Intelligent voice operation management system

Abstract

The invention discloses an intelligent voice operation management system which comprises a service system and an intelligent voice instruction system. The service system comprises a terminal device, and the intelligent voice instruction system comprises an audio collecting and processing module, a voice recognition model and a semantic analysis model. The intelligent voice instruction system analyzes the structured semantic analysis result and then feeds back the result to the service system for operation management of the project system, a user can quickly operate the project system through the voice instruction, manual operation is not needed, expansibility is high, voice operation is more natural and convenient, and the method is suitable for various scenes.

Description

Technical Field

[0001] The present invention relates to the field of speech recognition technology, and in particular to an intelligent speech operation management system. Background Art

[0002] Intelligent voice technology has been widely used in recent years, especially in smart homes, mobile devices and enterprise applications. These systems can perform various operations through voice commands, such as controlling home appliances, querying information, and navigating.

[0003] In the current project management system, it is common to use a mouse or touch screen to perform operations such as query, add, modify, and delete to meet daily needs. For example, similar operations are also performed during project demonstrations or product promotion. If intelligent voice recognition can be used to operate and manage the project system, this will make the operation easier and faster. Summary of the invention

[0004] In order to solve the deficiencies in the prior art, the present invention provides an intelligent voice operation management system, which obtains audio streams through terminal devices, analyzes the structured semantic analysis results through the intelligent voice command system, and then feeds back to the business system for operation management of the project system. Users can quickly operate the project system through voice commands without manual operation. It has strong scalability, and voice operation is more natural and convenient, and is suitable for various scenarios.

[0005] To achieve the above object, the present invention adopts the following technical solution:

[0006] The present invention provides an intelligent voice operation management system, comprising:

[0007] Business system; the business system includes a terminal device, which is used to obtain an audio stream in real time and transmit it to an intelligent voice command system, receive a structured semantic analysis result sent by the intelligent voice command system, and perform feedback operations according to the structured semantic analysis result;

[0008] Intelligent voice command system; the intelligent voice command system includes an audio acquisition and processing module, a speech recognition model and a semantic analysis model. The audio acquisition and processing module is used to receive and process the audio stream sent by the business system, the speech recognition model is used to identify the processed audio stream, and the semantic analysis model is used to analyze the structured semantic analysis results based on the speech recognition results and send them to the business system.

[0009] The intelligent voice operation management system of the present invention has a terminal device that is suitable for a variety of devices, such as mobile phones, interactive screens, gooseneck microphones, and other devices with voice collection functions. It is not limited to which project system it is in and has a very high degree of liberalization. The audio stream obtained by the terminal device passes through the intelligent voice command system and is recognized by the voice recognition model. The semantic analysis model then performs semantic analysis on the recognition result to obtain a structured semantic analysis result, which is then sent to the business system to execute related project system operations.

[0010] The intelligent voice operation management system of the present invention obtains audio streams through terminal devices, and after the intelligent voice command system analyzes the structured semantic analysis results, it feeds back to the business system for operation management of the project system. Users can quickly operate the project system through voice commands without manual operation. It has strong scalability, and voice operation is more natural and convenient, which is suitable for various scenarios.

[0011] In a further technical solution, it also includes:

[0012] An interactive keyword configuration system; the interactive keyword configuration system includes an interactive keyword visualization configuration module, an interactive keyword real-time update module and a model real-time online update module. The interactive keyword visualization configuration module is used to visualize the keywords in the interaction, the interactive keyword real-time update module is used to update the configured interactive keywords in real time, and the model real-time upper limit update system is used to send the updated interactive keywords to the semantic analysis model of the intelligent voice command system to update the model.

[0013] The configuration of interactive keywords is of great significance to voice recognition and semantic analysis, which greatly improves the accuracy of semantic analysis. At the same time, the visual configuration greatly facilitates user operation and supports personalized configuration with higher freedom and accuracy.

[0014] In a further technical solution, it also includes:

[0015] A project system interface module, wherein the project system interface module is used to interact with the project system and execute operations specified by the user in the structured semantic analysis results.

[0016] By configuring the project system interface module and connecting with the project system, the structured semantic analysis results can be directly transmitted to the project system for execution, which is more efficient.

[0017] In a further technical solution, it also includes:

[0018] An intelligent supervision module is used to intelligently monitor structured semantic analysis results and analyze data for risk prevention and control.

