Chinese opera culture promotion system and method
By collecting and processing opera data, and using generative adversarial networks and recurrent neural networks, personalized promotion of opera culture is achieved, the problem of single traditional communication channels is solved, and user experience and data quality is improved.
Patent Information
- Application Number
- CN202510479708.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-17
- Publication Date
- 2025-07-25
AI Technical Summary
Traditional opera performance forms and communication channels cannot meet the personalized needs of modern audiences, resulting in a single opera culture communication model, lack of integration with modern society and matching user preferences.
By collecting opera performance videos, scripts and music data, using generative adversarial networks and recurrent neural networks for data processing and feature extraction, establishing a user feedback mechanism, optimizing opera promotion content, and realizing personalized promotion.
It enriches the diversity and readability of opera data, improves data quality and model expression capabilities, and enhances user experience and promotion effects.
Smart Images
Figure CN120373556A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of cultural dissemination, and more specifically, to a system and method for promoting traditional Chinese opera culture. Background Art
[0002] As an important part of traditional Chinese culture, traditional Chinese opera still has important historical, cultural and artistic values in today's society. However, with the development of social transformation and cultural diversity, the inheritance and development of traditional Chinese opera culture face many challenges in modern society.
[0003] With the rapid development of information technology, the Internet has become an important platform for people to obtain information, exchange views and cultural dissemination. The traditional forms of opera performance and dissemination channels cannot fully meet the needs of modern audiences. There is a lack of dialogue and difficulty in integrating between the tradition and classic of opera art and modern times. The current opera performance methods cannot be personalized promoted according to user preferences, resulting in a too single dissemination model and development path of traditional Chinese opera culture. Summary of the Invention
[0004] In view of the technical problems existing in the prior art, the present invention provides a system and method for promoting traditional Chinese opera culture to solve the problems raised in the above background art.
[0005] The technical solution of the present invention to solve the above technical problems is as follows: A system for promoting traditional Chinese opera culture includes a data acquisition module, a data processing module, a model training module and a promotion and release module;
[0006] Data acquisition module: Collect opera performance videos, opera literary works and opera music, connect to the data interaction interfaces of libraries, literary websites and script publishing houses, and integrate them into the opera data set;
[0007] Data processing module: Add data annotations and divide the opera data set into a training set, a validation set and a test set, parse the opera data set into a readable format, and preprocess the opera data set;
[0008] Model training module: Use a generative adversarial network to establish a generator network and a discriminator network, create a sliding window for filtering the video-related data in the opera data set, perform color space conversion and extract features, convert the music-related data in the opera data set to the frequency domain and extract frequency domain features, and build a recurrent neural network model to output personalized opera promotion content according to the user preference layer and the test set;
[0009] Promotion and release module: Connect to the user feedback center to collect user feedback data, associate the user feedback data with the opera promotion content and establish a mapping relationship, establish a linear regression model, and optimize and feedback to the recurrent neural network model to optimize the opera promotion content;
[0010] In a preferred embodiment, the data acquisition module determines the channels for obtaining opera culture data through public cultural institutions, art troupes, opera performance platforms, and online video websites, collects opera performance videos covering different regions and genres in different forms, connects to the data interaction interfaces of libraries, literary websites, and script publishing houses and collects opera script literary works, and integrates the collected data into the opera dataset.
[0011] In a preferred embodiment, the data processing module adds data annotations through word segmentation, stop word removal, part-of-speech tagging, and keyword extraction, divides the opera dataset into a training set, a validation set, and a test set, where 70% of the feature representations are used as the training set, 20% of the feature representations are used as the validation set, and 10% of the feature representations are used as the test set, parses the opera dataset into a readable format, checks for missing values in the opera dataset, selects and deletes the records of the opera dataset corresponding to the missing values, and fills in the corresponding values using the self-similar data of the opera dataset, detects duplicate records in the opera dataset, and selects and deletes the duplicate values.
[0012] In a preferred embodiment, the model training module uses a generative adversarial network to establish a generator network and a discriminator network for various types of data in the opera dataset, and continuously improves the generator network to generate synthetic category data and continuously improves the discriminator network's ability to distinguish the authenticity of data by alternately training the generator network and the discriminator network until the synthetic category data reaches a quality level and stops training.
