Digital protection and dissemination of red cultural resources based on the Internet of Things and big data

Abstract

This paper examines the construction and application of a digital protection and dissemination system for red cultural resources based on the Internet of Things (IoT) and big data technologies. By utilizing high-precision sensors and big data tools, the system enables real-time monitoring and data collection of red cultural resources. Additionally, the random forest model is applied for feature extraction and classification prediction. The research encompasses data acquisition and remote processing, model construction and optimization, IoT architecture design, and application system development. Through system evaluation and user feedback analysis, the effectiveness of the digital protection measures and the communication impact are assessed, demonstrating improvements in both the protection of red cultural resources and user engagement. The optimized IoT architecture and model significantly enhance data accuracy and system stability, offering robust support for the long-term protection and widespread dissemination of red cultural resources. This study offers a scientific approach and practical pathway for the digital transformation of red cultural heritage, serving as a valuable reference in the field of cultural heritage protection.

Keywords

Internet of Things big data red culture digital protection communication research

Introduction

The background of this study is rooted in the rapid advancement of information technology, particularly with the widespread adoption of the Internet of Things (IoT) and big data technologies. These developments present new opportunities and challenges for the protection and dissemination of red cultural resources, which are a significant part of Chinese revolutionary history with profound historical and cultural value. However, traditional methods of protecting and communicating these resources suffer from various limitations, such as incomplete data, restricted communication channels, and low public engagement. IoT technology enables real-time monitoring and data collection for red cultural resources, while big data allows for deep analysis of large datasets, offering a novel solution for their digital protection and dissemination. The goal of this study is to establish a system for the digital protection and dissemination of red cultural resources, utilizing big data from IoT devices. This will enable comprehensive protection and broad dissemination through efficient data collection, precise feature extraction, and optimized model development. Ultimately, the research seeks to enhance the protection and communication of red cultural resources, raise public awareness and recognition of red culture, promote its inheritance and development in the modern era, and offer new technical methods and practical approaches for cultural heritage protection and communication.

Regarding the research on the digital protection and dissemination of red cultural resources based on the Internet of Things and big data, the application of the Internet of Things (IoT) and big data technology in the protection of cultural heritage has received extensive attention in recent years. Although Cao studied the relationship between Chinese culture and animal welfare, it also pointed out the compatibility between technology and cultural protection, which has similar challenges to the digital protection of red cultural resources, that is, how to protect the integrity and authenticity of historical culture under modern technological conditions.¹ Li et al. discussed the relationship between religious culture and the dynamic capability of enterprises, indicating that technological innovation can help improve the dynamic protection and management capability of cultural resources, and provide theoretical support for the digital protection of red culture, that is, enhance the vitality and dissemination scope of cultural resources through technology.² In terms of data security and privacy protection, Cui and Qi propose a legal framework for the protection of personal information, which is also essential for the digital protection of red cultural resources. When collecting and processing red cultural resources, big data technology must ensure data security and avoid information loss or misuse in the digitization process, which is of great significance for the long-term protection of cultural resources.³ Yu et al. showed that the application of Internet of Things technology in daily operations and environmental protection can provide strong support for the physical protection of red cultural resources, especially in the environmental monitoring and protection of historical sites and revolutionary memorial sites. Internet of Things technology can monitor environmental changes in real time and prevent the erosion of cultural resources by natural factors.⁴ Yan et al. further emphasizes the importance of cross-cultural communication by studying the spread of Chinese culture in Cuba, which provides reference for the international dissemination of red cultural resources, and can realize the promotion of red cultural resources on a global scale through the Internet of Things and big data technology.⁵ Zhong discussed the application of blockchain technology in the protection of cultural resources, indicating that the use of blockchain technology can ensure the safety of digital storage of red cultural resources and prevent data tampering or loss, which plays an important role in the long-term preservation and effective dissemination of red cultural resources.⁶ Through the Internet of Things and big data technology, the protection and dissemination of red cultural resources can be more efficient and cover a wider range, while ensuring the integrity and security of cultural heritage.

This study is of great significance and value in the field of digital protection and dissemination of red cultural resources. First of all, from the perspective of theoretical significance, by introducing the Internet of Things and big data technology, this study innovates the research methods for the protection and dissemination of red cultural resources and enriches the theoretical system in related fields. The Internet of Things technology realizes the real-time monitoring and accurate data collection of red cultural resources, while the big data technology provides a powerful means of data analysis and mining, which effectively improves the scientific nature of resource protection and the accuracy of communication. Secondly, from the perspective of practical significance, the digital protection and communication system constructed in this study can not only greatly improve the protection level of red cultural resources and avoid the damage and loss of resources, but also expand the influence and dissemination scope of red culture through diversified communication channels, and enhance the cultural cognition and recognition of the public. In addition, this study is also of great value for improving the management efficiency and decision-making level of red cultural resources. Through big data analysis, it can provide data support and scientific basis for resource management and optimize resource allocation and protection strategies. In general, this study has significant value in promoting the digital transformation of red cultural resources, promoting the protection and dissemination of cultural heritage, and enhancing cultural soft power, etc., and provides useful reference and reference for the research and practice in related fields.