[0019] The data is controlled and controlled through the intelligent supervision module, making the system more secure.

[0020] In a further technical solution, the audio acquisition processing module includes a noise suppression component, an echo cancellation component and a sound source localization component.

[0021] By preprocessing with the noise suppression component, the echo cancellation component and the sound source localization component, the preprocessing effect of the audio stream is better and the semantic analysis is more accurate.

[0022] In a further technical solution, the speech recognition model uses a natural language processing algorithm to process speech, and is trained based on a GPT model to quickly retrieve and recognize voice commands, and directly maps speech signals to interactive keywords through a sequence-to-sequence model, an attention mechanism, and a Transformer architecture.

[0023] Through a variety of algorithms and architectures, speech recognition is more accurate.

[0024] In a further technical solution, the semantic analysis model uses a small amount of labeled data and a large amount of unlabeled data through a semi-supervised learning algorithm to improve the performance of the model, uses the DQN reinforcement learning algorithm to learn the optimal strategy through trial and error in the environment, and finally uses a deep learning algorithm to perform a large amount of model training to accurately parse the structured semantic analysis results.

[0025] Improve the accuracy of structured semantic analysis results through a variety of algorithms.

[0026] The beneficial effects are:

[0027] 1. The intelligent voice operation management system of the present invention obtains the audio stream through the terminal device, and the intelligent voice command system analyzes the structured semantic analysis results and then feeds them back to the business system for operation management of the project system. Users can quickly operate the project system through voice commands without manual operation. It has strong scalability, and voice operation is more natural and convenient, which is suitable for various scenarios.

[0028] 2. The configuration of interactive keywords is of great significance to voice recognition and semantic analysis, which greatly improves the accuracy of semantic analysis. At the same time, the visual configuration greatly facilitates user operation and supports personalized configuration with higher freedom and accuracy.

[0029] 3. By configuring the project system interface module and connecting with the project system, the structured semantic analysis results can be directly transmitted to the project system for execution, which is more efficient.

[0030] 4. The data is controlled and controlled through the intelligent supervision module, making the system more secure.

[0031] 5. Preprocessing is performed through noise suppression components, echo cancellation components and sound source localization components, which can achieve better preprocessing effect on audio stream and more accurate semantic analysis.

[0032] 6. Through multiple algorithms and architectures, speech recognition is more accurate.

[0033] 7. Improve the accuracy of structured semantic analysis results through a variety of algorithms. DETAILED DESCRIPTION

[0034] The present invention will be further described below:

[0035] Example:

[0036] An intelligent voice operation management system, comprising:

[0037] Business system; the business system includes a terminal device, which is used to obtain an audio stream in real time and transmit it to an intelligent voice command system, receive a structured semantic analysis result sent by the intelligent voice command system, and perform feedback operations according to the structured semantic analysis result;

[0038] Intelligent voice command system; the intelligent voice command system includes an audio acquisition and processing module, a speech recognition model and a semantic analysis model. The audio acquisition and processing module is used to receive and process the audio stream sent by the business system, the speech recognition model is used to identify the processed audio stream, and the semantic analysis model is used to analyze the structured semantic analysis results based on the speech recognition results and send them to the business system.

[0039] The intelligent voice operation management system of the present invention has a terminal device that is suitable for a variety of devices, such as mobile phones, interactive screens, gooseneck microphones, and other devices with voice collection functions. It is not limited to which project system it is in and has a very high degree of liberalization. The audio stream obtained by the terminal device passes through the intelligent voice command system and is recognized by the voice recognition model. The semantic analysis model then performs semantic analysis on the recognition result to obtain a structured semantic analysis result, which is then sent to the business system to execute related project system operations.

[0040] The intelligent voice operation management system of the present invention obtains audio streams through terminal devices, and after the intelligent voice command system analyzes the structured semantic analysis results, it feeds back to the business system for operation management of the project system. Users can quickly operate the project system through voice commands without manual operation. It has strong scalability, and voice operation is more natural and convenient, which is suitable for various scenarios.

[0041] In another embodiment, it also includes:

[0042] An interactive keyword configuration system; the interactive keyword configuration system includes an interactive keyword visualization configuration module, an interactive keyword real-time update module and a model real-time online update module. The interactive keyword visualization configuration module is used to visualize the keywords in the interaction, the interactive keyword real-time update module is used to update the configured interactive keywords in real time, and the model real-time upper limit update system is used to send the updated interactive keywords to the semantic analysis model of the intelligent voice command system to update the model.

