Construction method and application of culture resource big data reference model

Abstract

The invention discloses a culture resource big data reference model construction method, and relates to the technical field related to culture big data. Comprising the following steps: setting a multi-source data integration module, setting a real-time data perception module, creating a data fusion engine, setting a user feedback analysis module, constructing a cultural heritage damage prediction calculation module, creating an emotion analysis feedback system, and setting a cultural resource digital display platform. The invention further discloses application of the culture resource big data reference model, specifically, the culture resource big data reference model is applied to an online and offline system, and a framework of public culture digital resources should grasp effective online and offline combination. By analyzing emotional feedback of visitors, not only can cultural understanding and resonance be enhanced and inheritance and innovation of cultural heritage be improved, but also cultural education and experience can be optimized, cultural policies and market decisions can be supported, cultural diversity and inclusiveness can be promoted, and development of the cultural industry can be promoted.

Description

Technical Field

[0001] The present invention relates to the technical field related to cultural big data, and in particular to a method for constructing a cultural resource big data reference model and its application. Background Art

[0002] The significance of building a big data reference model for cultural resources is far-reaching. It can not only better protect and inherit cultural heritage, but also promote the development and innovation of the cultural industry. Through in-depth analysis of cultural resources, we can dig out more about the historical value of cultural heritage and the stories behind it, and provide rich materials for the inheritance of traditional culture. With the continuous development of information technology, the application scope of cultural resources has rapidly radiated to various fields. Building a big data model for cultural resources to strengthen digital construction has become an inevitable trend. With the development of science and technology and the advancement of big data technology, the big data reference model for cultural resources will play an increasingly important role in the cultural field and make greater contributions to the protection of human cultural heritage and the prosperity of the cultural industry.

[0003] At present, the cultural resource big data reference model lacks feedback on visitors' emotions and reference in improving exhibition layout and interactive design, which reduces the personalized and emotional visiting experience, makes it difficult to optimize cultural education and experience, and promote the development of the cultural industry. On the other hand, it has poor prediction effect on cultural heritage damage, making it difficult to provide technical support for the protection, restoration and management of cultural heritage and extend the life of cultural heritage.

[0004] Therefore, it is necessary to propose a construction method and application of a cultural resource big data reference model to solve the above problems. Summary of the invention

[0005] 1. Technical issues to be resolved

[0006] In view of the deficiencies in the prior art, the present invention provides a method for constructing and applying a cultural resource big data reference model, which solves the problems that the existing cultural resource big data reference model lacks reference in improving exhibition layout and interactive design, reduces the personalized and emotional visiting experience, and is difficult to promote the development of the cultural industry; on the other hand, it is difficult to provide technical support for the protection, restoration and management of cultural heritage, and is difficult to extend the life of cultural heritage.

[0007] (II) Technical solution

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0009] A method for constructing a cultural resource big data reference model includes the following steps:

[0010] S1: Set up a multi-source data integration module, which is responsible for collecting data from many data sources, including museums, libraries, cultural heritage monitoring, social media, and web crawlers. The data sources include text, pictures, videos, and 3D scans in different formats;

[0011] S2: Establish a real-time data perception module to use IoT devices to monitor cultural relics in real time, collect environmental data and visitor flow information, and transmit them to the data fusion engine in real time;

[0012] S3: Create a data fusion engine to align and fuse data from different sources and formats in time and space, creating a multi-dimensional cultural resource database to facilitate subsequent analysis and utilization;

[0013] S4: Set up a user feedback analysis module to collect user feedback information for improving system functions and service quality and achieving continuous optimization of the system;

[0014] S5: Construct a cultural heritage damage prediction and calculation module, analyze and process the data in the multi-source data integration module and the real-time data perception module, use algorithms to analyze the damage patterns and risk factors of cultural heritage, realize early warning and formulate protection measures, and digitally restore and enhance damaged or aging cultural resources (such as paintings and documents) using image recognition and processing technology, and enhance their quality through AI for research and display;

[0015] S6: Create a sentiment analysis feedback system to analyze user comments on social platforms and cultural resource platforms to gain insights into user preferences and optimize recommendation algorithms accordingly. Identify audience emotions by analyzing language in social media, on-site feedback, or online interactions. Use sentiment analysis technology to evaluate audience emotional responses to specific cultural exhibits or activities, which is used to improve exhibition design and event planning.