In recent years, the application of Internet of Things (IoT) and big data technologies in cultural heritage protection has gained significant attention. IoT enables real-time monitoring of environmental conditions around cultural relics, while big data analytics allow for comprehensive data processing and predictive maintenance. Previous studies have demonstrated the successful integration of IoT sensors to monitor parameters such as temperature, humidity, and air quality in museums and historical sites. Big data technologies have been employed to analyze the collected data, helping to predict potential risks to cultural assets and optimize preservation strategies. These technologies have also been used to create digital archives and interactive platforms, enhancing public engagement with cultural heritage. The innovation in this research lies in combining IoT and big data to protect and disseminate red cultural resources specifically, which has not been extensively explored. By applying advanced feature extraction models like random forest and incorporating user feedback analysis, the approach offers a more dynamic and data-driven method of preserving and promoting red cultural resources, setting it apart from previous methods that primarily focused on physical preservation without leveraging large-scale data analytics.

Red cultural resources refer to the historical sites, memorials, literature, and other forms of cultural heritage related to significant revolutionary events under the leadership of the Communist Party of China. These resources embody the revolutionary spirit, patriotism, and collectivism that are crucial to understanding China’s modern history. The importance of red cultural resources lies not only in their historical value but also in their cultural and educational roles. They serve as a bridge to connect present and future generations with the nation’s revolutionary legacy, inspiring national pride and social responsibility. Red culture contributes to cultural identity, national cohesion, and the promotion of socialist core values. Protecting these resources is essential for maintaining their integrity and ensuring that future generations can learn from and be inspired by them. In the digital age, it is especially important to adopt modern technological methods to preserve and disseminate these cultural assets, as traditional methods alone are insufficient to protect them from degradation and to reach broader audiences. By ensuring the long-term preservation and wide dissemination of red cultural resources, their role in shaping modern society and reinforcing cultural values can be strengthened.

Overview of relevant theories

Brief introduction of red culture theory

Red culture refers to the cultural phenomena and heritage related to historical periods such as revolutionary struggles, nation-building, and reforms led by the Communist Party of China. It includes various elements such as ideologies, moral values, literature, art, and memorial sites. With its distinct revolutionary history and spiritual significance, red culture embodies the revolutionary spirit, patriotism, and collectivism of the Communist Party and its martyrs. Red culture not only reflects the social and cultural attributes of a particular historical period, but it also plays a crucial role in inspiring patriotic sentiment and a sense of historical responsibility by honoring and promoting the deeds and spirits of revolutionary predecessors. As a unique cultural form, red culture holds significant historical and cultural value and contributes to the construction of spiritual civilization and moral education in modern society. The preservation and dissemination of red culture have a lasting influence on the inheritance and promotion of China’s cultural traditions, as well as on enhancing national cohesion and cultural confidence.

Big data analysis theory

Big data analysis theory refers to the framework and methodologies used to collect, store, process, and analyze large, complex, and diverse datasets through modern information technology. The core of big data analysis lies in extracting valuable information and knowledge from vast datasets using efficient processing techniques and advanced algorithms. It involves several stages, including data preprocessing, mining, machine learning, and data visualization. During the preprocessing phase, raw data is cleaned, transformed, and reduced to enhance quality and efficiency. In the data mining and machine learning stages, patterns and rules are identified through classification, clustering, and association analysis. Finally, data visualization displays analytical results in charts and graphs for easier interpretation and decision-making. Big data analysis spans various technical fields, including computer science, statistics, and mathematics, and is applied across industries such as economics, healthcare, and transportation. This approach enables a comprehensive understanding of complex systems and accurate predictions, providing valuable support for scientific research, business decisions, and public governance.⁷

Digital communication theory

Digital communication theory explores how information is produced, transmitted, and received within a digital media environment. It focuses on the profound changes brought about by rapid advances in information and communication technologies, particularly the Internet and mobile communication. Key aspects of digital communication include encoding, storage, transmission, and decoding of information, with an emphasis on how digital technology transforms the entire communication process. The theory highlights the seamless reproduction and rapid dissemination of information, the diversification of communication channels, and the increased audience participation and interaction. Digital communication breaks the traditional constraints of time and space, enabling real-time transmission and global sharing of information. Moreover, it encourages the shift from one-way communication to multi-directional interaction, enhancing audience engagement and feedback. Issues such as information overload, the digital divide, and privacy concerns are also central to the theory, reflecting the opportunities and challenges posed by the digital age. A deeper understanding of digital communication mechanisms helps optimize information strategies and improve the overall effectiveness of communication efforts.⁸

Data acquisition and feature extraction

Data acquisition methods and tools

Data acquisition methods and tools refer to the technologies and equipment used to collect, capture, and store data for analysis. The primary methods include sensor-based acquisition, web crawling, manual input, and third-party data interfaces. Sensor acquisition involves real-time monitoring of environmental or equipment conditions using IoT devices like temperature sensors, humidity sensors, and RFID tags. Web crawlers automatically extract public data from the internet, while manual entry is typically employed for structured data and is suitable for small-scale, precise data collection. Third-party interfaces obtain data via API calls from external sources. Data acquisition tools include both hardware devices and software systems. Hardware, such as sensors and data acquisition cards, is responsible for collecting and transmitting real-time data. Software, including data management systems and data cleaning tools, ensures proper data storage, management, and processing. Effective acquisition methods and tools ensure data accuracy, completeness, and timeliness, laying a strong foundation for further analysis. As technology evolves, data acquisition methods continue to improve, adapting to the growing complexity of different data types and application scenarios.

