JP2025059012A - system - Google Patents

Abstract

To provide a system that analyzes voice data, automatically generates appropriate responses, and converts them into voice.SOLUTION: A system according to an embodiment includes an analysis unit, a generation unit, and a voice conversion unit. The analysis unit analyzes voice data. The generation unit generates a response based on the data analyzed by the analysis unit. The voice conversion unit converts the response generated by the generation unit into voice.SELECTED DRAWING: Figure 1

Description

The technology disclosed herein relates to a system.

Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including a description of the chatbot character and an associated instruction sentence, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance.

JP 2022-180282 A

Conventional technology does not automate the process of analyzing voice data and generating appropriate responses and converting them into voice, leaving room for improvement.

The system according to the embodiment aims to analyze voice data and automatically generate and voice appropriate responses.

The system according to the embodiment includes an analysis unit, a generation unit, and a voice conversion unit. The analysis unit analyzes the voice data. The generation unit generates a response based on the data analyzed by the analysis unit. The voice conversion unit voices the response generated by the generation unit.

The system according to the embodiment can analyze voice data and automatically generate and voice appropriate responses.

1 is a conceptual diagram showing an example of a configuration of a data processing system according to a first embodiment. 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. FIG. FIG. 11 is a conceptual diagram showing an example of a configuration of a data processing system according to a second embodiment. FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. FIG. 13 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. FIG. 13 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. FIG. 13 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. FIG. 13 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. 1 shows an emotion map onto which multiple emotions are mapped. 1 shows an emotion map onto which multiple emotions are mapped.

Below, an example of an embodiment of a system related to the technology disclosed herein is described with reference to the attached drawings.

First, let us explain the terminology used in the following explanation.

In the following embodiments, the signed processor (hereinafter simply referred to as the "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. The processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), or a TPU (Tensor Processing Unit).

In the following embodiments, a signed random access memory (RAM) is a memory in which information is temporarily stored and is used as a working memory by the processor.

In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs and various parameters. Examples of non-volatile storage devices include flash memory (Solid State Drive (SSD)), magnetic disks (e.g., hard disks), and magnetic tapes.

In the following embodiments, a communication I/F (Interface) with a code is an interface including a communication processor and an antenna. The communication I/F controls communication between multiple computers. Examples of communication standards applied to the communication I/F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), and Bluetooth (registered trademark).

In the following embodiments, "A and/or B" is synonymous with "at least one of A and B." In other words, "A and/or B" means that it may be only A, only B, or a combination of A and B. In addition, in this specification, the same concept as "A and/or B" is also applied when three or more things are expressed by connecting them with "and/or."

[First embodiment]
FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

As shown in FIG. 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

The data processing device 12 includes a computer 22, a database 24, and a communication I/F 26. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I/F 26 are also connected to the bus 34. The communication I/F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and/or a LAN (Local Area Network).

The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I/F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

The reception device 38 includes a touch panel 38A and a microphone 38B, and receives user input. The touch panel 38A detects contact with an indicator (e.g., a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 (see FIG. 2) acquires the data indicating the user input.

The output device 40 includes a display 40A and a speaker 40B, and presents data to the user by outputting the data in a form of expression that the user can perceive (e.g., voice and/or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs voice according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system including a lens, an aperture, and a shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

The communication I/F 44 is connected to the network 54. The communication I/Fs 44 and 26 are responsible for transmitting and receiving various types of information between the processor 46 and the processor 28 via the network 54.

Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

As shown in FIG. 2, in the data processing device 12, specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotion, including estimation and prediction of the user's emotion, but is not limited to such examples. Furthermore, the estimation and prediction of emotion also includes, for example, analysis of emotions.

In the smart device 14, the specific processing is performed by the processor 46. The storage 50 stores a specific processing program 60. The specific processing program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the specific processing program 60 from the storage 50 and executes the specific processing program 60 on the RAM 48. The specific processing is realized by the processor 46 operating as the control unit 46A in accordance with the specific processing program 60 executed on the RAM 48. The smart device 14 has a data generation model and an emotion identification model similar to the data generation model 58 and the emotion identification model 59, and can also use these models to perform processing similar to that of the specific processing unit 290.