[0016] S7: Set up a digital display platform for cultural resources, use AR / VR technology to provide audiences with an immersive cultural resource display experience, including virtual museum tours and historical scene reproduction, and build a multimodal interactive interface. Design an easy-to-use user interface that supports multiple interactive methods (voice, touch, and somatosensory) to lower the threshold for use and increase user participation;

[0017] S8: Configure a central processing server to process information from the multi-source data integration module, real-time data perception module, user feedback analysis module and sentiment analysis feedback system, and display the information using the cultural resources digital display platform.

[0018] Optionally, the cultural resource big data reference model also provides a multilingual voice guide system, develops a multilingual intelligent voice guide system based on speech recognition and natural language processing technology, provides convenience for tourists from different countries and regions, and provides an intelligent question and answer system, in which users ask questions in natural language and the system gives answers, providing information about the historical, artistic and scientific value of cultural resources.

[0019] Optionally, the cultural resource big data reference model also sets up a data security protection system to implement data encryption, access control and privacy protection measures to ensure the security of cultural resource data.

[0020] Optionally, when collecting data in S1, the RGB color collection mode is implemented for the color of cultural resources. The specific formula is: C=R+G+B, C represents a certain color, wherein R, G, and B represent the intensities of red, green, and blue, respectively, and the intensity value range of each color is 0 to 255. When R=255 (maximum brightness of red), G=0 (no green), and B=0 (no blue), the color presented is pure red. When the brightness of the three primary colors is 255, the mixed color is white. When the brightness of the three primary colors is 0, black is displayed.

[0021] Optionally, the cultural resource database in S3 adopts a distributed database system, including data representation, data sharding, query processing, and data synchronization and consistency maintenance of cultural resources. Cultural resources include historical relics, artworks, and documentary materials. These resources include multiple data types such as text, images, audio, and video. The data representation model is as follows:

[0022] R = {r1, r2, ..., r n}

[0023] Among them, R represents the collection of all cultural resources, r n Represents the nth resource item, each of which contains multiple data types and attributes;

[0024] Sharding is performed based on the geographical location, type, and age attributes of resources. The data sharding model formula is:

[0025] S = {s1, s2, ..., s m}

[0026] Among them, S is the set of all data shards, s m is the mth shard, each shard contains a set of related cultural resources;

[0027] Query processing includes query decomposition, localization processing, and result aggregation. For cultural resource queries, a cost-based model is used to optimize the query plan as follows:

[0028] Cost(Q)=Cost I / O (Q)+Cost CPU (Q)+Cost Network (Q)

[0029] Where Cost(Q) is the total cost of querying Q, including I / O cost, CPU cost, and network communication cost;

[0030] Cultural resource data involves synchronization of multiple data centers and integration with external data sources. Consistency maintenance is achieved through data replication and distributed transaction management. The specific model is expressed as follows:

[0031] Consistency=Replication Factor×Update Protocol

[0032] Among them, Replication Factor is the data replication factor, and Update Protocol is the update protocol, such as the two-phase commit protocol (2PC).

[0033] Optionally, the cultural heritage damage prediction calculation module in S5 mainly involves the target detection algorithm Y0L0v5, the deep learning model of text recognition and image processing, the Y0L0v5 target detection algorithm is used to detect and identify the damage type in the image in real time, the algorithm divides the input image into multiple regions, predicts the bounding box and category probability of each region, and the mathematical model formula is:

[0034]

[0035] Among them, P(Object) represents the probability of having an object, C i is the score predicted by the model, and the positioning loss is calculated by the intersection over union (IoU) loss:

[0036] Loss IoU =1-IoU

[0037] The total loss function is:

[0038] Loss total =Loss coord +Loss obj +Loss cls +Loss IoU

[0039] Large models such as BERT and GPT-3 play an important role in text recognition and image processing of historical documents by extracting and analyzing features from scanned images;

[0040] The text recognition model formula is:

[0041]

[0042] Among them, P(w | x) represents the probability of predicting text w given an image x, f(x) is the feature representation of the image, and v w is the vector representation of text w, V is the vocabulary;