Data cleaning and teleprocessing technology

Data cleaning and teleprocessing are critical steps in the data analysis process, aimed at improving data quality and ensuring the accuracy and validity of subsequent analysis. Data cleaning typically involves removing noisy data, handling missing values, and detecting and correcting outliers. Noise removal eliminates errors and irrelevant information, often through filtering or clustering techniques. Missing data can be handled by removing records, interpolation, or mean filling, depending on the dataset and application context. Outlier detection and correction are achieved through statistical methods or machine learning algorithms to ensure data consistency. Teleprocessing further refines the cleaned data by applying standardization, normalization, and discretization techniques. Standardization adjusts data to a uniform scale for better feature comparison, normalization maps values into a fixed range to eliminate dimensional biases, and discretization converts continuous data into categorical values for certain models. These techniques significantly enhance data quality and consistency, providing a solid foundation for feature extraction and model development.

Data processing involves several critical steps to ensure accuracy and reliability. The data cleaning process begins with the removal of noise, which is achieved through filtering techniques that eliminate outliers and irrelevant information captured by sensors. Missing values are handled using methods such as mean imputation for continuous data or by removing records with extensive missing information. Outlier detection is applied using statistical techniques and machine learning algorithms to identify abnormal data points, which are either corrected or removed. After cleaning, the data undergoes teleprocessing, where remote servers handle data aggregation, normalization, and transformation. Data normalization involves scaling features to a consistent range, ensuring compatibility for machine learning models, while standardization adjusts the data to fit a normal distribution. This remote processing stage leverages cloud computing for large-scale data management, enabling efficient and secure handling of large datasets. These steps are crucial in preparing the data for further analysis, including feature extraction and model training, ensuring that the processed data is accurate and representative of the underlying conditions of red cultural resources.

Feature engineering method

Feature engineering involves extracting feature variables that can significantly enhance model performance by processing and transforming raw data. In the context of the digital protection and dissemination of red cultural resources based on IoT and big data, feature engineering plays a crucial role. First, feature extraction transforms multidimensional data collected by IoT devices into easily analyzable features, such as converting sensor data—temperature, humidity, and light intensity—into comprehensive indicators representing environmental conditions. Next, feature construction generates new features through mathematical or logical operations tailored to the specific attributes and research requirements of red cultural resources. For instance, regional distribution features may be derived from geographical data, or time series analysis may be used to extract trends in access frequency. Feature selection, on the other hand, involves identifying the most impactful features for model prediction from a large pool of features. Common approaches include filtering, wrapping, and embedding, which help reduce data dimensionality, improve computational efficiency, and prevent overfitting. An effective feature engineering approach can significantly improve model accuracy and performance in areas such as resource classification, protection assessment, and communication mode analysis, providing a robust data foundation and technical support for the digital preservation and dissemination of red cultural resources.⁹

Feature selection technology

As shown in Table 1, feature selection technology is crucial in the study of digital protection and dissemination of red cultural resources, which helps to improve model performance and computational efficiency. In this study, the random forest importance score, recursive feature elimination (RFE) and L1 regularization were used to screen and evaluate the features. Table data show that preservation status score (0.85), digitization degree score (0.78) and content richness score (0.75) are outstanding in random forest importance score, so they are reserved for model training. A score of 0.65 for the number of visitors was similarly retained after its importance was determined by the RFE method, while a lower score (0.30) for the glocalization code was discarded in the RFE evaluation, showing that it contributed less to the model’s predictive effect. In addition, chronological features were evaluated by L1 regularization method with an importance score of 0.55 and were retained in the feature set. Through feature selection, the most representative and predictive features are retained, the data dimension is reduced, and the computational efficiency and prediction accuracy of the model are improved. This not only avoids the computational overhead caused by redundant features, but also effectively prevents overwriting. Finally, the selected feature combination performs well in the classification and prediction of red cultural resources, providing a solid data foundation and technical support for digital conservation and dissemination strategies. This process verifies the key role of feature selection technology in improving model performance, and provides an important reference for subsequent research and practice.¹⁰

Table 1.

Evaluation of feature selection techniques for red cultural resources.

Feature	Importance score	Selection method	Retention status
Preservation status score	0.85	Random forest importance	Retained
Vitalization score	0.78	Random forest importance	Retained
Visitor count	0.65	Recursive feature elimination (RFE)	Retained
Geographic location code	0.30	Recursive feature elimination (RFE)	Discarded
Era	0.55	L1 regularization	Retained
Content richness score	0.75	Random forest importance	Retained

Feature selection plays a critical role in improving model performance and computational efficiency by identifying the most relevant features for prediction. In this case, features such as Preservation Status Score, Vitalization Score, and Visitor Count were chosen based on their high importance scores, as shown in the random forest importance and Recursive Feature Elimination (RFE) evaluations. These features have strong predictive power for red cultural resource classification, as they directly correlate with the preservation quality, engagement, and usage metrics of the resources. Discarding less influential features, such as Geographic Location Code, reduces data complexity, avoids overfitting, and enhances model accuracy. Feature selection refines the model’s focus, improves its predictive capacity, and ensures faster computation by reducing dimensionality.^11,12 The selected features lead to a more stable and reliable model, as reflected in the classification results and prediction effectiveness for resource protection and dissemination.

