Intangible cultural heritage element feature extraction method based on knowledge graph

Abstract

The application relates to the technical field of data processing, and provides a non-heritage element feature extraction method based on a knowledge graph. The method comprises the following steps: defining interface arrangement parameters; checking the interface arrangement parameters; developing an interface arrangement management system based on the interface arrangement parameters when the checking is passed; determining a system to be connected, integrating through the interface arrangement management system, and obtaining a multi-source non-heritage element set; initializing a non-heritage element knowledge graph, identifying the multi-source non-heritage element set, and obtaining key non-heritage elements; performing multi-scale feature extraction on the key non-heritage elements, and outputting multi-scale fusion non-heritage elements; fusing the multi-scale fusion non-heritage elements, and outputting non-heritage element extraction results, so as to solve the technical problems that in the existing non-heritage element feature extraction process, interface arrangement complexity is high, it is difficult to cope with the complexity and heterogeneity of multi-source data, non-heritage element feature recognition accuracy is low, and feature fusion effect is poor.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a method for extracting intangible cultural heritage elements based on knowledge graphs. Background Technology

[0002] Intangible cultural heritage (ICH) refers to various traditional cultural expressions passed down through generations and closely related to the lives of people of all ethnic groups, including oral traditions, performing arts, festivals, and traditional handicrafts. These cultural expressions are an important part of national culture and possess significant historical, cultural, and social value. However, with the acceleration of globalization and modernization, many ICH projects face difficulties in transmission and are on the verge of disappearing, urgently requiring protection and transmission through modern technology. Existing ICH protection methods mainly rely on manual collection and organization, lacking systematic and intelligent means to comprehensively and accurately identify and extract ICH elements. This method is not only inefficient but also prone to omissions and errors, failing to meet the needs of large-scale ICH protection and utilization. Furthermore, existing technologies often face challenges in processing multi-source heterogeneous data, such as difficulty in data integration, inaccurate element identification, and poor feature extraction, resulting in unsatisfactory transmission and promotion effects of ICH elements in different application scenarios. Summary of the Invention

[0003] This invention provides a knowledge graph-based method for extracting intangible cultural heritage (ICH) element features, aiming to solve the technical problems of high interface arrangement complexity, difficulty in dealing with the complexity and heterogeneity of multi-source data, resulting in low accuracy of ICH element feature recognition and poor feature fusion effect in the existing ICH element feature extraction process.

[0004] In view of the above problems, the present invention provides a method for extracting intangible cultural heritage elements based on knowledge graphs.

[0005] This invention provides a method for extracting intangible cultural heritage (ICH) element features based on a knowledge graph. The method includes: defining interface orchestration parameters, which include the number of interfaces, interface classification, and transmission method; verifying the interface orchestration parameters; if the verification passes, developing an interface orchestration management system based on the defined interface orchestration parameters; identifying multiple systems to be accessed, integrating the multiple systems to be accessed through the interface orchestration management system to obtain a multi-source ICH element set; initializing an ICH element knowledge graph, identifying the multi-source ICH element set based on the ICH element knowledge graph to obtain multiple key ICH elements; performing multi-scale element feature extraction on the multiple key ICH elements to output multi-scale fused ICH elements; and fusing the multi-scale fused ICH elements to output the ICH element extraction result.

[0006] One or more technical solutions provided in this invention have at least the following technical effects or advantages:

[0007] The aforementioned knowledge graph-based intangible cultural heritage (ICH) feature extraction method first defines interface-related parameters, including quantity, classification, and transmission method, to guide the orchestration of these interfaces. Then, these defined parameters are validated to ensure accuracy before an interface orchestration management system is constructed. This system integrates multiple systems requiring access and extracts diverse ICH element information from these systems, forming a multi-source ICH element set. Next, an ICH element knowledge graph is built to aid in the identification and organization of these elements. Using this knowledge graph, several key ICH elements are extracted from the multi-source elements. For these key elements, features are extracted at various scales to gain a more comprehensive understanding of each element's characteristics. These features at different scales are then fused to obtain a more comprehensive and richer representation of the ICH elements. Finally, these fused elements are integrated to form the ICH element extraction result. This process improves the accuracy and efficiency of ICH element feature extraction, enabling better protection, inheritance, and display of intangible cultural heritage.

[0008] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a flowchart illustrating a method for extracting intangible cultural heritage elements based on knowledge graphs in one embodiment.

[0011] Figure 2 This is a flowchart illustrating the similarity matrix construction process of a knowledge graph-based intangible cultural heritage element feature extraction method in one embodiment. Detailed Implementation

[0012] This invention provides a knowledge graph-based method for extracting intangible cultural heritage (ICH) element features, which addresses the technical problems of high interface complexity, difficulty in handling the complexity and heterogeneity of multi-source data, and resulting in low accuracy of ICH element feature recognition and poor feature fusion effects in existing ICH element feature extraction processes.

[0013] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0014] It should be noted that the terms “comprising” and “having”, and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to such process, method, product, or device.