[0043] The image processing model formula is:

[0044] G(x)=σ(W·f(x)+b)

[0045] Among them, G(x) represents the processed image, W and b are weights and biases, and σ is the activation function;

[0046] The application of generative adversarial network (GAN) in cultural heritage restoration, the mathematical model formula is:

[0047] G θ (z) = t, D φ (t) = 1-D φ (G θ (z))

[0048] Among them, G θ represents the generator, D φ Represents the discriminator, z is random noise, t is real data, the generator generates data close to the real data, and the discriminator tries to distinguish the generated data from the real data. The two are continuously optimized through the adversarial process.

[0049] Optionally, the sentiment analysis feedback system in S5 is used to identify and analyze the sentiment tendency in the text data. In sentiment analysis, the text is converted into a word frequency vector, and the probability that the text belongs to different sentiment categories is calculated. The mathematical model formula is:

[0050]

[0051] Where P(c | x) is the probability that the text belongs to category c given the feature vector x, P(x | c) is the probability of feature vector x given category c, P(c) is the prior probability of category c, and P(x) is the probability of feature vector x;

[0052] The contextual information in the text is captured by the self-attention mechanism. The mathematical model formula is:

[0053]

[0054] Among them, Q, K, and V represent query vector, keyword vector, and value vector respectively. k represents the dimension of the keyword vector;

[0055] MultiHead(Q,K,V)=Concat(head1,head x l, ..., head h )W O

[0056] Among them, head i represents single-head attention, h represents the number of heads; W O represents the output weight matrix.

[0057] An application of a cultural resource big data reference model, specifically including a multi-source data integration module, a real-time data perception mechanism, a data fusion engine, a dynamic cultural resource recommendation system, a cultural heritage damage prediction calculation module, and a sentiment analysis feedback system. Specifically, the cultural resource big data reference model is applied to online and offline systems. The framework of public cultural digital resources should grasp the effective combination of online and offline. The application can be subdivided into the application of library data resources, the application of museum data resources, the application of public cultural cloud data resources, and the application of cultural center data resources.

[0058] (III) Beneficial effects

[0059] The present invention provides a method for constructing a cultural resource big data reference model and its application, which has the following beneficial effects:

[0060] 1. The cultural heritage damage prediction and calculation module provided in the present invention provides strong technical support for the protection of cultural heritage. The advantage of cultural heritage damage prediction and calculation is that it can provide scientific, efficient and forward-looking technical support for the protection, restoration and management of cultural heritage, and extend the life of cultural heritage and improve the efficiency of protection work by utilizing high-precision digitization, image recognition and restoration, voice and text analysis and other means.

[0061] 2. By analyzing the emotional feedback of visitors, the present invention can improve the exhibition layout and interactive design, and provide a more personalized and emotional visiting experience. The application of the user emotion analysis feedback system in the cultural data model can not only enhance cultural understanding and resonance, promote the inheritance and innovation of cultural heritage, but also optimize cultural education and experience, support cultural policies and market decisions, promote cultural diversity and inclusiveness, and promote the development of the cultural industry.

[0062] 3. The present invention sets up a distributed database system with high availability, scalability and flexibility. The system stores data in multiple physical nodes in a dispersed manner. Each node has its own database management system, which forms a unified logical database through network connection. It can be deployed across regions and cloud platforms, which facilitates data management and access, reduces communication costs, and improves response speed. It adopts a distributed transaction management mechanism to ensure that data remains consistent among multiple nodes. Through asynchronous replication and consistency mechanisms, while ensuring performance, it ensures data synchronization among various nodes. The system can select the data copy closest to the user according to distance for operation, which reduces communication time and improves the system's processing power and response speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 A schematic diagram of a method for constructing a cultural resource big data reference model of the present invention;

[0064] Figure 2 It is a schematic diagram of the cultural resource big data reference model of the present invention. DETAILED DESCRIPTION

[0065] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0066] Example 1

[0067] According to Figure 1 and 2 As shown, the present invention provides a technical solution:

[0068] A method for constructing a cultural resource big data reference model includes the following steps:

[0069] S1: Set up a multi-source data integration module, which is responsible for collecting data from many data sources, including museums, libraries, cultural heritage monitoring, social media, and web crawlers. The data sources include text, pictures, videos, and 3D scans in different formats;

[0070] S2: Establish a real-time data perception module to use IoT devices to monitor cultural relics in real time, collect environmental data and visitor flow information, and transmit them to the data fusion engine in real time;

[0071] S3: Create a data fusion engine to align and fuse data from different sources and formats in time and space, creating a multi-dimensional cultural resource database to facilitate subsequent analysis and utilization;

[0072] S4: Set up a user feedback analysis module to collect user feedback information for improving system functions and service quality and achieving continuous optimization of the system;

[0073] S5: Construct a cultural heritage damage prediction and calculation module, analyze and process the data in the multi-source data integration module and the real-time data perception module, use algorithms to analyze the damage patterns and risk factors of cultural heritage, realize early warning and formulate protection measures, and digitally restore and enhance damaged or aging cultural resources (such as paintings and documents) using image recognition and processing technology, and enhance their quality through AI for research and display;

[0074] S6: Create a sentiment analysis feedback system to analyze user comments on social platforms and cultural resource platforms to gain insights into user preferences and optimize recommendation algorithms accordingly. Identify audience emotions by analyzing language in social media, on-site feedback, or online interactions. Use sentiment analysis technology to evaluate audience emotional responses to specific cultural exhibits or activities, which is used to improve exhibition design and event planning.

[0075] S7: Set up a digital display platform for cultural resources, use AR / VR technology to provide audiences with an immersive cultural resource display experience, including virtual museum tours and historical scene reproduction, and build a multimodal interactive interface. Design an easy-to-use user interface that supports multiple interactive methods (voice, touch, and somatosensory) to lower the threshold for use and increase user participation;

[0076] S8: Configure a central processing server to process information from the multi-source data integration module, real-time data perception module, user feedback analysis module and sentiment analysis feedback system, and display the information using the cultural resources digital display platform.

[0077] In this embodiment, the application of big data technology can effectively improve the ability to perceive, acquire, manage and process cultural resources, thereby improving overall management efficiency. By digitizing cultural heritage, it can be preserved for a long time and avoid resource loss due to changes in time and environment. Cultural big data provides rich materials for cultural and artistic creation and promotes the development and prosperity of cultural undertakings. Through digitized cultural resources, excellent culture can be disseminated more widely, breaking the limitations of time and distance and benefiting more people. Through data analysis of cultural projects and policies, their effects can be evaluated in a timely manner, and policy measures can be adjusted and optimized. Building a cultural resource big data model has significant advantages in many aspects, from improving management efficiency, protecting cultural heritage, supporting industrial development, promoting cultural inheritance, and promoting scientific and technological innovation, optimizing public services, and assisting decision-making. Each item reflects the huge potential and value of big data in the cultural field.

[0078] Example 2

[0079] According to Figure 2 As shown, this embodiment is optimized as follows on the basis of Example 1. Specifically, a multilingual voice guide system is also set in the cultural resource big data reference model, and a multilingual intelligent voice guide system based on voice recognition and natural language processing technology is developed to provide convenience for tourists from different countries and regions. An intelligent question and answer system is also set to implement an intelligent question and answer system. Users ask questions in natural language, and the system gives answers to provide information about the historical, artistic and scientific value of cultural resources.

[0080] In this embodiment, the benefit of the intelligent question-answering system is that it can provide fast, accurate and continuously optimized information retrieval and processing capabilities, thereby significantly improving user experience and work efficiency. This system uses advanced natural language processing technology to not only parse and understand user questions, but also quickly find and provide accurate answers in large amounts of data.

[0081] Example 3

[0082] According to Figure 2 As shown, this embodiment is optimized as follows based on Example 1. Specifically, the cultural resource big data reference model also sets up a data security protection system to implement data encryption, access control and privacy protection measures to ensure the security of cultural resource data.

[0083] In this embodiment, the data security protection system can effectively prevent network attacks and data leakage risks, and ensure the safe operation of critical information infrastructure. By establishing a data security protection system, it can ensure that personal data is not illegally collected, used or leaked.