[0013] Further, a sliding window for video-related data is created in the opera dataset for pixel-by-pixel movement. For each pixel point within each sliding window, calculate the average value of the neighboring pixels around the pixel point, and use this average value as the new value of the current pixel point. The specific calculation formula is:
[0014]
[0015] where l′(x,y) represents the pixel value of the filtered video-related data, l(a,b) represents the coordinates of the neighboring pixels, and k 2 represents the size of the filter. Convert the video-related data in the opera dataset from the RGB color space to the grayscale space, and use grayscale stretching to adjust the brightness and contrast. Extract shape features through edge detection, extract color features through color histograms, extract texture features through wavelet transforms, provide texture transformation information of the video-related data in the opera dataset at different frequencies and times, and convert the music-related data in the opera dataset to the frequency domain and extract frequency domain features by performing spectral analysis on the corresponding electrical signals of the music-related data in the opera dataset.
[0016] Furthermore, an LSTM unit is selected to build a recurrent neural network model. By introducing a gating mechanism to control the accumulation speed of features, the input gate is divided into a user preference layer and an opera dataset input layer. According to the training set of the opera dataset input at the current time step and the hidden state at the previous time step, the activation value of the input gate is calculated through the Sigmoid activation function. The specific formula is:
[0017] R l = α(W l *[I t-1 , x t + b l )
[0018] Among them, R l represents the activation value of the input gate, α() represents the Sigmoid activation function, W l represents the weight matrix, I t-1 represents the hidden state at the previous time step, x t represents the training set of the opera dataset input at the current time step, b l represents the bias term. The output gate selects a fully connected layer and calculates the activation value of the output gate through the Sigmoid activation function according to the training set of the opera dataset input at the current time step, the hidden state at the previous time step, and the dataset cell state at the current time step. The specific formula is:
[0019] R o = α(W o *[I t-1 , x t + b o )
[0020] Among them, R o represents the activation value of the forget gate, α() represents the Sigmoid activation function, W o represents the weight matrix, I t-1 represents the hidden state at the previous time step, x t represents the training set of the opera dataset input at the current time step, b o represents the bias term. Through the fully connected layer, personalized opera promotion content is output according to the user preference layer and the test set.
[0021] In a preferred embodiment, the promotion and release module is connected to the user feedback center to collect user feedback data and track and monitor the user click-through rate, viewing duration, and interaction behavior. The user feedback data is associated with the opera promotion content to establish a mapping relationship, and a linear regression model is established. The specific formula is:
[0022] Y = β0 + β1X1 + β2X2 + … + β n X n + ε
[0023] Among them, Y represents user feedback data, X1, X2, … X n represents the nth opera promotion content, β1, β2 … β n represents the influence of user feedback data on the nth opera promotion content, β0 represents the intercept, and ε represents the error term. By finding the best influence to minimize the correlation prediction error between user feedback data and opera promotion content, and optimizing the feedback to the recurrent neural network model for optimizing opera promotion content.
[0024] In a preferred embodiment, it specifically includes the following steps:
[0025] S101. Collect opera performance videos covering different regions and genres in different forms, collect music of different opera repertoires from different sources, connect the data interaction interfaces of libraries, literary websites, and script publishing houses and collect opera script literary works, and integrate them into the opera dataset;
[0026] S102. Add data annotations through word segmentation, stop word removal, part-of-speech tagging, and keyword extraction, parse the opera dataset into a readable format, and preprocess the opera dataset;
[0027] S103. Use a generative adversarial network to establish a generator network and a discriminator network, filter, perform color space conversion and extract features for video-related data in the opera dataset, convert music-related data in the opera dataset to the frequency domain and extract frequency domain features, and build a recurrent neural network model to output personalized opera promotion content;
[0028] S104. Connect to the user feedback center to collect user feedback data, establish a linear regression model to minimize the correlation prediction error between user feedback data and opera promotion content by finding the best influence, and optimize the feedback to the recurrent neural network model for optimizing opera promotion content;
[0029] The beneficial effects of the present invention are as follows: By collecting data through multiple channels, the comprehensiveness and diversity of opera-related data are ensured, which helps to demonstrate the rich diversity of opera culture, enriches the sources and types of opera. By adding data annotations, the readability and usability of the opera dataset are improved. Through the preprocessing of the opera dataset, the integrity and accuracy of the dataset are guaranteed, and the quality and credibility of the dataset are improved. By using a generative adversarial network to balance the opera dataset, it helps to generate more realistic opera-related data and improve the data quality. By building a recurrent neural network model with LSTM units and introducing a gating mechanism, it helps to better capture the temporal relationships and sequence information between opera datasets, and improve the expressive ability and generation performance of the model. Associating user feedback data with opera promotion content and establishing a mapping relationship helps to deeply understand user behavior and preferences, and provides a basis for personalized customization of promotion content. Feeding the optimized feedback results back to the recurrent neural network model and using a closed-loop feedback mechanism to continuously optimize the promotion strategy can improve the user experience and promotion effect. BRIEF DESCRIPTION OF THE DRAWINGS
[0030] Figure 1 is a flowchart of the method of the present invention;
[0031] Figure 2 is a block diagram of the system structure of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0032] Next, the technical solutions in the embodiments of the present application will be clearly and completely described with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts shall fall within the protection scope of the present application.