[0015] Examples, such as Figure 1 As shown, this invention provides a method for extracting intangible cultural heritage elements based on knowledge graphs, the method comprising:

[0016] Define interface orchestration parameters, which include the number of interfaces, interface categories, and transmission methods.

[0017] In this embodiment of the invention, the requirements and functions of the entire system are first analyzed to preliminarily determine the number of systems to be integrated and the interaction requirements between these systems, thereby determining the total number of interfaces. The determination of the number of interfaces needs to consider the system's scalability and future integration needs, so that the interface orchestration can support seamless integration of multiple systems. Subsequently, the interfaces are classified according to the characteristics of each system, such as function, data type, and communication protocol. Interface classification includes data interfaces, service interfaces, and management interfaces. Each type of interface may involve different communication methods, data formats, and security requirements; therefore, classification helps to clarify the functional positioning and technical implementation path of the interfaces. Next, a suitable transmission method is selected for each type of interface. Transmission methods can include synchronous communication, asynchronous communication, message queues, RESTful APIs, SOAP, etc. When selecting a transmission method, the real-time performance, reliability, security, and system architecture requirements of data transmission need to be comprehensively considered. For example, for real-time data interfaces requiring fast response, a WebSocket-based transmission method can be selected; for interfaces with large data volumes and requiring guaranteed transmission reliability, a message queue method can be selected. After determining the interface classification and transmission method, multiple interfaces are generated, and a set of interface orchestration parameters suitable for the current system requirements are defined based on these interfaces, providing clear guidance for subsequent interface development and system integration.

[0018] The interface orchestration parameters are validated. If the validation passes, an interface orchestration management system is developed based on the defined interface orchestration parameters.

[0019] In one embodiment, after obtaining the interface orchestration parameters, a verification model is established for the defined interface orchestration parameters. This model includes a consistency verification channel and a security verification channel. By inputting the interface orchestration parameters into the pre-built verification model, consistency and security are verified through the two channels within the verification model, respectively. When the verification results of both channels are greater than or equal to preset thresholds, it indicates that the interface orchestration parameters are set reasonably and meet the system requirements. A verification pass instruction is then issued, allowing the next development phase to proceed. Upon receiving the verification pass instruction, the development of the interface orchestration management system begins. This interface orchestration management system is used to integrate systems such as regional e-commerce, smart tourism platforms, and regional intangible cultural heritage archive management. Specifically, the interface parameters are defined as Interface(I1,I2,…I… nThe interface orchestration method is defined as follows: I represents a single interface, n represents the number of interfaces, and Type represents the interface category. For each interface I, its specific function and behavior can be abstractly defined as I(A(c), A(p), Desr, Tans, Tag), where A(c) represents the abstract list of front-end request services, A(p) represents the abstract list of back-end response services, Desr represents different interface descriptions, Tans represents the transport connection method, and Tag represents the interface abstraction identifier. These definitions provide the foundation for interface orchestration and ensure the coordination between different interfaces. The interface orchestration method is defined as Method(Sim(I... i ,I j ),Mat(Max(Sim),I),Arr(Uni,Sec)), where Sim(I i ,I j The similarity of interfaces is calculated. Mat(Max(Sim), I) is used to obtain the interface with the highest similarity and perform matching. Arr(Uni, Sec) is used to check the consistency and security of interface orchestration. Sim(I) is used to calculate the similarity. i ,I j The similarity between different interfaces is calculated to determine matching interface pairs. Then, `Mat(Max(Sim),I)` is used for interface matching to ensure smooth data transmission. Subsequently, `Arr(Uni,Sec)` is used to perform consistency and security checks to ensure overall stability and security during the orchestration process. Once interface I passes the orchestration operation of the `Method` (i.e., through similarity calculation, interface matching, and consistency and security checks), the interface orchestration logic is completed. Afterwards, based on the defined interfaces and orchestration methods, the system architecture is designed, including interface management modules and data processing modules, to build an interface orchestration management system. This enables efficient identification and processing of multi-source intangible cultural heritage elements, ensuring high stability and security in complex application scenarios.

[0020] Furthermore, such as Figure 2 As shown, this invention provides a method for validation before using predefined interface orchestration parameters, including:

[0021] Perform pairwise similarity calculations on all interfaces to obtain a similarity set, wherein the similarity calculation includes interface function, input / output parameters, and data type; generate a similarity matrix based on the similarity set, wherein each element in the matrix represents the similarity value between two interfaces; based on the similarity matrix, obtain the first interface pair with the highest similarity, and perform matching based on the first interface pair.