[0084] Example 4

[0085] This embodiment is optimized as follows on the basis of Example 1. Specifically, when collecting data, the RGB color collection mode is implemented for the color of cultural resources. The specific formula is: C=R+G+B, C represents a certain color, wherein R, G, and B represent the intensities of the three colors red, green, and blue, respectively, and the intensity value range of each color is 0 to 255. When R=255 (maximum brightness of red), G=0 (no green), and B=0 (no blue), the color presented is pure red. When the brightness of the three primary colors is 255, the mixed color is white. When the brightness of the three primary colors is 0, black is displayed.

[0086] In this embodiment, RGB color acquisition can accurately capture and reproduce the rich colors in nature, and represents colors by mixing the three basic colors of red, green and blue. This representation method is intuitive and easy to understand. The RGB color space can represent more than 16 million different colors, which makes it very close to the colors of nature. In digital image processing, color balance, brightness and contrast can be adjusted, and color saturation can be changed by adjusting the values ​​of the RGB color channels. The RGB color space can be converted with other color spaces. For example, other color feature spaces can be derived through linear or nonlinear transformations. In short, the advantage of RGB color acquisition is that it can accurately capture and reproduce rich colors, providing an intuitive, extensive, mature, precise, diverse, compatible and flexible color application experience.

[0087] Example 5

[0088] This embodiment is based on Example 1 and has been optimized as follows. Specifically, the cultural resource database adopts a distributed database system, including data representation, data sharding, query processing, and data synchronization and consistency maintenance of cultural resources. Cultural resources include historical relics, artworks, and literature. These resources include multiple data types such as text, images, audio, and video. The data representation model is as follows:

[0089] R = {r1, r2, ..., r n}

[0090] Among them, R represents the collection of all cultural resources, r n Represents the nth resource item, each of which contains multiple data types and attributes;

[0091] Sharding is performed based on the geographical location, type, and age attributes of resources. The data sharding model formula is:

[0092] S = {s1, s2, ..., s m}

[0093] Among them, S is the set of all data shards, s m is the mth shard, each shard contains a set of related cultural resources;

[0094] Query processing includes query decomposition, localization processing, and result aggregation. For cultural resource queries, a cost-based model is used to optimize the query plan as follows:

[0095] Cost(Q)=Cost I / O (Q)+Cost CPU (Q)+Cost Network (Q)

[0096] Where Cost(Q) is the total cost of querying Q, including I / O cost, CPU cost, and network communication cost;

[0097] Cultural resource data involves synchronization of multiple data centers and integration with external data sources. Consistency maintenance is achieved through data replication and distributed transaction management. The specific model is expressed as follows:

[0098] Consistency=Replication Factor×Update Protocol

[0099] Among them, Replication Factor is the data replication factor, and Update Protocol is the update protocol, such as the two-phase commit protocol (2PC);

[0100] In this embodiment, the main advantages of the distributed database system are its high availability, scalability and flexibility. The system stores data in a dispersed manner on multiple physical nodes. Each node has its own database management system (DBMS). A unified logical database is formed through network connection. It can be deployed across regions and cloud platforms, which facilitates data management and access, reduces communication costs, and improves response speed. A distributed transaction management mechanism is adopted to ensure that data remains consistent among multiple nodes. Through asynchronous replication and consistency mechanisms, data synchronization between nodes is ensured while ensuring performance. The system can select the data copy closest to the user for operation based on distance, reduce communication time, distribute requests to different nodes through load balancing technology, improve the system's processing power and response speed, and adopt a scalable distributed Database technology is used to store massive data and ensure the integrity, consistency and high availability of the data. The specific models and formulas of the cultural resource distributed database system involve multiple aspects, including data representation, data sharding, query processing, and data synchronization and consistency maintenance of cultural resources. In the cultural resource distributed database system, the data model needs to be able to fully represent the diversity and complexity of cultural resources. Cultural resources include but are not limited to historical relics, artworks, and documentary materials. These resources may include text, images, audio, and video data types. In summary, the design and implementation of the cultural resource distributed database system involves multiple complex models and formulas, which are designed to deal with data representation, sharding, query processing, data synchronization, and consistency maintenance issues. These models and formulas together ensure the effective management and efficient access to cultural resource data.