In the feature extraction and selection process, specific features were chosen based on their relevance to red cultural resource classification and preservation analysis. The Preservation Status Score was selected because it directly reflects the current state of the resource and is critical for predicting future conservation needs. The Vitalization Score was included as it measures the degree of digital transformation and engagement, providing insights into the effectiveness of preservation efforts. Visitor Count is another important feature, representing user interaction and public interest, which can directly influence the sustainability of the resource. The Geographic Location Code was discarded due to its low correlation with the preservation outcome, as determined by Recursive Feature Elimination (RFE). The feature selection process, particularly the use of random forest importance scores, significantly improves model performance by focusing on high-impact features and removing irrelevant or redundant data. This enhances the model’s accuracy and reduces computational complexity, leading to more efficient resource management and better prediction of conservation outcomes.

Model construction and sot design

After the completion of data collection and feature extraction, the next step is to build an effective model and design a suitable sot architecture to realize the digital protection and dissemination of red cultural resources. The model construction part will introduce the principle, construction, training, parameter optimization and evaluation methods of the random forest model in detail, while the sot design part includes the architecture design and application system development to ensure the efficient collection, transmission and processing of data, and finally realize the comprehensive protection and wide dissemination of red cultural resources. The specific model construction process is shown in Figure 1.

Figure 1.

Schematic diagram of model construction and sot design.

Principles of random forest model

Random forest model is an ensemble learning method that combines multiple decision trees to improve the accuracy and robustness of the model. The basic principle is to construct multiple independent decision trees through random sampling and feature selection, and synthesize the predicted results of these trees to obtain the final classification or regression results. The construction process of random forest consists of two main steps: Firstly, random sampling with retractions is carried out on the training set to generate several different subsets; Secondly, a decision tree is trained for each subset, and some features are randomly selected to split during the construction of each decision tree. In this way, the random forest not only reduces the overwriting problems that may exist in a single decision tree, but also improves the overall prediction performance by integrating the results of multiple models.

The prediction formula of random forest is as follows:

\hat{y} = \frac{1}{N} \sum_{i = 1}^{N} h_{i} (x)

(1)

The main advantages of the random forest model include strong noise resistance, the ability to handle high-dimensional data, and good robustness to data distribution and missing values. In addition, the random forest can output a feature importance score, which helps to understand the feature’s contribution to the model’s prediction results.

Model construction

Before building the random forest model, the collected data is made into a table, which includes various data reflecting the characteristics of red cultural resources, such as resource type, preservation state, number of visitors, etc., as shown in Table 2.

Table 2.

Red cultural resources feature data table.

Resource ID	Resource type	Preservation status score	Monthly visitors	Vitalization score	Geographic location code
1	Revolutionary site	7.3	167.89	8.1	101
2	Revolutionary museum	8.1	243.67	9.4	102
3	Revolutionary relic	6.9	98.43	7.8	103
4	Red literature	7.5	123.56	8.6	104
5	Memorial sculpture	8.3	187.24	9.0	105

Construct a random forest model with three decision trees, N = 3, to predict the conservation effect score of a certain red cultural resource. The characteristic data of this resource is set as follows: the resource type is Revolutionary memorial Hall, the preservation status score is 8.1, the number of monthly visitors is 243.67, the digitization degree score is 9.4, and the geographic location code is 102.

Prediction of decision tree 1:

Input characteristics: 8.1 (save status score), 243.67 (number of visitors per month), 9.4 (Digitization score), 102 (glocalization code)

Prediction result: 8.5 (protection effect score)

Prediction of decision tree 2:

Input characteristics: 8.1, 243.67, 9.4, 102

Forecast result: 8.7

Prediction of decision tree 3:

Input characteristics: 8.1, 243.67, 9.4, 102

Forecast result: 8.6

Based on the prediction results of the three decision trees, the final protection effect score is

\hat{y} = \frac{1}{3} (8.5 + 8.7 + 8.6) = \frac{1}{3} \times 25.8 = 8.6

(2)

Through the random forest model, the final conservation effect score of red cultural resources is obtained by combining the prediction results of multiple decision trees. This process not only improves the stability and accuracy of the model, but also provides importance analysis of features to help optimize resource conservation and dissemination strategies.

Model training

After building the random forest model, it is necessary to train the model so that it can accurately predict the conservation effect and dissemination impact of red cultural resources. The key steps of model training include sample division, feature selection, model training and evaluation. These steps are described in detail and the relevant formulas are provided below.

First, the data set is divided into a training set and a test set. Let’s assume that dataset D contains m samples, where $D_{train}$ is the training set and contains the sample of $\frac{2}{3}$ , and $D_{test}$ is the test set and contains the sample of $\frac{1}{3}$ .