[0022] Preferably, to more accurately verify interface orchestration parameters, the system terminal first determines the criteria for calculating interface similarity, including interface function, input / output parameters, and data type. Comparing the similarity of two interface functions determines whether they perform similar operations or achieve similar goals. Comparing the input / output parameters of two interfaces checks whether their types, quantities, and structures are consistent. Comparing the data types transmitted by two interfaces determines whether their data formats are compatible. Subsequently, all interfaces in the system are paired, and the similarity of each pair of interfaces is calculated using the corresponding similarity calculation method according to the above criteria. For interface functions, a set of functional features is first defined for each interface, such as the main operations of the interface (e.g., read, write, delete), the frequency and timeliness of the operations, etc., and the functional features are vectorized to form feature vectors for the two interfaces. For example, binary values ​​are used to represent the existence of a function (1 indicates existence, 0 indicates non-existence). Then, the Jaccard similarity coefficient or cosine similarity is used to calculate the functional similarity between the two interfaces to obtain the interface functional similarity. For input / output parameters, the input and output parameters are first listed. Common parameters include integers, strings, floating-point numbers, JSON objects, etc. Next, the input and output parameters of the two interfaces are compared to check if their types are consistent, and the similarity of parameter names is calculated using Levenshtein distance. Then, the similarity of the input and output parameters is averaged to obtain the input-output parameter similarity. For data types, the specific data types transmitted by each interface are first listed, such as JSON, XML, CSV, and text, and then the similarity of the data types between the two interfaces is calculated using Levenshtein distance to obtain the data type similarity. After obtaining the interface function similarity, input-output parameter similarity, and data type similarity, the three similarities are averaged to obtain the similarity between the two interfaces, which is then added to the similarity set. This process is repeated until all interface pairs have been calculated. Afterward, a similarity matrix is ​​constructed based on the obtained similarity set. This matrix is ​​a symmetric matrix where rows and columns represent interfaces in the system, and each element represents the similarity value between the interfaces in the corresponding row and column. Then, the off-diagonal element with the largest similarity value is found in the similarity matrix. This value represents the pair of interfaces with the highest similarity among all interface combinations. Based on the found maximum similarity value, the corresponding interface pair I is determined. i and I jThese interfaces are most similar in functionality, input / output parameters, and data types. Finally, based on the most similar interface pairs, they are initially matched together. This means that these interfaces can work collaboratively or interchangeably in system calls and processing. Through this process, the system can efficiently identify and match the most similar interface pairs, laying the foundation for the subsequent development of an interface orchestration and management system. This method not only improves interface compatibility but also optimizes the overall system performance and data processing capabilities.

[0023] Furthermore, this invention provides validation through predefined interface orchestration parameters, including:

[0024] An orchestration verification model is obtained, comprising a consistency verification channel and a security verification channel. The consistency verification channel performs data transmission verification on the defined interface orchestration parameters and outputs a consistency verification index. The security verification channel performs data security verification on the defined interface orchestration parameters and outputs a security verification index. If both the consistency verification index and the security verification index are greater than or equal to a preset threshold, a verification pass result is obtained. If either the consistency verification index or the security verification index is less than the preset threshold, an orchestration update instruction is obtained. The interface orchestration parameters are redefined according to the orchestration update instruction until a verification pass result is obtained.

[0025] Preferably, after the interface orchestration parameters are calculated based on similarity, a pre-built orchestration verification model is acquired and activated. This model includes a consistency verification channel and a security verification channel. The consistency verification channel verifies the consistency and reliability of data transmission between the interface orchestration parameters. This channel primarily verifies the integrity and efficiency of data transmission. The security verification channel evaluates the security of the interface orchestration parameters, ensuring that the encryption level and user authentication complexity meet requirements during data transmission. Once activated, the interface orchestration parameters are input into the model for verification. In the consistency verification channel, a consistency verification index is obtained by weighting data transmission integrity and efficiency, and then transmitted to the input layer. In the security verification channel, a security verification index is obtained by weighting encryption level and user authentication complexity, and then transmitted to the input layer, where it is output along with the consistency verification index. After the orchestration verification model outputs the consistency and security verification indices, they are compared with preset thresholds. If both indices are greater than or equal to the preset thresholds, the verification is successful, and a successful verification result is generated. Conversely, if either the consistency check metric or the security check metric falls below a preset threshold, an orchestration update instruction is generated, prompting the need to redefine and adjust the interface orchestration parameters. Based on this instruction, the number, category, transmission method, or other parameters of the interfaces are adjusted, orchestration is re-executed, and verification is performed again through the orchestration verification model until both check metric meets the preset threshold, resulting in a successful verification. This process ensures the reliability and security of the interface orchestration management system when integrating and calling multiple interfaces, providing a solid foundation for the system's stable operation.

[0026] Furthermore, the present invention provides a consistency verification channel and a security verification channel, including:

[0027] The consistency verification channel is used to verify the data transmission of the predefined interface orchestration parameters, obtain the data transmission integrity, data transmission efficiency, and transmission mode matching, and calculate the data transmission integrity, data transmission efficiency, and transmission mode matching through the first weighted network layer to output the consistency verification index. The security verification channel is used to verify the data security of the predefined interface orchestration parameters, obtain the data encryption level, user authentication complexity, and access control protection, and calculate the data encryption level, user authentication complexity, and access control protection through the second weighted network layer to output the security verification index.