Note that a device other than the data processing device 12 may have the data generation model 58. For example, a server device (e.g., a generation server) may have the data generation model 58. In this case, the data processing device 12 obtains a processing result (such as a prediction result) using the data generation model 58 by communicating with the server device having the data generation model 58. The data processing device 12 may also be a server device, or a terminal device owned by a user (e.g., a mobile phone, a robot, a home appliance, etc.). Next, an example of processing by the data processing system 10 according to the first embodiment will be described.

(Example 1)
The customer response automation system according to the embodiment of the present invention is a system that uses large-scale voice data to create a voice tempo and intonation model, and combines it with a generation AI to fully automatically respond to inquiries from customers and the like. This system first uses large-scale voice data to model the voice tempo and intonation. This model learns the rhythm and intonation of natural conversation through analysis of voice data. Next, a response to the customer's inquiry is generated using the generation AI. This generation AI has been fine-tuned in advance and has knowledge of specific business and services. Furthermore, the response content generated by the generation AI is converted into voice using the voice tempo and intonation model. As a result, the text-based response generated by the generation AI is provided to the customer as a natural voice. For example, a quick and accurate response to customer inquiries is possible. In addition, by performing fine tuning, the response content of the generation AI can be made closer to a more accurate one. In this way, the present invention combines large-scale voice data and generation AI to realize automation of customer responses, thereby improving business efficiency and customer satisfaction. As a result, the customer response automation system is able to respond quickly and accurately to inquiries from customers.

The customer support automation system according to the embodiment includes an analysis unit, a generation unit, and a voice conversion unit. The analysis unit analyzes voice data. The analysis unit converts voice data into text data, for example, using voice recognition technology. The analysis unit can also analyze the content of the voice data using natural language processing technology. For example, the analysis unit analyzes the phonemes and phonology of the voice data and models the tempo and intonation of the voice. The generation unit uses a generation AI to generate a response based on the data analyzed by the analysis unit. The generation unit generates a response, for example, using a text generation AI (for example, LLM). The generation unit can also use a generation AI to generate a response that has knowledge about a specific business or service. For example, the generation unit generates an appropriate response to an inquiry about customer support. The voice conversion unit voices the response generated by the generation unit. The voice conversion unit converts text data into voice data, for example, using voice synthesis technology. The voice conversion unit can also provide the generated voice data to the customer. For example, the voice conversion unit provides the generated voice data to the customer via telephone or the Internet. This enables the customer response automation system according to the embodiment to respond quickly and accurately to customer inquiries. Some or all of the above-mentioned processes in the analysis unit, generation unit, and voice conversion unit may be performed, for example, using AI, or may be performed without using AI. For example, the analysis unit can input voice data to the AI and have the AI analyze the voice data. The generation unit can input data analyzed by the analysis unit to the AI and have the AI generate a response. The voice conversion unit can input the response generated by the generation unit to the AI and have the AI convert it into voice.

The analysis unit analyzes the voice data. The analysis unit converts the voice data into text data, for example, using voice recognition technology. Specifically, the voice recognition technology converts the voice signal into digital data, and analyzes the digital data to identify phonemes and phonology. This makes it possible to extract information contained in the voice data in text format. The analysis unit can also analyze the contents of the voice data using natural language processing technology. Natural language processing technology analyzes the grammatical structure and meaning of the text data, and performs understanding based on the context. For example, the analysis unit analyzes the phonemes and phonology of the voice data and models the tempo and intonation of the voice. This makes it possible to grasp the speaker's emotions and intentions more accurately. Furthermore, the analysis unit can also use filtering technology to remove background noise and echoes from the voice data. This makes it possible to improve the quality of the voice data and increase the accuracy of the analysis. The analysis unit combines these technologies to analyze the voice data with high accuracy and output it as text data. The analysis unit can share the results of the analysis of the voice data with other systems and departments, and can, for example, work with a customer support system or database to improve the efficiency of customer support.