D_{train} = {(x_{i}, y_{i})}_{i = 1}^{m_{train}}, D_{test} = {(x_{i}, y_{i})}_{i = 1}^{m_{test}}

(3)

To select the most important features for training, the common method is to calculate the importance score of the features. The training of a random forest model involves building multiple decision trees, each of which is trained using a different subset. Suppose the random forest contains N, let’s say N = 3. The training formula for each tree is

T_{i} = TrainDecisionTree (D_{train}^{(i)})

(4)

The final prediction of the model is the average of the predictions of all trees:

\hat{y} = \frac{1}{N} \sum_{i = 1}^{N} h_{i} (x)

(5)

The model is trained using the following feature data, as shown in Table 3.

Table 3.

Model parameters.

Resource ID	Preservation status score	Monthly visitors	Vitalization score	Protection effectiveness score
1	7.3	167.89	8.1	7.8
2	8.1	243.67	9.4	8.5
3	6.9	98.43	7.8	7.1
4	7.5	123.56	8.6	7.6
5	8.3	187.24	9.0	8.2

1. Build the training set of decision tree 1:

Generate training subsets using sampling with retractions $D_{train}^{(1)}$

2. Training formula of decision tree 1:

T_{1} = TrainDecisionTree (D_{train}^{(1)})

(6)

3. Build the training set of decision Tree 2:

Generate training subsets using sampling with retractions $D_{train}^{(2)}$

4. Training formula of decision tree 2:

T_{2} = TrainDecisionTree (D_{train}^{(2)})

(7)

5. Build the training set of decision Tree 3:

Generate training subsets using sampling with retractions $D_{train}^{(3)}$

6. Training formula of decision tree 3:

T_{3} = TrainDecisionTree (D_{train}^{(3)})

(8)

7. Synthesize the prediction results of all decision trees:

\hat{y} = \frac{1}{3} (h_{1} (x) + h_{2} (x) + h_{3} (x))

(9)

The final protection effect score is

\hat{y} = \frac{1}{3} (8.4 + 8.5 + 8.6) = \frac{1}{3} \times 25.5 = 8.5

(10)

In this way, the random forest model can synthesize the prediction results of multiple decision trees to improve the accuracy and stability of the prediction.

Parameter optimization

Parameter optimization is a critical step in enhancing the performance of the random forest model, aiming to improve its prediction accuracy and stability by adjusting key model parameters. The primary parameters include the number of trees (N), the maximum depth of each tree (d), the minimum number of samples per node (min_samples_split), and the number of features used for splitting (max_features). During the optimization process, Grid Search and Cross-Validation are employed to systematically explore different parameter combinations and assess the performance of each on a validation set. These methods help identify the optimal parameter set, ensuring the model generalizes well across various datasets. For instance, tuning the number of trees and the maximum depth allows control over the model’s complexity, preventing overfitting or underfitting. Additionally, adjusting the minimum sample size and the number of split features improves computational efficiency and model accuracy. Through iterative optimization, the configuration with the best performance is selected, enabling the random forest model to excel in predicting the protection and dissemination of red cultural resources, thus providing robust technical support for practical application and promotion.

Model evaluation

Model evaluation is essential to ensure the effectiveness and reliability of the random forest model in real-world applications.¹³ The evaluation process begins by assessing accuracy, a fundamental metric representing the proportion of correct predictions in the overall classification task. Precision and recall are also analyzed to gauge the model’s performance in different categories, where precision measures the correct proportion of predicted positive samples, and recall measures the correct proportion of actual positive samples. The F1 score is then calculated as a harmonic mean of precision and recall, balancing the two metrics. To further verify the model’s stability and generalizability, cross-validation is employed, where the model’s performance is evaluated on various data partitions using repeated training. This ensures that the model is not biased by the peculiarities of any specific dataset. Finally, plotting the ROC curve and calculating the AUC value (Area Under the Curve) provides insight into the model’s classification capability under different threshold settings. These comprehensive evaluation metrics ensure that the random forest model achieves high accuracy, stability, and reliability in predicting the protection and dissemination of red cultural resources.

Internet of Things architecture design

Model evaluation is essential to ensure the effectiveness and reliability of the random forest model in real-world applications. The evaluation process begins by assessing accuracy, a fundamental metric representing the proportion of correct predictions in the overall classification task. Precision and recall are also analyzed to gauge the model’s performance in different categories, where precision measures the correct proportion of predicted positive samples, and recall measures the correct proportion of actual positive samples. The F1 score is then calculated as a harmonic mean of precision and recall, balancing the two metrics. To further verify the model’s stability and generalizability, cross-validation is employed, where the model’s performance is evaluated on various data partitions using repeated training. This ensures that the model is not biased by the peculiarities of any specific dataset. Finally, plotting the ROC curve and calculating the AUC value (Area Under the Curve) provides insight into the model’s classification capability under different threshold settings. These comprehensive evaluation metrics ensure that the random forest model achieves high accuracy, stability, and reliability in predicting the protection and dissemination of red cultural resources.