[0028] Optionally, a consistency verification channel is used to verify various metrics of interface orchestration parameters in terms of data transmission, such as data transmission integrity, data transmission efficiency, and transmission mode matching, to ensure that data transmission in the system is reliable and efficient. To construct the consistency verification channel, a large amount of historical interface orchestration case data is first acquired, including both successful and failed cases. Each case contains the actual interface orchestration parameters and their corresponding results for transmission integrity, transmission efficiency, and transmission mode matching. Then, transmission integrity, transmission efficiency, and transmission mode matching are labeled for each case for supervised learning. Subsequently, based on a multilayer perceptron, the consistency verification channel structure is set up according to business requirements and data characteristics, including an input layer, a hidden layer, and an output layer. The historical interface orchestration case data is then divided to generate training, validation, and test sets, and the weights and biases of the consistency verification channel are initialized. Common initialization methods include random initialization, Gaussian distribution initialization, and Xavier initialization. Afterward, the training set is input into the consistency verification channel, passing through the input layer and hidden layer, and finally passed to the output layer to generate prediction results. The mean squared error (MSE) is then used to calculate the error between the predicted and true values. After calculating the error, the gradient of the mean squared error loss function with respect to the consistency check channel parameters is calculated using the backpropagation algorithm. These gradients indicate the direction and magnitude of adjustment needed for the consistency check channel parameters. Based on the calculated gradients, the model's weights and biases are updated using a selected optimizer (such as SGD or Adam), gradually reducing the loss function value. This training process is performed in multiple batches. After each batch, the consistency check channel is evaluated using a validation set, and the validation loss is calculated. This helps monitor the consistency check channel's performance on unseen data, avoiding overfitting or underfitting. If the validation set loss does not significantly decrease in several consecutive batches, training is stopped early, and the consistency check channel is finally evaluated using a test set, calculating the test set loss and various performance metrics (such as R-squared). 2Consistency verification channels (such as MAE) are used to ensure the model performs well on unseen data. Once verification is successful, the current consistency verification channel is saved. The first weighted network layer is constructed by assigning weights to various verification parameters based on the actual needs of the system. For example, data transmission integrity may be more important than transmission efficiency, thus giving it a higher weight. After constructing the consistency verification channel and the first weighted network layer, the output layer of the consistency verification channel is connected to the first weighted network layer. Using the weights set internally by the first weighted network layer, the data transmission integrity, data transmission efficiency, and transmission method matching generated by the consistency verification channel are weighted and summed to generate a consistency verification index. Similarly, using the same construction method, a security verification channel is constructed based on historical interface orchestration case data and its corresponding data encryption level, user authentication complexity, and access control protection results. This channel is connected to the second weighted network layer to evaluate the data security protection capabilities of the interface orchestration parameters, including data encryption level, user authentication complexity, and access control protection, and to generate a security verification index to ensure the confidentiality, integrity, and availability of data during transmission. By integrating the consistency verification channel and the security verification channel, an orchestration verification model is constructed. This model evaluates the overall effectiveness of interface orchestration parameters by using consistency and security verification metrics output from two verification channels, ensuring high reliability and data security in actual system operation.

[0029] Multiple systems to be connected are identified, and these systems are integrated through the interface orchestration and management system to obtain a multi-source intangible cultural heritage element set.

[0030] In one embodiment, various data sources related to intangible cultural heritage (ICH) elements are identified based on business needs. These data sources may include regional e-commerce systems, smart tourism platforms, and ICH archive management systems. The role and function of each system in the extraction and processing of ICH elements are then clarified. For example, a regional e-commerce system may provide sales data for ICH products, a smart tourism platform may provide tourist behavior data, and an ICH archive management system may provide specific information about ICH. Subsequently, the interfaces of each system to be connected are called through an interface orchestration management system to obtain ICH element data from these systems. It is ensured that the data can be transmitted according to the predetermined interface orchestration parameters. Afterward, the data obtained from each system is synchronized and integrated. During this process, duplicates and abnormal data are removed, and data from different systems is converted to the same data format. Then, the data from different systems is aligned according to timestamps to ensure data consistency and accuracy. Finally, the cleaned and aligned data is aggregated into a multi-source ICH element set. This set will contain ICH element information from all systems to be connected, providing a reliable data foundation for subsequent data analysis and applications.

[0031] Furthermore, the present invention provides multiple systems to be accessed, including:

[0032] The multiple systems to be connected include at least the regional e-commerce platform, the regional smart tourism platform, the regional media convergence platform, the regional intangible cultural heritage archive management system, the regional intangible cultural heritage ontology and regional knowledge graph system, the regional intangible cultural heritage and cultural tourism resource distribution GIS system, and the regional intangible cultural heritage virtual simulation interactive system.