The generation unit uses the generation AI to generate a response based on the data analyzed by the analysis unit. The generation unit generates a response using, for example, a text generation AI (e.g., LLM). Specifically, the generation AI receives text data provided by the analysis unit as input and generates an appropriate response based on the content. The generation AI has learned a large amount of text data and has the ability to understand grammar and context, so it can generate natural and fluent responses. The generation unit can also use the generation AI to generate responses with knowledge of specific business operations and services. For example, the generation unit generates an appropriate response to an inquiry about customer support. The generation AI has learned knowledge about specific business operations and services in advance and can also respond to specialized questions. Furthermore, the generation unit can evaluate the quality of the generated response and make corrections as necessary. For example, if the response generated by the generation AI is inappropriate, the generation unit reevaluates the content of the response and generates a more appropriate response. The generation unit can also store the generated response in a database and use it as a reference for future inquiries. This allows the generation unit to generate quick and accurate responses, improving the efficiency and quality of customer support.

The voice conversion unit converts the response generated by the generation unit into voice. The voice conversion unit converts text data into voice data, for example, using voice synthesis technology. Specifically, the voice synthesis technology receives text data as input and generates natural voice based on the content. The voice synthesis technology generates natural and easy-to-listen voice by analyzing combinations of phonemes and phonological elements and adding appropriate intonation and tempo. The voice conversion unit can also provide the generated voice data to customers. For example, the voice conversion unit provides the generated voice data to customers via telephone or the Internet. In the case of telephone, the voice conversion unit transmits the generated voice data to a telephone line in real time and responds directly to the customer. In the case of the Internet, the voice conversion unit distributes the generated voice data in a streaming format so that the customer can listen to the voice through a web browser or a dedicated application. Furthermore, the voice conversion unit can evaluate the quality of the generated voice data and make corrections as necessary. For example, if the intonation or tempo of the voice is unnatural, the voice conversion unit readjusts the voice synthesis technology to generate a more natural voice. The voice conversion unit can also store the generated voice data in a database and use it as a reference for future inquiries. This allows the voice conversion unit to provide fast and accurate voice responses, improving the efficiency and quality of customer service.

The generation unit can include an adjustment unit that performs fine tuning. The adjustment unit performs fine tuning of the generation AI. The adjustment unit improves the accuracy of the response by, for example, adjusting parameters of the generation AI. The adjustment unit can also select training data. For example, the adjustment unit selects data related to a specific business or service and uses it for training the generation AI. In this way, the generation unit can make the response content of the generation AI more accurate by performing fine tuning. Some or all of the above-mentioned processing in the adjustment unit may be performed, for example, using AI, or may be performed without using AI. For example, the adjustment unit can cause the AI to adjust parameters of the generation AI.

The voice conversion unit may include a providing unit that provides the generated voice to the customer. The providing unit provides the generated voice to the customer. The providing unit provides the voice to the customer, for example, via telephone. The providing unit may also provide the voice to the customer via the Internet. For example, the providing unit provides the voice through a website or a mobile app. In this way, by providing the generated voice to the customer, it becomes possible to respond in a natural voice. Some or all of the above-mentioned processing in the providing unit may be performed, for example, using AI, or may be performed without using AI. For example, the providing unit may input the generated voice data to AI and cause the AI to provide the voice.

The analysis unit can analyze multiple pieces of voice data and model the tempo and intonation of the voice. The analysis unit analyzes multiple pieces of voice data, such as telephone voices and recorded voices. The analysis unit analyzes the voice waveform and extracts rhythm patterns. For example, the analysis unit analyzes the phonemes and phonology of the voice data and models the tempo and intonation of the voice. In this way, the rhythm and intonation of natural conversation can be learned by analyzing large-scale voice data. