the dynamic sensing data refers to timely information reflecting the real-time operational status and environmental changes of historical and cultural resources. Specifically, real-time visitor flow in the area where the historical and cultural resources are located reflects the current attractiveness, popularity, and potential carrying capacity of the resources; environmental temperature and humidity directly affect the preservation of historical and cultural resources and the visitor experience, as excessively high or low temperatures and humidity may damage cultural relics or affect visitor comfort; and noise levels in decibels reflect the tranquility of the surrounding environment, which is of great significance for assessing its cultural atmosphere and visitor experience.

[0040] This application's solution, by clearly defining the specific composition of static attribute data and dynamic perception data, ensures a comprehensive and detailed capture of the various characteristics of historical and cultural resources during the evaluation process. Static attribute data provides a stable, intrinsic value foundation for the evaluation, enabling more accurate judgments of the resource's historical, cultural, and artistic value. Dynamic perception data supplements this with real-time, external operational status and environmental impact information, allowing the evaluation results to reflect the resource's activity, attractiveness, and challenges in the current context. Therefore, by combining these two types of data, a more three-dimensional and dynamic evaluation perspective can be constructed, providing high-quality input for subsequent topological association model construction and feature aggregation processing.

[0041] In some implementations, step A2 includes: A201. The historical and cultural resources are abstracted into nodes, and the static attribute data and the dynamic perception data are concatenated to obtain the feature vector of the node; the node is a graph node; A202. Calculate the distance between the geographic centers of any two of the aforementioned historical and cultural resources; A203. If the distance between the geographic centers is less than a preset threshold, an initial edge is established between the two corresponding nodes, and an initial weight is assigned to the initial edge based on the reciprocal of the distance between the geographic centers.

[0042] Abstracting historical and cultural resources into nodes means representing each independent historical and cultural entity (e.g., ancient buildings, ruins, historical districts, museums, etc.) within the target area as an independent graph node in the topological association model. The static attribute data and the dynamic sensing data are considered as feature vectors of this node, integrated through a concatenation method to comprehensively describe the inherent characteristics and real-time operational status of the historical and cultural resource. For example, static attribute data may include the resource type, architectural style, historical period, etc., while dynamic sensing data may include real-time pedestrian traffic, environmental temperature and humidity, noise levels, etc.

[0043] Furthermore, calculating the geographic center distance between any two historical and cultural resources refers to obtaining the geographic coordinates of each historical and cultural resource through a Geographic Information System (GIS) or other positioning technology, and calculating the straight-line distance or actual path distance between them.

[0044] In a preferred implementation, when the geographical center distance is less than a preset threshold, it indicates that the two historical and cultural resources are sufficiently close in space and may have a direct or indirect connection. In this case, an initial edge is established between the two corresponding nodes to represent this potential connection. The weight of the initial edge is determined based on the reciprocal of the geographical center distance (e.g., by multiplying the reciprocal of the geographical center distance by a preset proportional coefficient). This means that the closer the resources are, the stronger their initial connection, and vice versa. The preset threshold can be adjusted according to the actual application scenario and the requirements for the tightness of the connection; for example, it can be set to 500 meters, 1 kilometer, etc.

[0045] This application's approach visualizes historical and cultural resources as graph nodes, using their static attribute data and dynamic perception data as feature vectors for these nodes, providing a rich data foundation for subsequent graph learning algorithms. By calculating the geographic center distance between historical and cultural resources and setting a threshold, spatially adjacent or related resources can be effectively identified. Initial weights are assigned based on the reciprocal of the distance (e.g., using [reference]), thus initially constructing a topological structure reflecting spatial relationships. This method of constructing initial relationships based on geographic distance lays the foundation for subsequent feature aggregation processing, enabling the model to understand the interactions between historical and cultural resources from a spatial dimension.

[0046] In practical implementation, efficient and accurate feature aggregation, as well as precise identification of relationship types and quantification of relationship strength, are crucial for improving the accuracy of evaluation results. To address this, this application further proposes specific steps to optimize the application of graph learning algorithms in feature aggregation processing. Specifically, step A3 may include: A301. The feature vectors of the nodes are transformed and aggregated using the layered attention layers of the graph attention network to determine the attention weights between adjacent nodes; A302. Based on the attention weights, a classifier is used to identify the relationship type between adjacent nodes, and a regressor is used to calculate the relationship strength between adjacent nodes.