[0033] Preferably, the multiple systems to be integrated include regional e-commerce, regional smart tourism platforms, regional integrated media platforms, regional intangible cultural heritage (ICH) archive management systems, regional ICH ontology and regional knowledge graph systems, regional ICH and cultural tourism resource distribution GIS systems, and regional ICH virtual simulation interactive systems. Among these, the regional e-commerce, regional smart tourism, and regional integrated media platforms are existing systems, while the regional ICH archive management systems, regional ICH ontology and regional knowledge graph systems, regional ICH and cultural tourism resource distribution GIS systems, and regional ICH virtual simulation interactive systems are newly developed systems. The regional ICH archive management system is developed based on MongoDB and stores and manages data according to ICH categories. The regional ICH ontology and regional knowledge graph system is also developed based on MongoDB, ensuring that the system can provide accurate ICH element identification and knowledge association. The regional ICH and cultural tourism resource distribution GIS system is developed using GIS technology and MongoDB to realize geographic information management and display of resources. The ICH virtual simulation interactive system uses 3D and VR technologies for virtual simulation and interactive design of ICH, enhancing the user experience.

[0034] Initialize the intangible cultural heritage element knowledge graph, identify the multi-source intangible cultural heritage element set based on the intangible cultural heritage element knowledge graph, and obtain multiple key intangible cultural heritage elements.

[0035] In one embodiment, the nodes and relationships in the intangible cultural heritage (ICH) element structure sub-graphs are first determined. Nodes represent different ICH elements, and relationships represent the connections between these elements. Then, based on the data types of the interface orchestration management system, multiple ICH element structure sub-graphs are constructed. Subsequently, node information is merged based on the similarity of node attributes between different ICH element structure sub-graphs to generate an ICH element knowledge graph. Afterward, multi-source ICH element sets are matched within the ICH element knowledge graph, and the matching results are filtered and labeled according to a preset matching degree to generate multiple key ICH elements for subsequent feature extraction, analysis, and application.

[0036] Furthermore, this invention provides an initial knowledge graph of intangible cultural heritage elements, including:

[0037] Define the node attribute information and node dependency relationships for each intangible cultural heritage element structural sub-graph; integrate the data types of the interface orchestration management system, and obtain multiple intangible cultural heritage element structural sub-graphs according to the node attribute information and node dependency relationships, wherein each intangible cultural heritage element structural sub-graph corresponds to a data type, and each node of each intangible cultural heritage element structural sub-graph corresponds to an intangible cultural heritage element; calculate the node attribute similarity between different intangible cultural heritage element structural sub-graphs, merge the information of nodes with a similarity greater than a preset value, and fuse the multiple intangible cultural heritage element structural sub-graphs to generate an intangible cultural heritage element knowledge graph.

[0038] Preferably, attribute information is defined for nodes in each intangible cultural heritage element structure sub-graph based on different data types, such as video, image, text, and audio. For example, video node attributes may include the video's name, length, content description, related figures or events, etc. Image node attributes may include the image's shape, color, cultural significance, historical background, etc. Text node attributes may include the text's content, language, source, and the event or story it describes, etc. Audio node attributes may include the audio's title, length, sound quality, content description, and sound source (e.g., the sound of a dragon). Dependency relationships are then defined for nodes in each sub-graph. For example, the relationship between an image and text or historical event, such as a dragon image and a dragon story. The association between text and audio, such as a story describing a dragon and an audio recording of a dragon's roar. Subsequently, the various types of data obtained from different data sources by the interface orchestration management system are integrated. Data type consistency is ensured, and these are mapped to corresponding node attributes and dependency relationships, including names and multiple intangible cultural heritage element structure sub-graphs. Each intangible cultural heritage element structure sub-graph corresponds to a data type, and each node in the intangible cultural heritage element structure sub-graph corresponds to an intangible cultural heritage element. Taking video and text as examples, sub-graphs describing specific skills or historical events are generated based on video data. Each node represents an element in the video, such as a skill action or event scene. Sub-graphs describing stories or historical events are generated based on text data, with each node representing text content, such as the story of the dragon. Then, similarity calculation standards are defined for nodes in different sub-graphs, such as based on node name, description, historical background, cultural significance, etc., and similar similarity calculation methods are used to calculate similarity. For example, node names are extracted from the dragon image and the text describing the dragon story, and Levenshtein distance is used for name similarity. Feature vectors of the dragon image and the text describing the dragon story are obtained, and cosine similarity is used to calculate cultural similarity. The two similarities are then averaged to obtain node attribute similarity. After calculating the node attribute similarity for every two nodes, these node attribute similarities are compared with a preset similarity, and the information of nodes with node attribute similarities greater than the preset similarity is merged. For example, a dragon image, a dragon story text, and an audio recording of a dragon's roar can be merged into a single node, forming a comprehensive node with multiple meanings. By merging similar nodes from multiple intangible cultural heritage element structural sub-graphs, a complete intangible cultural heritage element knowledge graph can be generated. Each node in this graph may contain information from multiple data sources such as video, images, text, and audio, forming a node with multiple meanings, thus achieving the effect of processing multi-source data simultaneously.

[0039] Furthermore, this invention provides a method for identifying the multi-source intangible cultural heritage element set based on the intangible cultural heritage element knowledge graph, and obtaining multiple key intangible cultural heritage elements, including:

[0040] Based on the intangible cultural heritage element knowledge graph, node matching is performed on the multi-source intangible cultural heritage element set to obtain the data feature vector matching degree of each node; according to the size of the data feature vector matching degree, k nodes with a matching degree greater than a preset vector are selected, the k nodes are marked on the intangible cultural heritage element knowledge graph, and the intangible cultural heritage elements corresponding to the k nodes are extracted to obtain multiple key intangible cultural heritage elements.