[0101] Example 6

[0102] This embodiment is based on Example 1 and has made the following optimizations. Specifically, the cultural heritage damage prediction calculation module mainly involves the target detection algorithm Y0L0v5, the deep learning model of text recognition and image processing, and the Y0L0v5 target detection algorithm is used to detect and identify the damage type in the image in real time. The algorithm divides the input image into multiple regions and predicts the bounding box and category probability of each region. The mathematical model formula is:

[0103]

[0104] Among them, P(Object) represents the probability of having an object, C i is the score predicted by the model, and the positioning loss is calculated by the intersection over union (IoU) loss:

[0105] Loss IoU =1-IoU

[0106] The total loss function is:

[0107] Loss total =Loss coord +Loss obj +Loss cls +Loss IoU

[0108] Large models such as BERT and GPT-3 play an important role in text recognition and image processing of historical documents by extracting and analyzing features from scanned images;

[0109] The text recognition model formula is:

[0110]

[0111] Among them, P(w | x) represents the probability of predicting text w given an image x, f(x) is the feature representation of the image, and v w is the vector representation of text w, V is the vocabulary;

[0112] The image processing model formula is:

[0113] G(x)=σ(W·f(x)+b)

[0114] Among them, G(x) represents the processed image, W and b are weights and biases, and σ is the activation function;

[0115] The application of generative adversarial network (GAN) in cultural heritage restoration, the mathematical model formula is:

[0116] G θ (z) = t, D φ (t) = 1-D φ (Gθ (z))

[0117] Among them, G θ represents the generator, D φ Represents the discriminator, z is random noise, t is real data, the generator generates data close to the real data, and the discriminator tries to distinguish the generated data from the real data. The two are continuously optimized through the adversarial process.

[0118] In this embodiment, Y0L0v5 is an efficient target detection algorithm suitable for real-time detection and identification of damage types in images. The algorithm divides the input image into multiple regions and predicts the bounding box and category probability of each region. Secondly, large models such as BERT and GPT-3 play an important role in text recognition and image processing of historical documents. They improve the accuracy of restoration and protection by extracting and analyzing features of scanned images. By training a neural network specifically for extracting visual features of cultural assets, GAN can automatically repair the damaged parts based on these features. This method effectively retains the historical information and artistic value of the original cultural relics. The calculation method and model formula for cultural heritage damage prediction cover multiple aspects from target detection to deep learning to generative adversarial networks. These methods not only improve the accuracy of prediction and repair, but also provide strong technical support for the protection of cultural heritage. The benefit of cultural heritage damage prediction calculation is that it can provide scientific, efficient and forward-looking technical support for the protection, restoration and management of cultural heritage, and extend the life of cultural heritage and improve the efficiency of protection by using high-precision digitization, image recognition and restoration, speech and text analysis and other means.

[0119] Example 7

[0120] This embodiment is based on Example 1 and is optimized as follows. Specifically, the sentiment analysis feedback system is used to identify and analyze the sentiment tendency in the text data. In the sentiment analysis, the text is converted into a word frequency vector, and the probability of the text belonging to different sentiment categories is calculated. The mathematical model formula is:

[0121]

[0122] Where P(c | x) is the probability that the text belongs to category c given the feature vector x, P(x | c) is the probability of feature vector x given category c, P(c) is the prior probability of category c, and P(x) is the probability of feature vector x;

[0123] The contextual information in the text is captured by the self-attention mechanism. The mathematical model formula is:

[0124]

[0125] Among them, Q, K, and V represent query vector, keyword vector, and value vector respectively. k represents the dimension of the keyword vector;

[0126] MultiHead(Q,K,V)=Concat(head1,head x l,...,head h )W O

[0127] Among them, head i represents single-head attention; h represents the number of heads; W O represents the output weight matrix.