[0033] In the description of the present application, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of the described features. In the description of the present application, "a plurality" means two or more unless otherwise specifically defined.
[0034] In the description of this application, the term "for example" is used to mean "serving as an example, illustration, or explanation". Any embodiment described in this application as "for example" is not necessarily to be construed as more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to implement and use the present invention. In the following description, details are set forth for purposes of explanation. It should be understood that those of ordinary skill in the art can recognize that the present invention can be implemented without the use of these specific details. In other instances, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is to be accorded the widest scope consistent with the principles and features disclosed in this application.
[0035] Embodiment 1
[0036] This embodiment provides a Figure 1 shown opera culture promotion system, specifically including: a data acquisition module, a data processing module, a model training module, and a promotion and release module;
[0037] Data acquisition module: Collect opera performance videos, opera script literary works, and opera music, connect to the data interaction interfaces of libraries, literary websites, and script publishing houses, and integrate them into the opera dataset;
[0038] Data processing module: Add data annotations and divide the opera dataset into a training set, a validation set, and a test set, parse the opera dataset into a readable format, and preprocess the opera dataset;
[0039] Model training module: Use a generative adversarial network to establish a generator network and a discriminator network, create a sliding window for filtering the video-related data in the opera dataset, perform color space conversion and extract features, convert the music-related data in the opera dataset to the frequency domain and extract frequency domain features, and build a recurrent neural network model to output personalized opera promotion content according to the user preference layer and the test set;
[0040] Promotion and release module: Connect to the user feedback center to collect user feedback data, associate the user feedback data with the opera promotion content and establish a mapping relationship, establish a linear regression model, and optimize the feedback to the recurrent neural network model for optimizing the opera promotion content.
[0041] Embodiment 2
[0042] This embodiment provides a Figure 2 shown opera culture promotion method, specifically including the following steps:
[0043] S101. Collect opera performance videos covering different regions and genres in different forms, collect music of different opera repertoires from different sources, connect to the data interaction interfaces of libraries, literary websites, and script publishing houses, collect opera script literary works, and integrate them into the opera dataset;
[0044] Further, determine the channels for obtaining opera culture data through public cultural institutions, art groups, opera performance platforms, and online video websites, and collect opera performance videos covering different regions and genres in different forms. The different forms include classic performances, modern adaptations, and stage recordings. The different regions and genres include Peking Opera, Pingju Opera, Huangmei Opera, and Yue Opera. Collect music of different opera repertoires from different sources. The different sources include live performance recordings, music albums, and film scores. The music of different opera repertoires includes arias and music pieces. Connect to the data interaction interfaces of libraries, literary websites, and script publishing houses, and collect opera script literary works, including plot summaries and character lines. Integrate the collected data into the opera dataset.