Application system development

The development of the application system is the practical implementation of IoT and big data technology for the digital protection and dissemination of red cultural resources. The first step is to define the system’s functional requirements, including data collection and monitoring, data storage and management, data analysis and display, and user access and interaction. The system adopts a modular design to ensure the independent development and maintenance of each functional component, improving scalability and reliability. The data acquisition module integrates various sensors and equipment for real-time monitoring of red cultural resources, while the data storage and management module uses cloud storage and database technology to ensure secure and efficient data handling. The data analysis and display module employs big data platforms and visualization tools to clean, process, and analyze the collected data, presenting the results in charts, reports, and other formats to support decision-making. The user access and interaction module provides a user-friendly interface, offering multi-platform access (e.g., PC, mobile, and tablet) for inquiries, browsing, and interaction. To guarantee the system’s safety and reliability, strict testing and verification processes, including functional, performance, and security tests, are conducted. Additionally, the system is designed with a robust maintenance mechanism to support real-time monitoring, fault diagnosis, and automatic recovery, ensuring long-term stable operation. Through this systematic and modular development process, the application system efficiently supports the digital protection and dissemination of red cultural resources, enabling effective data management and broad information dissemination.

The design of the application system interface focuses on providing a user-friendly and intuitive experience that supports the digital protection and dissemination of red cultural resources. The interface is structured to facilitate easy navigation, with clear categorization of resources such as revolutionary sites, museums, and literature. Users can quickly access detailed information, view historical data, and interact with real-time monitoring through an interactive dashboard. Visual elements such as charts, graphs, and icons are utilized to display data analysis results and key performance indicators, improving usability and user engagement. User feedback has been positive, highlighting the clarity of information presentation and the simplicity of interactions. Access to the system is available across multiple platforms, including PC, mobile, and tablet, ensuring that users can engage with the system seamlessly from any device. This multi-terminal design enhances the system’s practical usability, increasing user satisfaction and participation in the preservation and communication of red cultural heritage.

Data interpretation and result analysis

Analysis of classification effect of red cultural resources data

As shown in Figure 2, for revolutionary sites, the classification accuracy of the model reaches 98.5%, the accuracy rate is 97.8%, and the recall rate is 98.0%, indicating that the model can efficiently identify and classify such resources with a low error rate. The classification results of the Revolutionary Memorial show that the accuracy rate of the model is 99.2%, the accuracy rate and the recall rate are 99.0% and 99.1%, respectively, which means that the model is very accurate in the identification of resources in this category, with almost no misjudgments. The accuracy rate of the classification of revolutionary cultural relics was 97.8%, the accuracy rate and the recall rate were 97.5% and 97.6%, respectively, which was slightly lower than other categories, but still showed a high classification ability. The classification accuracy of red books, documents and memorial sculptures is 98.3% and 98.9%, respectively, and the accuracy and recall rates are close to 98%, which further verifies the stability and reliability of the model in the classification of various red cultural resources. On the whole, the model shows a high level of accuracy and consistency in the classification of different categories of red cultural resources, especially in the classification of revolutionary memorials and memorial sculptures. This indicates that the random forest model can effectively support the digital protection and dissemination of red cultural resources, and provide a solid data foundation and technical support for the subsequent resource management and utilization.

Figure 2.

Evaluation of classification effectiveness of red cultural resources.

Digital protection effect evaluation

As shown in Figure 3, the vitalization degree score is high, with the highest score of 9.2 (resource ID is 2) and the lowest score of 7.8 (resource ID is 3), indicating that all resources have invested and achieved remarkable results in digital protection. The data integrity score is generally close to perfect, reaching a maximum of 9.5 (resource ID is 2), which indicates that the preservation and maintenance of data is well guaranteed during the digitization process. The frequency increase score reflects the improvement of the attractiveness of digital protection measures to resources. The highest rated resource ID is 2, at 8.9, while the lowest resource ID is 3, also has a score of 7.2, which means that digital protection significantly improves the frequency of access to the resource and user engagement. In terms of user satisfaction scores, all resources have high scores, with the highest being 9.2 (resource ID 2) and the lowest being 8.3 (resource ID 3), which indicates that digital protection measures are highly recognized and satisfied by users. The data recovery score evaluates the recovery ability of the system after data loss or damage. The score is generally high, with the highest being 9.0 (resource ID is 2) and the lowest being 8.3 (resource ID is 3), indicating that the digital protection system has the ability to respond quickly and recover, and ensure the continuous availability of data. The digital protection measures of red cultural resources perform well in the aspects of digitization degree, data integrity, increase of access frequency, user satisfaction and data recovery ability. These results verify the importance and effectiveness of digital protection for red cultural resources, and provide solid data support and guidance for future promotion and optimization.

Figure 3.

Evaluation of digital preservation effectiveness of red cultural resources.

Communication mode and user feedback analysis

As shown in Figure 4, firstly, the communication coverage score reflects the coverage breadth of red cultural resources in each communication channel. The highest score is the resource ID of 2, reaching 9.1, indicating that the resource has the best dissemination effect and the widest coverage on various digital platforms. The lowest score, 7.8 (resource ID 3), is slightly lower, but still performs well, indicating that most resources can be widely disseminated through digital platforms. The Interactive engagement score measures how often users interact with red cultural resources on digital platforms. The interaction participation score of resource 2 is the highest (8.5), indicating that users are highly engaged with this resource. Resource ID 3 has the lowest score of 7.4, indicating that the resource has room for improvement in terms of interactivity. The content richness score evaluates the diversity and richness of digital resource content. The highest rated resource ID is 2 and 9.0, indicating that the resource has advantages in content production and presentation. The lowest score is 8.0 (resource ID is 3), indicating that content richness is generally high, but still needs further improvement. The information accuracy score reflects the accuracy and reliability of digital resources in information transmission. Resource ID 2 once again performs well with a score of 9.4, indicating high accuracy of its information. The lowest score is 8.7 (resource ID is 3), which is slightly lower, but the overall accuracy of information delivery is good.