[0043] The configuration of interactive keywords is of great significance to voice recognition and semantic analysis, which greatly improves the accuracy of semantic analysis. At the same time, the visual configuration greatly facilitates user operation and supports personalized configuration with higher freedom and accuracy.

[0044] In another embodiment, it also includes:

[0045] A project system interface module, wherein the project system interface module is used to interact with the project system and execute operations specified by the user in the structured semantic analysis results.

[0046] By configuring the project system interface module and connecting with the project system, the structured semantic analysis results can be directly transmitted to the project system for execution, which is more efficient.

[0047] In another embodiment, it also includes:

[0048] An intelligent supervision module is used to intelligently monitor structured semantic analysis results and analyze data for risk prevention and control.

[0049] The data is controlled and controlled through the intelligent supervision module, making the system more secure.

[0050] In another embodiment, the audio acquisition processing module includes a noise suppression component, an echo cancellation component and a sound source localization component.

[0051] By preprocessing with the noise suppression component, the echo cancellation component and the sound source localization component, the preprocessing effect of the audio stream is better and the semantic analysis is more accurate.

[0052] In another embodiment, the speech recognition model uses a natural language processing algorithm to process speech, is trained based on a GPT model to quickly retrieve and recognize voice commands, and directly maps speech signals to interactive keywords through a sequence-to-sequence model, an attention mechanism, and a Transformer architecture.

[0053] Through a variety of algorithms and architectures, speech recognition is more accurate.

[0054] In another embodiment, the semantic analysis model uses a small amount of labeled data and a large amount of unlabeled data through a semi-supervised learning algorithm to improve the performance of the model, uses a DQN reinforcement learning algorithm to learn the optimal strategy through trial and error in the environment, and finally uses a deep learning algorithm to perform a large amount of model training to accurately parse the structured semantic analysis results.

[0055] Improve the accuracy of structured semantic analysis results through a variety of algorithms.

[0056] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. An intelligent voice operation management system, characterized in that: include: Business system; the business system includes a terminal device, which is used to obtain an audio stream in real time and transmit it to an intelligent voice command system, receive a structured semantic analysis result sent by the intelligent voice command system, and perform feedback operations according to the structured semantic analysis result; Intelligent voice command system; the intelligent voice command system includes an audio acquisition and processing module, a speech recognition model and a semantic analysis model. The audio acquisition and processing module is used to receive and process the audio stream sent by the business system, the speech recognition model is used to identify the processed audio stream, and the semantic analysis model is used to analyze the structured semantic analysis results based on the speech recognition results and send them to the business system.

2. The intelligent voice operation management system according to claim 1, characterized in that: Also includes: An interactive keyword configuration system; the interactive keyword configuration system includes an interactive keyword visualization configuration module, an interactive keyword real-time update module and a model real-time online update module. The interactive keyword visualization configuration module is used to visualize the keywords in the interaction, the interactive keyword real-time update module is used to update the configured interactive keywords in real time, and the model real-time upper limit update system is used to send the updated interactive keywords to the semantic analysis model of the intelligent voice command system to update the model.

3. The intelligent voice operation management system according to claim 1, characterized in that: Also includes: A project system interface module, wherein the project system interface module is used to interact with the project system and execute operations specified by the user in the structured semantic analysis results.

4. The intelligent voice operation management system according to claim 1, characterized in that: Also includes: An intelligent supervision module is used to intelligently monitor structured semantic analysis results and analyze data for risk prevention and control.

5. The intelligent voice operation management system according to claim 1, characterized in that: The audio collection and processing module includes a noise suppression component, an echo cancellation component and a sound source localization component.

6. The intelligent voice operation management system according to claim 2, characterized in that: The speech recognition model uses a natural language processing algorithm to process speech and is trained based on a GPT model to quickly retrieve and recognize voice commands. It directly maps speech signals to interactive keywords through a sequence-to-sequence model, an attention mechanism, and a Transformer architecture.

7. The intelligent voice operation management system according to claim 1, characterized in that: The semantic analysis model uses a small amount of labeled data and a large amount of unlabeled data through a semi-supervised learning algorithm to improve the performance of the model, uses the DQN reinforcement learning algorithm to learn the optimal strategy through trial and error in the environment, and finally uses a deep learning algorithm to perform a large amount of model training to accurately parse the structured semantic analysis results.