[0045] S102. Add data annotations through word segmentation, stop word removal, part-of-speech tagging, and keyword extraction, parse the opera dataset into a readable format, and preprocess the opera dataset;
[0046] Further, add data annotations through word segmentation, stop word removal, part-of-speech tagging, and keyword extraction, including video type, actor information, repertoire name, music genre, and script content. Divide the opera dataset into a training set, a validation set, and a test set. Among them, 70% of the feature representations are used as the training set, 20% of the feature representations are used as the validation set, and 10% of the feature representations are used as the test set. Parse the opera dataset into a readable format, including converting video-related data into NumPy arrays and text-related data into word vector representations. Check whether there are missing values in the opera dataset, select to delete the records corresponding to the missing values in the opera dataset, and use the self-similar data in the opera dataset for corresponding filling. Detect duplicate records in the opera dataset and select to delete the duplicate values.
[0047] S103. Use a generative adversarial network to establish a generator network and a discriminator network, filter, perform color space conversion, and extract features for video-related data in the opera dataset, convert music-related data in the opera dataset to the frequency domain and extract frequency domain features, and build a recurrent neural network model to output personalized opera promotion content;
[0048] Further, for various types of data in the opera dataset, a generator network and a discriminator network are established using a generative adversarial network. The generator network accepts random noise as input and outputs synthetic category data. The discriminator network accepts real category data and synthetic category data as input and attempts to distinguish the authenticity of the data. The generator network and the discriminator network are alternately trained to continuously improve the synthetic category data generated by the generator network and continuously enhance the ability of the discriminator network to distinguish the authenticity of the data until the synthetic category data reaches a quality level and the training stops.
[0049] Further, a sliding window for video-related data is created in the opera dataset for pixel-by-pixel movement. For each pixel point within each sliding window, the average value of the neighboring pixels around the pixel point is calculated, and this average value is used as the new value of the current pixel point. The specific calculation formula is as follows:
[0050]
[0051] where l′(x,y) represents the pixel value of the filtered video-related data, l(a,b) represents the coordinates of the neighboring pixels, and k 2 represents the size of the filter. The video-related data in the opera dataset is converted from the RGB color space to the grayscale space, and the brightness and contrast are adjusted using grayscale stretching. Shape features are extracted through edge detection, including the feature information of the contour boundaries in the video-related data in the opera dataset. Color features are extracted through a color histogram, including the feature information of the color distribution in the video-related data in the opera dataset. Texture features are extracted through wavelet transform, including the feature information of the texture in the video-related data in the opera dataset. The wavelet transform is a continuous wavelet transform. The video-related data in the opera dataset is continuously transformed using different scales and translation parameters to provide texture transformation information of the video-related data in the opera dataset at different frequencies and times. By performing spectral analysis on the electrical signals corresponding to the music-related data in the opera dataset, the music-related data in the opera dataset is converted to the frequency domain and frequency domain features are extracted. The bandwidth of the spectrum is obtained to describe the frequency distribution of the music-related data signals in the opera dataset, and the frequency components are obtained to describe the degree of uniformity of the energy distribution of the music-related data in the opera dataset.
[0052] Further, an LSTM unit is selected to build a recurrent neural network model. By introducing a gating mechanism to control the accumulation speed of features, where the gating mechanism includes an input gate, a forget gate, and an output gate. The input gate is divided into a user preference layer and an opera dataset input layer, and according to the training set of the opera dataset input at the current time step and the hidden state at the previous time step, the activation value of the input gate is calculated through the Sigmoid activation function. The specific formula is as follows:
[0053] R l =α(Wl *[I t-1 ,x t +b l )
[0054] Among them, R l represents the activation value of the input gate, α() represents the Sigmoid activation function, W l represents the weight matrix, I t-1 represents the hidden state of the previous time step, x t represents the training set of the opera dataset input at the current time step, b l represents the bias term. The output gate selects the fully connected layer and calculates the activation value of the output gate through the Sigmoid activation function according to the training set of the opera dataset input at the current time step, the hidden state of the previous time step, and the cell state of the dataset at the current time step. The specific formula is:
[0055] R o = α(W o *[I t-1 ,x t +b o )
[0056] Among them, R o represents the activation value of the forget gate, α() represents the Sigmoid activation function, W o represents the weight matrix, I t-1 represents the hidden state of the previous time step, x t represents the training set of the opera dataset input at the current time step, b o represents the bias term, and personalized opera promotion content is output through the fully connected layer according to the user preference layer and the test set.