[0128] In this embodiment, through in-depth analysis of user emotions, we can more comprehensively understand emotional expressions and resonance points in different cultural backgrounds, which helps to explore the deep emotional elements in cultural works, thereby promoting cross-cultural understanding and communication. User emotion analysis can reveal people's attitudes and emotional tendencies towards traditional culture, which is crucial for the protection and inheritance of cultural heritage. By understanding the emotional acceptance of traditional culture in modern society, we can more effectively formulate protection strategies and education plans. In cultural experience activities such as museums and art exhibitions, by analyzing visitors' emotional feedback, we can improve exhibition layout and interactive design to provide a more personalized and emotional visiting experience. The application of user emotion analysis feedback system in cultural data model can not only enhance cultural understanding and resonance, enhance the inheritance and innovation of cultural heritage, but also optimize cultural education and experience, support cultural policies and market decisions, promote cultural diversity and inclusiveness, and promote the development of cultural industry.

[0129] An application of a cultural resource big data reference model, specifically including a multi-source data integration module, a real-time data perception mechanism, a data fusion engine, a dynamic cultural resource recommendation system, a cultural heritage damage prediction calculation module, and a sentiment analysis feedback system. Specifically, the cultural resource big data reference model is applied to online and offline systems. The framework of public cultural digital resources should grasp the effective combination of online and offline. The application can be subdivided into the application of library data resources, the application of museum data resources, the application of public cultural cloud data resources, and the application of cultural center data resources.

[0130] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A method for constructing a cultural resource big data reference model, characterized in that: The following steps are involved: S1: Set up a multi-source data integration module, which is responsible for collecting data from many data sources, including museums, libraries, cultural heritage monitoring, social media, and web crawlers. The data sources include text, pictures, videos, and 3D scans in different formats; S2: Establish a real-time data perception module to use IoT devices to monitor cultural relics in real time, collect environmental data and visitor flow information, and transmit it to the data fusion engine in real time; S3: Create a data fusion engine to align and fuse data from different sources and formats in time and space to create a multi-dimensional cultural resource database; S4: Set up a user feedback analysis module to collect user feedback information for improving system functions and service quality and achieving continuous optimization of the system; S5: Construct a cultural heritage damage prediction and calculation module, analyze and process the data in the multi-source data integration module and the real-time data perception module, use algorithms to analyze the damage patterns and risk factors of cultural heritage, realize early warning and formulate protection measures, and digitally restore damaged or aging cultural resources through digital restoration and enhancement, using image recognition and processing technology; S6: Create a sentiment analysis feedback system to analyze user comments on social platforms and cultural resource platforms to gain insights into user preferences and optimize recommendation algorithms accordingly. Identify audience emotions by analyzing language in social media, on-site feedback, or online interactions, and use sentiment analysis technology to evaluate audience emotional responses to specific cultural exhibits or activities. S7: Set up a digital display platform for cultural resources: Use AR / VR technology to provide audiences with an immersive cultural resource display experience, including virtual museum tours and historical scene re-enactments, and build a multimodal interactive interface that supports a variety of interactive methods; S8: Configure a central processing server to process information from the multi-source data integration module, real-time data perception module, user feedback analysis module and sentiment analysis feedback system, and display the information using the cultural resources digital display platform.

2. The method for constructing a cultural resource big data reference model according to claim 1, characterized in that: The cultural resource big data reference model also provides a multilingual voice guide system, develops a multilingual intelligent voice guide system based on speech recognition and natural language processing technology, and provides an intelligent question-and-answer system. Users ask questions in natural language, and the system gives answers, providing information about the historical, artistic and scientific value of cultural resources.

3. The method for constructing a cultural resource big data reference model according to claim 1, characterized in that: The cultural resource big data reference model also sets up a data security protection system, implements data encryption, access control and privacy protection measures, and ensures the security of cultural resource data.

4. The method for constructing a cultural resource big data reference model according to claim 1, characterized in that: When collecting data in S1, the RGB color collection mode is implemented for the color of cultural resources. The specific formula is: C=R+G+B, C represents a certain color, wherein R, G, and B represent the intensities of red, green, and blue, respectively, and the intensity value range of each color is 0 to 255. When R=255 (maximum brightness of red), G=0 (no green), and B=0 (no blue), the color presented is pure red. When the brightness of the three primary colors is 255, the mixed color is white. When the brightness of the three primary colors is 0, black is displayed.