[0041] Preferably, for each element in the multi-source intangible cultural heritage element set, its corresponding data feature vector is extracted. These feature vectors are obtained from data forms such as image features (extracted through existing convolutional neural networks), text features (extracted through existing embedding models), and audio features. Then, each element in the multi-source intangible cultural heritage element set is matched with nodes in the intangible cultural heritage element knowledge graph. Cosine similarity is used to calculate the matching degree between the feature vector of each element in the multi-source intangible cultural heritage element set and the feature vector of each node in the graph, obtaining the data feature vector matching degree of each node. Subsequently, according to business requirements, a preset vector matching degree is set, and this preset vector matching degree is used to filter the data feature vector matching degree of each node, obtaining k nodes whose data feature vector matching degree is greater than the preset vector matching degree. Afterwards, the k selected nodes are marked on the intangible cultural heritage element knowledge graph. Marking methods include changing node color, thickening node borders, adding annotations, etc., to make them stand out in the graph for easy identification. Then, the intangible cultural heritage elements corresponding to the K nodes are input into a pre-built user testing and evaluation model. This model evaluates these elements in terms of usability, satisfaction, and accuracy. The evaluation results are then weighted and summed to obtain k evaluation indicators. Finally, based on these pre-defined evaluation indicators, the k indicators are filtered to obtain several key intangible cultural heritage elements. These key elements will serve as core data for subsequent analysis, display, or application, representing the most representative cultural elements within the multi-source intangible cultural heritage element set, providing a solid data foundation for cultural protection, inheritance, and promotion.

[0042] Furthermore, the present invention provides a method for extracting the intangible cultural heritage elements corresponding to the k nodes and obtaining multiple key intangible cultural heritage elements, and also includes:

[0043] Establish a user testing and evaluation model; input the intangible cultural heritage elements corresponding to the k nodes into the user testing and evaluation model, including usability evaluation, satisfaction evaluation, accuracy evaluation, dissemination evaluation, and aesthetic evaluation, and obtain k evaluation indicators corresponding to the k intangible cultural heritage elements; output the intangible cultural heritage elements that exceed the preset evaluation indicators as the multiple key intangible cultural heritage elements.

[0044] Optionally, a large dataset of intangible cultural heritage (ICH) elements with known evaluation results is first collected. Each ICH element has a corresponding score on evaluation dimensions such as usability, satisfaction, accuracy, dissemination, and aesthetics. Usability assesses the ease of use of the ICH element, such as ease of operation and user experience. Satisfaction assesses the overall user satisfaction with the ICH element. Accuracy assesses the degree of matching between the ICH element and its actual cultural background or history. Dissemination assesses the dissemination effect of the ICH element in social media or educational scenarios. Aesthetics assesses the visual or auditory beauty of the ICH element. Subsequently, using the same method as described above for constructing consistency verification channels, a user testing and evaluation model is trained based on a multilayer perceptron, a third-weighted network layer, and the collected ICH element dataset. After training, the ICH elements corresponding to the selected k nodes are input into the trained user testing and evaluation model. The model outputs scores for each ICH element across five dimensions—usability, satisfaction, accuracy, dissemination, and aesthetics—based on the learned mapping relationships. These scores are then weighted and summed through the third-weighted network layer to obtain k evaluation indicators for the k ICH elements. Then, by comparing the k evaluation indicators with the preset evaluation indicators, the intangible cultural heritage elements with evaluation indicators greater than the preset evaluation indicators were selected, and the selected intangible cultural heritage elements were selected as multiple key intangible cultural heritage elements.

[0045] Multi-scale element feature extraction is performed on the multiple key intangible cultural heritage elements to output multi-scale fused intangible cultural heritage elements; the multi-scale fused intangible cultural heritage elements are then fused to output the intangible cultural heritage element extraction results.

[0046] In one embodiment, after obtaining multiple key intangible cultural heritage (ICH) elements, multi-scale element features are applied to these elements, including local and global scale features for each key ICH element. Local scale features extract local detail features of the ICH elements, such as texture, detailed structure, and features of specific parts, as well as volumetric structure. Global scale features extract overall features of the ICH elements, focusing on macroscopic information such as shape, outline, and overall structure. Subsequently, the extracted local and global features of each key ICH element are integrated to generate local and global feature vectors for each key ICH element, thus obtaining multi-scale fused ICH elements. Then, the local and global features of each key ICH element in the multi-scale fused ICH elements are fused and spliced ​​to generate the final ICH element extraction result. This ICH element extraction result contains key information of multiple ICH elements at different scales and has strong expressive power.

[0047] Furthermore, the present invention provides multi-scale element feature extraction for the aforementioned multiple key intangible cultural heritage elements, and also includes:

[0048] Local feature extraction is performed on the multiple key intangible cultural heritage elements to obtain the local feature vector corresponding to each key intangible cultural heritage element; global feature extraction is performed on the multiple key intangible cultural heritage elements to obtain the global feature vector corresponding to each key intangible cultural heritage element; the local feature vector and the global feature vector corresponding to each key intangible cultural heritage element are concatenated to output a comprehensive feature vector; the intangible cultural heritage element extraction result is output based on the comprehensive feature vector.

[0049] Optionally, for data from multiple key intangible cultural heritage elements, local feature extraction can be performed using image, text, and audio data as examples. For image data, key points in the image are located using feature point detection algorithms such as SIFT, SURF, and ORB. These key points represent important local regions in the image, such as edges and corners. The detected feature points are then described to generate local image feature vectors. Descriptors such as SIFT or SURF can capture the texture and shape information of local image regions. For text data, natural language processing techniques are used to extract key phrases or words, representing the core content of the text. These phrases or words are then input into existing text embedding models, such as Word2Vec and BERT, to generate corresponding local text feature vectors. For audio data, the audio signal is segmented into short time segments, and local spectral features are extracted using short-time Fourier transform or wavelet transform. The spectral features extracted from each time segment are then encoded to generate local audio feature vectors, which reflect the changes in audio over a specific time period. Subsequently, the local feature vectors of the image, text, and audio are concatenated to form a complete local feature vector, which contains information about the intangible cultural heritage elements at the local scale. For each key intangible cultural heritage element, its corresponding local feature vector is output for subsequent feature concatenation. Similarly, global feature extraction is performed on the image, text, and audio data of multiple key intangible cultural heritage elements. For image data, existing convolutional neural networks, such as ResNet and VGG, are used to extract global features. The global feature vector of the image is then output through the fully connected layer or the last convolutional layer of the network. For text data, a text embedding model is used to generate a global text feature vector representing the overall semantics of the text. For audio data, Mel-frequency cepstral coefficients (MFCC) are used to extract global audio features, generating an audio global feature vector. MFCC captures the overall frequency distribution and pitch characteristics of the audio. Finally, the extracted global feature vectors of the image, text, and audio are concatenated to obtain a complete global feature vector reflecting the information of the intangible cultural heritage elements at the overall level. For each key intangible cultural heritage element, its corresponding global feature vector is output as the basis for comprehensive feature concatenation. Then, the local and global feature vectors of each key intangible cultural heritage element are concatenated. The concatenation method is direct concatenation, linking the local and global feature vectors into a longer comprehensive feature vector. For each key intangible cultural heritage element again, its corresponding comprehensive feature vector is output. This vector combines local and global feature information, providing the most comprehensive description of the intangible cultural heritage element. Finally, all comprehensive feature vectors are summarized to generate the final intangible cultural heritage element extraction results. These results combine the local and global features of intangible cultural heritage elements, comprehensively reflecting their cultural connotations and forms of expression.

[0050] In summary, the embodiments of the present invention have at least the following technical effects:

[0051] This invention first defines interface orchestration parameters and develops an interface orchestration management system after verification. This system integrates multiple systems to be connected, acquires a multi-source intangible cultural heritage (ICH) element set, and identifies the multi-source ICH elements based on an initialized ICH element knowledge graph, extracting key ICH elements. Further, multi-scale element feature extraction is performed on these key ICH elements, and the extracted ICH elements are output after fusion. In addition, the method includes calculating interface similarity, using an orchestration verification model to verify the consistency and security of interface parameters, and redefining interface orchestration parameters as needed. The initialization of the ICH element knowledge graph is performed by integrating data types and calculating node attribute similarity. Finally, the key ICH elements are evaluated using a user testing and evaluation model, and a comprehensive feature vector is output based on the concatenation of local and global features, yielding the extracted ICH elements. These technologies collectively address the technical problems in existing intangible cultural heritage (ICH) element feature extraction processes, such as high interface orchestration complexity, difficulty in handling the complexity and heterogeneity of multi-source data, resulting in low accuracy of ICH element feature recognition and poor feature fusion effects. They achieve the effect of optimizing the feature fusion process and improving the accuracy and efficiency of ICH element feature extraction by defining interface orchestration parameters and optimizing the interface orchestration management system.

[0052] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0053] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0054] This specification and accompanying drawings are merely illustrative examples of the invention and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of the invention. Clearly, those skilled in the art can make various alterations and modifications to the invention without departing from its scope. Therefore, if such modifications and modifications fall within the scope of the invention and its equivalents, the invention is intended to include these modifications and modifications.

Claims

1. A method for extracting intangible cultural heritage element features based on a knowledge graph, characterized in that, The method comprises: defining interface arrangement parameters, the interface arrangement parameters comprising the number of interfaces, interface classification, and transmission mode; verifying through the defined interface arrangement parameters, and if the verification is passed, developing an interface arrangement management system based on the defined interface arrangement parameters; determining a plurality of systems to be accessed, integrating the plurality of systems to be accessed through the interface arrangement management system, and obtaining a multi-source intangible cultural heritage element set; initializing an intangible cultural heritage element knowledge graph, identifying the multi-source intangible cultural heritage element set according to the intangible cultural heritage element knowledge graph, and obtaining a plurality of key intangible cultural heritage elements; performing multi-scale element feature extraction on the plurality of key intangible cultural heritage elements, and outputting multi-scale fusion intangible cultural heritage elements; fusing the multi-scale fusion intangible cultural heritage elements, and outputting intangible cultural heritage element extraction results; The method for initializing the intangible cultural heritage element knowledge graph comprises: defining node attribute information and node dependency relationship of each intangible cultural heritage element structure sub-graph; integrating data types of the interface arrangement management system, obtaining a plurality of intangible cultural heritage element structure sub-graphs according to the node attribute information and the node dependency relationship, wherein each intangible cultural heritage element structure sub-graph corresponds to a data type, and the nodes of each intangible cultural heritage element structure sub-graph correspond to an intangible cultural heritage element; performing node attribute similarity calculation between different intangible cultural heritage element structure sub-graphs, merging nodes with a similarity greater than a preset similarity, fusing the plurality of intangible cultural heritage element structure sub-graphs, and generating an intangible cultural heritage element knowledge graph; performing multi-scale element feature extraction on the plurality of key intangible cultural heritage elements, outputting multi-scale fusion intangible cultural heritage elements, and fusing the multi-scale fusion intangible cultural heritage elements to output intangible cultural heritage element extraction results, the method further comprising: performing local feature extraction on the plurality of key intangible cultural heritage elements to obtain a local feature vector corresponding to each key intangible cultural heritage element; performing global feature extraction on the plurality of key intangible cultural heritage elements to obtain a global feature vector corresponding to each key intangible cultural heritage element; splicing the local feature vector and the global feature vector corresponding to each key intangible cultural heritage element to output a comprehensive feature vector corresponding to each key intangible cultural heritage element; according to the comprehensive feature vector, all comprehensive feature vectors are summarized to output intangible cultural heritage element extraction results.

2. The method of claim 1, wherein, Before verifying through the defined interface arrangement parameters, the method comprises: performing pairwise similarity calculation on all interfaces to obtain a similarity set, wherein the similarity calculation comprises interface function, input and output parameters, and data type; generating a similarity matrix according to the similarity set, wherein each element in the matrix represents the similarity value between two interfaces; based on the similarity matrix, obtaining a first interface pair with the maximum similarity, and matching according to the first interface pair.

3. The method of claim 1, wherein, The method for verifying through the defined interface arrangement parameters comprises: obtaining an arrangement verification model, the arrangement verification model comprising a consistency verification channel and a security verification channel, wherein the consistency verification channel is used for data transmission verification of the defined interface arrangement parameters, and outputs a consistency verification index, and the security verification channel is used for data security verification of the defined interface arrangement parameters, and outputs a security verification index; If the consistency check index and the security check index are greater than or equal to a preset threshold, a check pass result is obtained; If any of the consistency check index and the security check index is less than the preset threshold, an arrangement update instruction is obtained, and interface arrangement parameters are redefined according to the arrangement update instruction until a check pass result is obtained.

4. The method of claim 3, wherein, The consistency check channel is used for data transmission check of the defined interface arrangement parameters, and data transmission integrity, data transmission efficiency and transmission mode matching are obtained. The data transmission integrity, data transmission efficiency and transmission mode matching are calculated through a first weight network layer, and a consistency check index is output. The security check channel is used for data security check of the defined interface arrangement parameters, and data encryption level, user authentication complexity and permission control protection are obtained. The data encryption level, user authentication complexity and permission control protection are calculated through a second weight network layer, and a security check index is output.

5. The method of claim 1, wherein, According to the intangible element knowledge graph, the multi-source intangible element set is identified, and a plurality of key intangible elements are obtained. The method comprises: Based on the intangible element knowledge graph, node matching is performed on the multi-source intangible element set, and data feature vector matching degrees of each node are obtained. According to the size of the data feature vector matching degrees, k nodes greater than a preset vector matching degree are screened, the k nodes are identified on the intangible element knowledge graph, the intangible elements corresponding to the k nodes are extracted, and a plurality of key intangible elements are obtained.

6. The method of claim 5, wherein, The method for extracting the intangible elements corresponding to the k nodes to obtain a plurality of key intangible elements further comprises: Establishing a user test evaluation model; The intangible elements corresponding to the k nodes are input into the user test evaluation model, including usability evaluation, satisfaction evaluation, accuracy evaluation, propagation evaluation and aesthetic evaluation, and k evaluation indexes corresponding to the k intangible elements are obtained. Intangible elements greater than a preset evaluation index are taken as the plurality of key intangible elements.

7. The method of claim 1, wherein, The plurality of systems to be accessed at least include regional e-commerce, regional smart tourism platform and regional fusion media platform, regional intangible cultural heritage archive management system, regional intangible cultural heritage ontology and regional knowledge graph system, regional intangible cultural heritage and tourism resource distribution GIS system and regional intangible cultural heritage virtual simulation interaction system.

Images