[0047] Specifically, a Graph Attention Network (GAT) is a graph neural network that aggregates features of neighboring nodes by determining importance weights (i.e., attention weights) between nodes. The attention layer is the core component of the GAT, its role being to perform a weighted summation of the feature vectors of each node and its neighbors, thereby generating a new, more representative feature representation of that node. In this way, complex dependencies between neighboring nodes can be effectively captured, providing high-quality feature input for subsequent relation identification and strength quantification. Furthermore, after determining the attention weights between neighboring nodes, these weights reflect the degree to which different neighboring nodes contribute to the feature aggregation of the central node. Based on these attention weights, a classifier can be used to identify the type of relationship between neighboring nodes. For example, the classifier can be a Multilayer Perceptron (MLP) or a Support Vector Machine (SVM), whose input is the aggregated node features and attention weights, and whose output is a predefined relation type, such as "enhancing relationship," "conflicting relationship," or "irrelevant relationship." Simultaneously, a regressor can be used to calculate the relation strength between neighboring nodes. The regressor can also be a Multilayer Perceptron, with similar input to the classifier, but its output is a continuous numerical value representing the magnitude of the relation strength. By combining classifiers and regressors, the interactive relationship attributes between historical and cultural resources can be comprehensively identified and quantified.

[0048] This application's solution introduces a layered attention layer in a graph attention network, enabling dynamic determination of attention weights between neighboring nodes during feature aggregation. This mechanism overcomes the limitation of traditional graph learning algorithms that treat all neighboring nodes equally, thus allowing for a more refined capture of the differences in the importance of interactions between different historical and cultural resources. It is precisely this adaptive attention mechanism that enables the aggregated node features to more accurately reflect the true connections between historical and cultural resources. Furthermore, by combining classifiers and regressors to identify and quantify relationship types and strengths, the comprehensiveness and accuracy of interaction relationship attributes are ensured, providing a reliable basis for subsequent value correction component calculations.

[0049] In some implementations, step A4, calculating the individual assessment value for each of the historical and cultural resources, includes: Obtain the historical rating, artistic value score, average visitor traffic, and average visitor dwell time for each of the aforementioned historical and cultural resources; The historical rating, the art value score, the average foot traffic, and the average visitor dwell time are normalized. For each historical and cultural resource, the normalized historical grade, normalized artistic value score, normalized average visitor flow, and normalized average visitor dwell time are weighted and summed to obtain the individual evaluation value of the historical and cultural resource.

[0050] Obtaining historical ranking, artistic value rating, average visitor traffic, and average visitor dwell time for various historical and cultural resources involves collecting multi-dimensional data to comprehensively reflect the inherent value of these resources and their social activity and attractiveness from different perspectives. This can be achieved using various data sources and collection methods. For example, historical ranking can be obtained by consulting official cultural relic survey data, local chronicles, historical documents, or expert evaluation reports, typically expressed using a grading system (e.g., national, provincial, municipal cultural relic protection units) or a rating system (e.g., 1-5 points). Artistic value ratings can be assessed by professional art historians, architects, or cultural heritage experts, providing a quantitative score based on factors such as the resource's architectural style, craftsmanship, and cultural connotations. Average visitor traffic and average visitor dwell time can be obtained through real-time monitoring and data statistics at the entrance and internal areas of historical and cultural resources using multimodal sensor networks (e.g., Wi-Fi probes, Bluetooth beacons, infrared counters, video surveillance systems combined with AI analysis), or by analyzing indirect data sources such as online travel platforms, social media check-in data, and ticket sales records. These data can be averaged at daily, weekly, or monthly granularities to reflect the long-term attractiveness of the resources.

[0051] Normalizing historical ratings, artistic value scores, average foot traffic, and average visitor dwell time involves converting data with different dimensions and numerical ranges to a uniform scale, eliminating dimensional differences, and ensuring comparability between different evaluation dimensions. Various normalization methods can be used to achieve this. For example, min-max normalization can linearly scale data to the [0,1] interval. Z-score normalization (Standardization) can transform data into a distribution with a mean of 0 and a standard deviation of 1. The choice of normalization method depends on the distribution characteristics of the data and the evaluation requirements. For example, if the data contains extreme values, Z-score normalization may be more robust.

[0052] For each historical and cultural resource, a weighted sum is calculated based on the normalized historical grade, normalized artistic value score, normalized average visitor flow, and normalized average visitor dwell time to obtain the individual evaluation value of that resource. During the weighted summation process, different weights can be assigned to different indicators based on actual evaluation needs and expert experience to reflect their relative importance in the overall individual value composition. For example, in some scenarios, historical grade and artistic value score may be given higher weights, while in others, average visitor flow and average visitor dwell time may be more important.

[0053] This application's solution comprehensively considers the inherent attributes of historical and cultural resources (such as historical grade and artistic value rating) and their current operational status (such as average foot traffic and average visitor dwell time), and standardizes and weights these multi-dimensional data to ensure the comprehensiveness and objectivity of the individual value assessment of each historical and cultural resource. This method can systematically capture the key factors affecting resource value and present them in a quantitative form, laying a solid foundation for subsequent overall value assessment.

[0054] In some implementations, step A4, determining the corresponding value correction component based on the interaction relationship attribute, includes: If the relationship type is an enhanced relationship, then the value gain component is determined as the value correction component based on the product of the relationship strength, the sum of the individual evaluation values ​​of the two corresponding nodes, and the preset gain coefficient. If the relationship type is a conflict relationship, then the value loss component is determined as the value correction component based on the product of the relationship strength, the sum of the individual evaluation values ​​of the two corresponding nodes, and the preset loss coefficient.

[0055] Specifically, the interaction relationship attribute refers to the relationship type and strength between adjacent nodes identified and quantified in step A3. The relationship type can be understood as the nature of the interaction between historical and cultural resources, such as a "reinforcing relationship" that promotes mutual benefit and synergy, or a "conflicting relationship" that restricts each other and generates negative impacts. The relationship strength quantifies the closeness or influence of this interaction. When an reinforcing relationship is identified between two historical and cultural resources, it indicates a synergistic effect, enabling them to jointly enhance each other's value. In this case, the value correction component is determined as the value gain component, calculated as the product of the relationship strength, the sum of the individual evaluation values ​​of the two corresponding historical and cultural resources, and a preset gain coefficient. The gain coefficient is a preset parameter used to adjust the contribution of the reinforcing relationship to value enhancement. Conversely, when a conflicting relationship is identified between two historical and cultural resources, it indicates potential competition, interference, or negative impacts, which may lead to a decrease in overall value. In this case, the value correction component is determined as the value loss component, calculated as the product of the relationship strength, the sum of the individual evaluation values ​​of the two corresponding historical and cultural resources, and a preset loss coefficient. The loss coefficient is a preset parameter used to adjust the degree of impact of conflict relationships on value reduction. The individual assessment value refers to the value assessment result of each historical and cultural resource calculated in step A4, reflecting its inherent value.

[0056] This application's solution, by introducing a distinction between the types of interaction relationships when determining the value correction component and calculating value gain or value loss components separately based on their nature, can more accurately reflect the complex interactions between historical and cultural resources. Specifically, when there is an enhancing relationship between two historical and cultural resources—for example, when they are complementary in cultural themes, functional positioning, or visitor experience—their combined value will be greater than the value of simple summation. Calculating the value gain component quantifies the additional value brought about by this synergistic effect. Conversely, when there is a conflicting relationship between two historical and cultural resources—for example, when they contradict each other in spatial layout, functional use, or environmental impact—their combined value may be negatively affected. Calculating the value loss component quantifies the value reduction caused by this negative effect. This mechanism makes the value correction component no longer a single, directionless adjustment, but rather a targeted and intensive correction based on the specific nature of the interaction relationship, thus making the overall value assessment result closer to reality.

[0057] In some implementations, step A5 includes: A501. Calculate the sum of the individual assessment values, the sum of the value gain components, and the sum of the value loss components; A502. The overall value assessment result is obtained by adding the sum of the individual assessment values ​​to the sum of the value gain components and then subtracting the sum of the value loss components.

[0058] Specifically, in step A501, the individual assessment values ​​of all historical and cultural resources within the target area are first summed to obtain a total of these individual assessment values. These individual assessment values ​​are calculated independently for each historical and cultural resource in step A4 above. Simultaneously, all identified value gain components are summed to obtain a total of these value gain components; and all identified value loss components are summed to obtain a total of these value loss components. The value gain and value loss components are value correction components determined in step A4 above based on the interaction attributes (enhancing relationship or conflict relationship) between adjacent historical and cultural resources.

[0059] In step A502, the overall value assessment result is obtained by adding the sum of the individual assessment values ​​to the sum of the value gain components, and then subtracting the sum of the value loss components. This calculation method ensures that when assessing the overall value of the target area, not only the inherent value of each historical and cultural resource is considered, but also the positive (gain) and negative (loss) effects of their interactions are fully taken into account.

[0060] This application's solution achieves precise quantification of the overall value of a target area by explicitly adding and subtracting the sum of individual assessment values, the sum of value gain components, and the sum of value loss components. The sum of individual assessment values ​​reflects the cumulative effect of the independent value of all historical and cultural resources within the area; the sum of value gain components reflects the synergistic value-added effect generated by the enhanced relationships between historical and cultural resources—for example, two adjacent historical buildings, due to their complementary styles or functional synergy, jointly enhance the cultural appeal of the area; while the sum of value loss components reflects the value reduction caused by conflicting relationships between historical and cultural resources—for example, a historical site surrounded by incongruous modern buildings may lead to a decrease in its historical atmosphere and aesthetic value. This comprehensive calculation method can fully and objectively reflect the true overall value of historical and cultural resources in the context of urban renewal.

[0061] In some of the embodiments described above in this application, methods for assessing the value of historical and cultural resources have been proposed. However, in their implementation, traditional methods of dynamically sensing data and constructing relationships may fail to fully capture the deep, chain-like connections between historical and cultural resources driven by cultural experiences, potentially leading to an incomplete and inaccurate assessment of the overall value of the resources. For example, relying solely on geographical distance or simple attribute similarity to construct connections may overlook the cultural experience paths formed by visitors between different resources, and the unique value inherent in these paths. If these problems are not addressed, urban renewal decisions may fail to fully identify and utilize the potential cultural spillover effects and synergistic values ​​of historical and cultural resources, thereby affecting the cultural heritage and economic benefits of renewal projects.

[0062] In some implementations, the dynamic sensing data also includes visitor behavior trajectory data and cultural clue data; Before step A3, the following are also included: A6. Perform semantic processing on the static attribute data to extract the cultural themes and narrative roles of each historical and cultural resource; A7. Combining the visitor behavior trajectory data, the cultural clue data, and the cultural themes and narrative roles of the historical and cultural resources, perform sequence pattern mining and semantic matching to infer whether there are chain-like associations driven by cultural experiences among the historical and cultural resources, and determine the association strength of the chain-like associations.

[0063] Visitor behavior trajectory data refers to the movement path information of visitors between different historical and cultural resources. Specifically, multimodal sensor networks can be used to track visitor movement paths. For example, a multimodal sensor network can include Wi-Fi probes, Bluetooth beacons, and visual analysis modules. Wi-Fi probes can record the location and movement of visitors by detecting their Wi-Fi signals; Bluetooth beacons can emit low-power Bluetooth signals, and visitor devices can report their location information after receiving the signal; the visual analysis module can capture and analyze images of visitors' movement using cameras. Cultural cue data refers to information about markers related to cultural activities. Specifically, image recognition technology can be used to identify markers carried by visitors. For example, image recognition technology can identify specific badges worn by visitors, promotional materials held by them, or culturally themed clothing worn by them.

[0064] Step A6 refers to semantic processing of static attribute data to extract the cultural themes and narrative roles of each historical and cultural resource. Semantic processing transforms these discrete attribute data into information with greater cultural connotation and narrative appeal. Specifically, Natural Language Processing (NLP) techniques can be used to analyze the textual descriptions of historical and cultural resources, extracting keywords and thematic terms, and combining this with expert knowledge bases or ontology to construct cultural themes and narrative roles. A cultural theme refers to the core cultural connotation or narrative thread carried by a historical and cultural resource, such as "Silk Road culture." A narrative role refers to the functional role played by a historical and cultural resource in a specific cultural narrative or historical event, such as "starting point," "turning point," "end point," or "witness." The aim is to transform discrete static attribute information into structured, culturally meaningful semantic tags, providing richer context for subsequent correlation analysis.

[0065] Step A7 refers to combining visitor behavior trajectory data, cultural cue data, and the cultural themes and narrative roles of historical and cultural resources to perform sequence pattern mining and semantic matching. This infers whether there are chain-like associations driven by cultural experiences among historical and cultural resources and determines the strength of these associations. Sequence pattern mining involves discovering frequently occurring resource access sequences from visitor behavior trajectory data. Specifically, the Apriori algorithm or PrefixSpan algorithm can be used to identify common paths and sequences of visitor movement between different resources. Semantic matching involves associating visitors' cultural preferences with the cultural themes and narrative roles of resources. This can be achieved using methods such as ontology-based semantic matching, word vector similarity calculation, or knowledge graph reasoning. For example, if visitor behavior trajectory data shows that many visitors, after visiting an ancient building with a "historical heritage" theme, then go to a traditional handicraft workshop with a "non-heritage" theme, and cultural cue data shows that visitors carry items related to traditional culture, then it can be inferred that there is a chain-like association driven by cultural experiences between the ancient building and the handicraft workshop. The strength of chain association refers to the tightness of this cultural experience-driven association. It can be quantified based on factors such as the frequency of occurrence of sequence patterns, the similarity score of semantic matching, and the visitor dwell time. For example, the higher the frequency of occurrence, the higher the similarity score, and the longer the dwell time, the stronger the association. (For example, the association strength can be pre-assigned to different combinations of frequency ranges, similarity score ranges, and visitor dwell time ranges to form a lookup table. The corresponding association strength can be obtained by querying the lookup table based on the actual frequency of occurrence, similarity score, and dwell time.)

[0066] This application's solution, by introducing visitor behavior trajectory data and cultural clue data, and combining them with semantic processing of static attribute data, can more comprehensively capture the complex relationships between historical and cultural resources from two dimensions: the actual experience of visitors and the deep cultural connotations of the resources. Specifically, visitor behavior trajectory data reveals the spatial-temporal sequence relationships between resources, while cultural clue data reflects the degree of alignment between visitors' cultural interests and resource themes. By semantically processing static attribute data, cultural themes and narrative roles are extracted, providing high-level semantic information for understanding the cultural positioning of resources. Based on this, sequence pattern mining is used to identify typical visitor tour paths, and semantic matching is used to associate visitors' cultural preferences with the cultural themes of resources, thereby inferring chain-like associations driven by cultural experiences. This chain-like association transcends traditional geographical proximity or simple attribute similarity, revealing the synergistic role of resources in cultural narratives and experience flows, enabling the evaluation model to more accurately reflect the overall value of historical and cultural resources.

[0067] As a preferred embodiment, the solution of this application is specifically implemented as follows: In the assessment scenario of a city's historical and cultural district, the district contains multiple historical and cultural resources, such as ancient buildings A, traditional handicraft workshops B, and historical streets and alleys C.

[0068] First, a multimodal sensor network is deployed, including the installation of Wi-Fi probes, Bluetooth beacons, and visual analytics modules at key locations in the ancient building A, the handicraft workshop B, and the historical street C. When visitors enter the area, the Wi-Fi probes and Bluetooth beacons continuously track the signals from their devices, recording their movement paths and dwell times between A, B, and C, generating visitor behavior trajectory data. Simultaneously, the visual analytics module captures visitor images via cameras and uses image recognition technology to identify culturally relevant identifiers carried by visitors, such as "Intangible Cultural Heritage Festival" badges or "Ancient Building Guide Maps," generating cultural clue data.

[0069] Secondly, semantic processing was performed on the static attribute data of ancient building A, handicraft workshop B, and historical street C. For example, the static attributes of ancient building A include "Ming and Qing architecture" and "ancestral hall." After semantic processing, cultural themes such as "clan culture" and "architectural art" and the narrative role of "historical witness" were extracted. The static attributes of handicraft workshop B include "traditional wood carving" and "intangible cultural heritage." Cultural themes such as "intangible cultural heritage inheritance" and "craftsmanship spirit" and the narrative role of "skill inheritor" were extracted. The static attributes of historical street C include "bluestone pavement" and "commercial street." Cultural themes such as "urban culture" and "historical features" and the narrative role of "life scenes" were extracted.

[0070] Next, sequence pattern mining and semantic matching were conducted by combining visitor behavior trajectory data, cultural clue data, and the cultural themes and narrative roles of each resource. For example, sequence pattern mining revealed that 80% of visitors would visit handicraft workshop B after visiting ancient building A, and 60% of these visitors carried items related to traditional culture. Semantic matching revealed a high correlation in cultural connotation between the themes of "clan culture" and "architectural art" of ancient building A and the themes of "intangible cultural heritage" and "craftsmanship" of handicraft workshop B. Therefore, it was inferred that there is a chain-like association driven by cultural experience between ancient building A and handicraft workshop B, and based on visitor frequency, dwell time, and semantic matching degree, the association strength was determined to be 0.8.

[0071] These visitor behavior trajectory data, cultural clue data, semantically processed cultural themes and narrative roles, as well as the inferred chain connections and their strength, will be injected into the subsequent construction of topological connection models and feature aggregation of graph learning algorithms as richer dynamic perception data and connection information, so as to more accurately identify and quantify the interactive relationship attributes between historical and cultural resources, and then calculate the overall value assessment results.

[0072] Furthermore, step A2 may also include: A204. Inject the chain association as an edge type into the topological association model, and use the association strength of the chain association as the corresponding edge weight; In step A301, during the process of determining the attention weights in the graph attention network, a weight factor for the chain association is introduced to adjust the influence of the chain association on node feature aggregation. Following step A3, the following is also included: A8. Using a regression prediction model, predict the value spillover component corresponding to the chain association based on the aggregated node and edge features; In step A5, the process of calculating the overall value assessment result also includes accumulating the value spillover component.

[0073] In this approach, chain-like connections are injected into the topological association model as an edge type, and the strength of these connections is used as the corresponding edge weights. This means explicitly modeling culturally driven chain-like connections as an edge type within the topological association model and assigning its connection strength as a weight. Specifically, this can be achieved as follows: When constructing the topological association model, in addition to considering traditional connections such as geographical distance, new edges are added between corresponding historical and cultural resource nodes based on the chain-like connection information inferred in step A7. These new edges are labeled as "chain-like connections," and their weights are set to the chain-like connection strength determined in A7. Therefore, the topological association model can more precisely express the deep cultural connections between historical and cultural resources beyond geographical distance, providing richer and more semantic graph structure information for subsequent graph learning algorithms, thus enabling more accurate identification of interactions between resources.

[0074] In the process of determining attention weights in graph attention networks, a weight factor for chain-like associations is introduced to adjust their impact on node feature aggregation. This means that when the graph attention network performs node feature aggregation, an adjustable weight factor is introduced for edges of this specific type, chain-like associations. Specifically, this can be achieved by multiplying this weight factor into the attention calculation formula of the graph attention network when calculating the attention weights of edges involving chain-like associations. This weight factor can be a fixed value or a value that is dynamically adjusted according to the characteristics of chain-like associations. Therefore, when the model performs node feature aggregation, it can dynamically adjust the contribution of chain-like associations to node feature aggregation based on their strength and type. This ensures that chain-like associations driven by cultural experience are fully considered and weighted during feature learning, thus enabling the aggregated node features to more accurately reflect the complex interactions between resources.

[0075] The method of using a regression prediction model to predict the value spillover component corresponding to chain associations based on aggregated node and edge features refers to using a specialized regression prediction model after the graph learning algorithm has completed feature aggregation. This model takes the aggregated node and edge features (including chain association features) as input and predicts the additional value generated by the chain associations. Specifically, this can be achieved by constructing a multilayer perceptron (MLP) or support vector regression (SVR) model. Its input layer receives the node feature vectors aggregated by a graph attention network and the edge feature vectors of the chain associations (e.g., the strength and type of the chain associations). The output layer predicts a continuous numerical value representing the value spillover component brought about by the chain associations. This model can be trained using historical data to learn how to map the features of chain associations to their corresponding value spillovers. This directly quantifies the additional contribution of cultural experience-driven chain associations to the overall value, solving the problem of how to make this implicit value explicit.

[0076] The process of calculating the overall value assessment result also includes accumulating the value spillover component. This refers to adding the chain-related value spillover component predicted by the regression prediction model to the original individual assessment value and value correction component to obtain the final overall value assessment result. Specifically, this can be achieved as follows: In step A5, the original formula for calculating the overall value assessment result is the sum of individual assessment values ​​plus the sum of value gain components minus the sum of value loss components. Based on this, the value spillover component predicted in A8 is also added to this sum. For example, the final overall value assessment result = sum of individual assessment values ​​+ sum of value gain components - sum of value loss components + sum of value spillover components. This ensures that the additional value brought about by the chain-like connections driven by cultural experience can be fully incorporated into the final overall value assessment, making the assessment result more comprehensive, accurate, and able to more truly reflect the overall value of historical and cultural resources.

[0077] This application's solution explicitly injects the chain-like connections driven by cultural experiences and their strength into the topological connection model, and introduces weighting factors to adjust their influence on feature aggregation in the graph attention network. This allows the model to more comprehensively capture the deep, indirect interactions between historical and cultural resources. It is precisely this deep integration of chain-like connections that enables subsequent regression prediction models to accurately predict the value spillover components generated by these chain-like connections based on the aggregated node and edge features. By adding this value spillover component to the overall evaluation result, this application's solution can overcome the shortcomings of traditional evaluation methods in quantifying the added value effect brought about by continuous cultural experiences, thus enabling the evaluation results to more comprehensively and accurately reflect the overall value of historical and cultural resources.

[0078] refer to Figure 2 This application provides an intelligent assessment system for the value of historical and cultural resources in urban renewal, used to assess the overall value of a target area containing multiple historical and cultural resources; the system includes: Data acquisition module 1 is used to acquire static attribute data and dynamic sensing data of multiple historical and cultural resources within the target area (for details, please refer to step A1 above). Model building module 2 is used to build a topological association model based on the relationship between the historical and cultural resources; the topological association model uses the historical and cultural resources as nodes, and uses the static attribute data and the dynamic perception data as feature vectors of the nodes (for details, please refer to step A2 above). The aggregation processing module 3 is used to perform feature aggregation processing on the topological association model using graph learning algorithms to identify and quantify the interaction relationship attributes between adjacent nodes; the interaction relationship attributes include relationship type and relationship strength (the specific process can be referred to step A3 above). The component calculation module 4 is used to calculate the individual evaluation value of each of the historical and cultural resources, and determine the corresponding value correction component based on the interaction relationship attribute (for details, please refer to step A4 above). Result generation module 5 is used to combine the individual assessment values ​​and the value correction components to calculate the overall value assessment result of the target area (for details, please refer to step A5 above).

[0079] In some implementations, the dynamic sensing data also includes visitor behavior trajectory data and cultural clue data; The intelligent assessment system for the value of historical and cultural resources in this urban renewal project also includes: The semantic processing module is used to perform semantic processing on the static attribute data and extract the cultural themes and narrative roles of each historical and cultural resource (for details, please refer to step A6 above). The association determination module is used to combine the visitor behavior trajectory data, the cultural clue data, and the cultural themes and narrative roles of the historical and cultural resources to perform sequence pattern mining and semantic matching, infer whether there are chain associations driven by cultural experience among the historical and cultural resources, and determine the association strength of the chain associations (for details, please refer to step A7 above).

[0080] Furthermore, when constructing the topological association model, model building module 2 can also perform the following: The chain association is injected into the topological association model as an edge type, and the association strength of the chain association is used as the corresponding edge weight. When performing feature aggregation processing on the topological association model, the aggregation processing module 3 can also perform the following: During the process of the graph attention network learning the attention weights, a weight factor for the chain association is introduced to adjust the influence of the chain association on node feature aggregation. The intelligent assessment system for the value of historical and cultural resources in this urban renewal project also includes: The overflow prediction module is used to predict the value overflow component corresponding to the chain association based on the aggregated node features and edge features using a regression prediction model (the process can be referred to step A8 above). The result generation module 5 also adds the value spillover component during the process of calculating the overall value assessment result.

[0081] The above descriptions are merely some embodiments of the present invention. Those skilled in the art can make various modifications and improvements without departing from the inventive concept of the present invention, and these all fall within the scope of protection of the present invention.

Claims

1. A method for intelligently assessing the value of historical and cultural resources in urban renewal, used to evaluate the overall value of a target area containing multiple historical and cultural resources; characterized in that... The method includes the following steps: A1. Obtain static attribute data and dynamic sensing data of multiple historical and cultural resources within the target area; A2. Construct a topological association model based on the relationships between the historical and cultural resources; the topological association model uses the historical and cultural resources as nodes, and uses the static attribute data and the dynamic perception data as feature vectors of the nodes; A3. Use graph learning algorithms to perform feature aggregation processing on the topological association model to identify and quantify the interaction relationship attributes between adjacent nodes; the interaction relationship attributes include relationship type and relationship strength. A4. Calculate the individual evaluation value of each of the historical and cultural resources, and determine the corresponding value correction component based on the interaction relationship attributes; A5. By combining the individual assessment values ​​and the value correction components, the overall value assessment result of the target area is calculated.

2. The intelligent assessment method for the value of historical and cultural resources in urban renewal according to claim 1, characterized in that, The static attribute data includes the type, architectural style, and historical period of the historical and cultural resources; the dynamic sensing data includes the real-time pedestrian flow, ambient temperature and humidity, and noise level in the area where the historical and cultural resources are located.

3. The intelligent assessment method for the value of historical and cultural resources in urban renewal according to claim 1, characterized in that, Step A2 includes: A201. The historical and cultural resources are abstracted into nodes, and the static attribute data and the dynamic perception data are concatenated to obtain the feature vector of the node; the node is a graph node; A202. Calculate the distance between the geographic centers of any two of the aforementioned historical and cultural resources; A203. If the distance between the geographic centers is less than a preset threshold, an initial edge is established between the two corresponding nodes, and an initial weight is assigned to the initial edge based on the reciprocal of the distance between the geographic centers.

4. The intelligent assessment method for the value of historical and cultural resources in urban renewal according to claim 1, characterized in that, Step A3 includes: A301. The feature vectors of the nodes are transformed and aggregated using the layered attention layers of the graph attention network to determine the attention weights between adjacent nodes; A302. Based on the attention weights, a classifier is used to identify the relationship type between adjacent nodes, and a regressor is used to calculate the relationship strength between adjacent nodes.

5. The intelligent assessment method for the value of historical and cultural resources in urban renewal according to claim 1, characterized in that, In step A4, the individual assessment value of each of the historical and cultural resources is calculated, including: Obtain the historical rating, artistic value score, average visitor traffic, and average visitor dwell time for each of the aforementioned historical and cultural resources; The historical rating, the art value score, the average foot traffic, and the average visitor dwell time are normalized. For each historical and cultural resource, the normalized historical grade, normalized artistic value score, normalized average visitor flow, and normalized average visitor dwell time are weighted and summed to obtain the individual evaluation value of the historical and cultural resource.

6. The intelligent assessment method for the value of historical and cultural resources in urban renewal according to claim 1, characterized in that, In step A4, determining the corresponding value correction component based on the interaction relationship attribute includes: If the relationship type is an enhanced relationship, then the value gain component is determined as the value correction component based on the product of the relationship strength, the sum of the individual evaluation values ​​of the two corresponding nodes, and the preset gain coefficient. If the relationship type is a conflict relationship, then the value loss component is determined as the value correction component based on the product of the relationship strength, the sum of the individual evaluation values ​​of the two corresponding nodes, and the preset loss coefficient.

7. The intelligent assessment method for the value of historical and cultural resources in urban renewal according to claim 6, characterized in that, Step A5 includes: A501. Calculate the sum of the individual assessment values, the sum of the value gain components, and the sum of the value loss components; A502. The overall value assessment result is obtained by adding the sum of the individual assessment values ​​to the sum of the value gain components and then subtracting the sum of the value loss components.

8. The intelligent assessment method for the value of historical and cultural resources in urban renewal according to claim 4, characterized in that, The dynamic sensing data also includes visitor behavior trajectory data and cultural clue data; Before step A3, the following are also included: A6. Perform semantic processing on the static attribute data to extract the cultural themes and narrative roles of each historical and cultural resource; A7. Combining the visitor behavior trajectory data, the cultural clue data, and the cultural themes and narrative roles of the historical and cultural resources, perform sequence pattern mining and semantic matching to infer whether there are chain-like associations driven by cultural experiences among the historical and cultural resources, and determine the association strength of the chain-like associations.

9. The intelligent assessment method for the value of historical and cultural resources in urban renewal according to claim 8, characterized in that, Step A2 also includes: A204. Inject the chain association as an edge type into the topological association model, and use the association strength of the chain association as the corresponding edge weight; In step A301, during the process of determining the attention weights in the graph attention network, a weight factor for the chain association is introduced to adjust the influence of the chain association on node feature aggregation. Following step A3, the following is also included: A8. Using a regression prediction model, predict the value spillover component corresponding to the chain association based on the aggregated node and edge features; In step A5, the process of calculating the overall value assessment result also includes accumulating the value spillover component.

10. An intelligent assessment system for the value of historical and cultural resources in urban renewal, used to assess the overall value of a target area containing multiple historical and cultural resources; characterized in that, The system includes: The data acquisition module is used to acquire static attribute data and dynamic sensing data of multiple historical and cultural resources within the target area; The model building module is used to construct a topological association model based on the relationship between the historical and cultural resources; the topological association model uses the historical and cultural resources as nodes, and uses the static attribute data and the dynamic perception data as the feature vectors of the nodes. The aggregation processing module is used to perform feature aggregation processing on the topological association model using graph learning algorithms to identify and quantify the interaction relationship attributes between adjacent nodes; the interaction relationship attributes include relationship type and relationship strength. The component calculation module is used to calculate the individual evaluation value of each of the historical and cultural resources, and determine the corresponding value correction component based on the interaction relationship attribute. The result generation module is used to combine the individual assessment values ​​and the value correction components to calculate the overall value assessment result of the target area.

Images