Figure 4.

Evaluation of dissemination modes and user feedback for red cultural resources.

The user satisfaction score comprehensively reflects the user’s overall satisfaction with digital resources. The highest score of 8.8 (resource ID 2) indicates that users are most satisfied with it. The lowest score is 7.9 (resource ID is 3), indicating that the user experience needs to be further optimized. The communication mode and user feedback of red cultural resources perform well, especially the resource ID 2 is outstanding in all indicators, showing its successful experience in digital communication. The resource ID of 3 is relatively low, but the overall level remains high. These analysis results provide data support for further optimizing the communication mode of red cultural resources, and put forward improvement directions for improving user satisfaction.

Summarize the problems and research suggestions

Problem summary

In the study of digital protection and dissemination of red cultural resources, remarkable results have been achieved through the construction of random forest model and the design of Internet of Things architecture, but some problems and challenges have also been exposed. First of all, in the data acquisition and teleprocessing stage, there are problems of low data quality, such as noise and missing values in some sensor data, which affect the accuracy of the model. Secondly, in the process of feature engineering and feature selection, although the random forest model is used to achieve the importance ranking of features, the feature extraction method still needs to be further optimized to improve the predictive performance of the model. In the stage of model training and parameter optimization, although cross-validation and grid search techniques are adopted, the complexity and computational cost of the model still need to be balanced to avoid overwriting or underwriting. In addition, in the design of the Internet of Things architecture, the stability and security of data transmission still need to be strengthened, especially in the process of large-scale data transmission and real-time monitoring, which may encounter problems of network delay and data loss. In the process of application system development, user interface friendliness and system usability need to be further improved to ensure that users can easily access and interact. In the analysis of communication mode and user feedback, although each communication channel performs well, the communication strategies of different channels need to be further refined to better meet the needs of different user groups. Overall, this study has achieved initial results in the digital protection and dissemination of red cultural resources, but it still needs further improvement and optimization in data quality, model optimization, system stability and user experience.

Research suggestions

In order to further improve the digital protection and dissemination of red cultural resources, this study puts forward the following suggestions. First of all, in the data acquisition stage, sensor equipment with higher accuracy and stability should be used, and data cleaning and re-processing technology should be strengthened to ensure data quality and integrity. Secondly, in terms of feature engineering, it is suggested to introduce more advanced feature extraction and selection methods, such as deep learning technology, to improve the predictive performance and robustness of the model. In the process of model training and parameter optimization, more advanced optimization algorithms such as Bayesian optimization can be explored to further improve the accuracy and efficiency of the model. At the same time, the architecture design of the Internet of Things should strengthen the stability and security of data transmission, and adopt multi-level encryption and redundant transmission technology to ensure the reliability and security of data in the transmission process. In the application system development, we should pay attention to improving the friendliness and interactivity of the user interface, and optimize the usability of the system through user experience design (UXD), so that users can access and interact more conveniently. In addition, according to the characteristics of different communication channels, more detailed and targeted communication strategies should be formulated, and the extensive coverage and interaction of social media, the authority and credibility of official websites, and the rich display forms of online exhibition platforms should be utilized to maximize the influence of red cultural resources and user participation. Finally, it is recommended to conduct regular system evaluation and feedback collection to continuously optimize system function and performance through user feedback to ensure the continuous effectiveness and improvement of digital protection and communication measures.

Conclusions

In this study, the whole process from data collection, feature extraction, model construction and training to data interpretation and result analysis is systematically discussed through the construction of a digital protection and dissemination system of red cultural resources based on sot big data. First of all, in terms of data acquisition and feature extraction, high-precision sensors and advanced data cleaning and teleprocessing technologies are used to ensure the quality and integrity of the data. Through in-depth feature engineering, the key features that can significantly improve the performance of the model are extracted, which lays a solid foundation for the subsequent model construction. In the part of model construction, the introduction and optimization of random forest model significantly improved the accuracy and stability of classification and prediction of red cultural resources. The architecture design of the Internet of Things ensures the real-time collection, stable transmission and secure storage of data, and supports the efficient operation of the entire system. In the process of application system development, through the modular design and friendly user interface, it provides convenient access and interaction functions, and improves the user experience. In the part of data interpretation and result analysis, the classification effect of red cultural resources, digital protection effect, communication mode and user feedback are deeply analyzed, and the practical application effect of the system is verified. Although the research has achieved remarkable results, it also found some problems, such as low data quality, model optimization needs to be further improved, and system stability and security need to be strengthened. Based on these findings, this study proposes a series of improvement recommendations, including the use of higher precision sensor devices, the introduction of deep learning technology to optimize feature extraction, the use of advanced optimization algorithms to improve model performance, strengthen the security and stability of data transmission, improve the user interface friendliness of the system, and develop targeted communication strategies. In conclusion, this study provides a scientific and effective technical plan and practical path for the digital protection and dissemination of red cultural resources, and has important guiding significance for future research and practical application.