[0057] S104. Connect to the user feedback center to collect user feedback data, establish a linear regression model to minimize the correlation prediction error between the user feedback data and the opera promotion content by finding the best influence, and optimize and feedback it to the recurrent neural network model for optimizing the opera promotion content;
[0058] Furthermore, connect to the user feedback center to collect user feedback data and track and monitor the user click volume, viewing duration, and interaction behavior, associate the user feedback data with the opera promotion content and establish a mapping relationship, including matching the user click volume with the title of the opera promotion content, matching the viewing duration with the description of the opera promotion content, and establish a linear regression model. The specific formula is:
[0059] Y = β0 + β1X1 + β2X2 + … + β n X n + ε
[0060] Among them, Y represents user feedback data, X1, X2, … X n represents the nth opera promotion content, β1, β2 … β n represents the influence of user feedback data on the nth opera promotion content, β0 represents the intercept, and ε represents the error term. By finding the best influence to minimize the correlation prediction error between user feedback data and opera promotion content, and optimizing the feedback to the recurrent neural network model for optimizing opera promotion content.
[0061] It should be noted that in the above embodiments, the descriptions of each embodiment have their own focuses. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.
[0062] Those skilled in the art should understand that the embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0063] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the flows and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.
[0064] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured product including an instruction device, and the instruction device implements the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.
[0065] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus, causing a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one process or a plurality of processes and / or blocks Figure 1 one process or a plurality of processes and / or blocks Figure 1 or steps for implementing the functions specified in one block or a plurality of blocks.
[0066] Although the preferred embodiments of the present invention have been described, additional changes and modifications can be made by those skilled in the art once they learn of the basic inventive concept. Therefore, the appended claims are intended to be construed to cover the preferred embodiments as well as all changes and modifications falling within the scope of the present invention.
[0067] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.
Claims
1. An opera culture promotion system, characterized in that, Specifically include: Including a data acquisition module, a data processing module, a model training module, and a promotion and release module; Data acquisition module: Collect opera performance videos, opera script literary works, and opera music, connect to the data interaction interfaces of libraries, literary websites, and script publishing houses, and integrate them into the opera dataset; Data processing module: Add data annotations and divide the opera dataset into a training set, a validation set, and a test set, parse the opera dataset into a readable format, and preprocess the opera dataset; Model training module: Use a generative adversarial network to establish a generator network and a discriminator network, create a sliding window for filtering the video-related data in the opera dataset, perform color space conversion and feature extraction, convert the music-related data in the opera dataset to the frequency domain and extract frequency domain features, and build a recurrent neural network model to output personalized opera promotion content according to the user preference layer and the test set; Promotion and release module: Connect to the user feedback center to collect user feedback data, associate the user feedback data with the opera promotion content and establish a mapping relationship, establish a linear regression model, and optimize and feedback to the recurrent neural network model for optimizing the opera promotion content.
2. The opera culture promotion system according to claim 1, wherein: The data acquisition module determines the channels for obtaining opera culture data through public cultural institutions, art groups, opera performance platforms, and online video websites, collects opera performance videos covering different regions and genres in different forms, connects to the data interaction interfaces of libraries, literary websites, and script publishing houses and collects opera script literary works, and integrates the collected data into the opera dataset.
3. A traditional opera culture promotion system according to claim 1, characterized in that: The data processing module adds data annotations through word segmentation, stop word removal, part-of-speech tagging, and keyword extraction, divides the opera dataset into a training set, a validation set, and a test set, parses the opera dataset into a readable format, checks for missing values in the opera dataset, selects to delete the records corresponding to the missing values in the opera dataset, and fills in the corresponding values using the self-similar data of the opera dataset, detects duplicate records in the opera dataset, and selects to delete the duplicate values.
4. A traditional opera culture promotion system according to claim 1, characterized in that: The model training module uses a generative adversarial network to establish a generator network and a discriminator network for various types of data in the opera dataset. Create a sliding window for video-related data in the opera dataset for pixel-by-pixel movement. For each pixel point within each sliding window, calculate the average value of the neighboring pixels around the pixel point, and use this average value as the new value of the current pixel point. Convert the video-related data in the opera dataset from the RGB color space to the grayscale space, and adjust the brightness and contrast using grayscale stretching. Extract shape features through edge detection, extract color features through color histograms, extract texture features through wavelet transforms, and provide texture transformation information of the video-related data in the opera dataset at different frequencies and times. Through spectral analysis of the electrical signals corresponding to the music-related data in the opera dataset, convert the music-related data in the opera dataset to the frequency domain and extract frequency domain features, and select LSTM units to build a recurrent neural network model.