5. The method for constructing a cultural resource big data reference model according to claim 1, characterized in that: The cultural resource database in S3 adopts a distributed database system, including data representation, data sharding, query processing, and data synchronization and consistency maintenance of cultural resources. Cultural resources include historical relics, artworks, and documentary materials. These resources include multiple data types such as text, images, audio, and video. The data representation model is as follows: R＝{r1，r2，...，r n} Among them, R represents the collection of all cultural resources, r n Represents the nth resource item, each of which contains multiple data types and attributes; Sharding is performed based on the geographical location, type, and age of resources. The data sharding model is expressed as follows: S={s1,s2,...,s m } Among them, S is the set of all data shards, s m is the mth shard, each shard contains a set of related cultural resources; Query processing includes query decomposition, localization processing, and result aggregation. For cultural resource queries, a cost-based model is used to optimize the query plan as follows: Cost(Q)=Cost I / O (Q)+Cost CPU (Q)+Cost Network (Q) Where Cost(Q) is the total cost of querying Q, including I / O cost, CPU cost, and network communication cost; Cultural resource data involves the synchronization of multiple data centers and the integration with external data sources. Consistency maintenance is achieved through data replication and distributed transaction management. The specific model formula is: Consistenct=Repliation Factor×Update Protocol Among them, Replication Factor is the data replication factor, and Update Protocol is the update protocol.

6. The method for constructing a cultural resource big data reference model according to claim 1, characterized in that: The cultural heritage damage prediction calculation module in S5 mainly involves the target detection algorithm YOLOv5, text recognition and image processing deep learning models; YOLOv5 target detection algorithm is used to detect and identify the type of damage in images in real time. The algorithm divides the input image into multiple regions and predicts the bounding box and category probability of each region. The mathematical model formula is: Among them, P(Object) represents the probability of having an object, C i is the score predicted by the model, and the positioning loss is calculated by the intersection over union (IoU) loss: Loss IoU =1-IoU The total loss function is: Loss total =Loss coord +Loss obj +Loss cls +Loss IoU Large models such as BERT and GPT-3 play an important role in text recognition and image processing of historical documents by extracting and analyzing features from scanned images; The text recognition model formula is: Among them, P(w|x) represents the probability of predicting text w given an image x, f(x) is the feature representation of the image, and v w is the vector representation of text w, V is the vocabulary; The image processing model formula is: G(x)=σ(W·f(x)+b) Among them, G(x) represents the processed image, W and b are weights and biases, and σ is the activation function; The application of generative adversarial network (GAN) in cultural heritage restoration, the mathematical model formula is: G θ (z)=t,D φ (t)=1-D φ (G θ (z)) Among them, G θ Denotes the generator, Dφ denotes the discriminator, z is random noise, and t is real data. The generator generates data close to the real data, while the discriminator tries to distinguish the generated data from the real data. The two are continuously optimized through the adversarial process.

7. The method for constructing a cultural resource big data reference model according to claim 1, characterized in that: The sentiment analysis feedback system in S5 is used to identify and analyze the sentiment tendency in the text data. In sentiment analysis, the text is converted into a word frequency vector, and the probability of the text belonging to different sentiment categories is calculated. The mathematical model formula is: Where P(c|x) is the probability that the text belongs to category c given the feature vector x, P(x|c) is the probability of feature vector x given category c, P(c) is the prior probability of category c, and P(x) is the probability of feature vector x; The contextual information in the text is captured by the self-attention mechanism. The mathematical model formula is: Among them, Q, K, and V represent query vector, keyword vector, and value vector respectively. k represents the dimension of the keyword vector; MultiHead(Q,K,V)=Concat(head1,head2l,...,head h )W O Among them, head i represents single-head attention; h represents the number of heads, W O represents the output weight matrix.

8. The method for constructing a cultural resource big data reference model according to claim 1, characterized in that: The application of the cultural resource big data reference model specifically includes a multi-source data integration module, a real-time data perception mechanism, a data fusion engine, a dynamic cultural resource recommendation system, a cultural heritage damage prediction calculation module, and a sentiment analysis feedback system. Specifically, the cultural resource big data reference model is applied to online and offline systems. The framework of public cultural digital resources should grasp the effective combination of online and offline. The application can be subdivided into the application of library data resources, the application of museum data resources, the application of public cultural cloud data resources, and the application of cultural center data resources.

Images