The digital protection and dissemination system for red cultural resources, based on the Internet of Things and big data, has proven to be a comprehensive solution that enhances both the preservation and promotion of these valuable resources. By integrating high-precision sensors and big data technology, the system ensures efficient data collection and processing, leading to accurate real-time monitoring of cultural sites. The use of the random forest model for feature extraction and classification further refines the predictive accuracy, optimizing resource management and enhancing user participation. The modular IoT architecture design guarantees stable data transmission and secure storage, while the user-friendly interface improves accessibility across various platforms. Additionally, the analysis of communication effectiveness and user feedback highlights the system’s success in broadening the dissemination of red cultural heritage, increasing public awareness, and engagement. Overall, the system not only preserves the integrity of red cultural resources but also expands their reach and impact in the digital age, offering a sustainable solution for future cultural heritage protection.

The digital protection and dissemination system for red cultural resources, leveraging the Internet of Things and big data technologies, significantly enhances both preservation and communication strategies. Through the integration of high-precision sensors, the system ensures real-time monitoring and data acquisition, improving the accuracy and timeliness of information related to red cultural heritage. The use of advanced data cleaning and teleprocessing technologies ensures the data integrity, which supports more reliable feature extraction and model building. The random forest model has been particularly effective in analyzing key features such as preservation status and visitor engagement, leading to optimized resource management strategies. The IoT architecture, designed for scalability and security, ensures stable data transmission and supports large-scale monitoring, enabling broader communication reach. Additionally, user feedback has indicated that the system’s interface, with its intuitive design and cross-platform accessibility, enhances user interaction and satisfaction, further boosting public engagement with red cultural heritage. This comprehensive approach not only preserves the historical and cultural value of red resources but also promotes their dissemination to a wider audience, ensuring their relevance in the digital era and contributing to long-term cultural sustainability.

The innovative aspects of this study lie in the integration of IoT and big data technologies to create a dynamic system for the digital protection and dissemination of red cultural resources. By utilizing high-precision sensors and the random forest model for feature selection and prediction, the system offers a more accurate and efficient method for real-time monitoring and data-driven resource management. The application of these technologies enhances both the preservation efforts and the public’s engagement with red cultural heritage, demonstrating its practical value in addressing the limitations of traditional methods. In terms of future research, there is potential to further explore the integration of artificial intelligence and machine learning algorithms to optimize predictive models and develop more personalized user interaction features. Additionally, expanding the system’s application to other forms of cultural heritage, and enhancing international collaboration, could lead to broader impacts in the global context of cultural preservation and dissemination.

Footnotes

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Cao

. Is the concept of animal welfare incompatible with Chinese culture? Soc Anim 2022; 30(3): 284–296.

Zhang

. Intellectual property protection, religious culture and dynamic capabilities of enterprises: evidence from Chinese listed companies. Technol Anal Strateg 2024; 36(6): 1263–1277.

Cui

. The legal construction of personal information protection and privacy under the Chinese civil code. Comput Law Secur Rev 2021; 41: 105560.

Tso

Chen

, et al. Day-to-day organizational life and environmental protection: Evidence from Chinese chemical firms. Corpor Soc Resp Env Ma 2024; 31(5):4504–4511.

Yan

Lai

. Does Confucian culture influence corporate R&D investment? evidence from Chinese private firms. Financ Res Lett 2021; 40: 101719.

Zhong

. The development and dissemination of intelligent storage management software of Chinese excellent traditional culture based on blockchain in the era of financial media. Heliyon 2024; 10(7): e27262.

Zuo

Liu

. The information needs of Chinese family members of cancer patients in the online health community: what and why? Inf Process Manag 2021; 58(3): 102517.

Zhao

Kim

. The impact of traditional Chinese paper-cutting in digital protection for intangible cultural heritage under virtual reality technology. Heliyon 2024; 10(18): e38073.

Yang

Gao

Zhou

, et al. A study on the influence mechanism of consumer values on green consumption willingness under the background of Chinese culture. Fresen Environ Bull 2021; 30(4): 3632–3638.

10.

, et al. The inactivated ISKNV-I vaccine confers highly effective cross-protection against epidemic RSIV-I and RSIV-II from cultured spotted sea bass Lateolabrax maculatus. Microbiol Spectr 2023; 11(3): e0449522.

11.

Simonin

Roullier-Gall

Ballester

, et al. Bio-protection as an alternative to sulphites: impact on chemical and microbial characteristics of red wines. Front Microbiol 2020; 11: 1308.

12.

Wei

Cheng

Liao

, et al. Composition and function of the skin microbiota were altered of red leg syndrome in cultured bullfrog (Rana catesbeiana). Aquac Rep 2023; 29: 101487.

13.

Wang

Zhang

Zhao

, et al. Phenol red hydrogel as pH indicator with protection against nanoceria degradation. J Sci Adv Mater Devices 2023; 8(4): 100644.