5. The opera culture promotion system according to claim 4, characterized in that: The specific steps for establishing the generator network and the discriminator network are as follows: Continuously improve the generator network to generate synthetic category data and continuously enhance the discriminator network's ability to distinguish the authenticity of data by alternately training the generator network and the discriminator network until the synthetic category data reaches the quality level and then stop training.
6. The opera culture promotion system according to claim 4, characterized in that: The specific calculation formula for calculating the average value of the neighboring pixels around the pixel points of the video-related data is as follows: where l′(x,y) represents the pixel value of the filtered video-related data, l(a,b) represents the coordinates of the neighboring pixels, and k 2 represents the size of the filter.
7. The opera culture promotion system according to claim 4, characterized in that: The specific steps for selecting LSTM units to build a recurrent neural network model are as follows: Divide the input gate into a user preference layer and an input layer of the opera dataset, and calculate the activation value of the input gate through the Sigmoid activation function according to the training set of the opera dataset input at the current time step and the hidden state at the previous time step. The specific formula is as follows: R l = α(W l * [I t-1 , x t + b l ) where, R l represents the activation value of the input gate, α() represents the Sigmoid activation function, W l represents the weight matrix, I t-1 represents the hidden state of the previous time step, x t represents the training set of the opera data set input at the current time step, b l represents the bias term. The output gate selects the fully connected layer and calculates the activation value of the output gate through the Sigmoid activation function based on the training set of the opera data set input at the current time step, the hidden state of the previous time step, and the data set cell state at the current time step. The specific formula is as follows: R o = α(W o *[I t-1 ,x t +b o ) Among them, R o represents the activation value of the forget gate, α() represents the Sigmoid activation function, W o represents the weight matrix, I t-1 represents the hidden state at the previous time step, x t represents the training set of the opera data set input at the current time step, b o represents the bias term.
8. A traditional opera culture promotion system according to claim 1, characterized in that: The promotion and release module is connected to the user feedback center to collect user feedback data, track and monitor the user click-through rate, viewing duration, and interaction behaviors, associate the user feedback data with the opera promotion content and establish a mapping relationship, establish a linear regression model, minimize the prediction error of the association between the user feedback data and the opera promotion content by finding the best influence, and optimize and feedback it to the recurrent neural network model for optimizing the opera promotion content.
9. A traditional opera culture promotion system according to claim 8, characterized in that: The specific formula for the linear regression model is as follows: Y = β0 + β1X1 + β2X2 + … + β n X n + ε Among them, Y represents user feedback data, X1, X2, … X n represents the nth opera promotion content, β1, β2 … β n represents the influence of user feedback data on the nth opera promotion content, β0 represents the intercept, and ε represents the error term.
10. A method for promoting traditional opera culture is applied to a traditional opera culture promotion system as described in any one of claims 1-9, characterized in that: Specifically, it includes the following steps: S101. Collect opera performance videos covering different regions and genres in different forms, collect music of different opera repertoires from different sources, connect to the data interaction interfaces of libraries, literary websites, and script publishing houses and collect opera script literary works, and integrate them into the opera dataset; S102. Add data annotations through word segmentation, stop word removal, part-of-speech tagging, and keyword extraction, parse the opera dataset into a readable format, and preprocess the opera dataset; S103. Use a generative adversarial network to establish a generator network and a discriminator network, filter, perform color space conversion, and extract features for the video-related data in the opera dataset, convert the music-related data in the opera dataset to the frequency domain and extract frequency domain features, and build a recurrent neural network model to output personalized opera promotion content; S104. Connect to the user feedback center to collect user feedback data, establish a linear regression model, minimize the prediction error of the association between the user feedback data and the opera promotion content by finding the best influence, and optimize and feedback it to the recurrent neural network model for optimizing the opera promotion content.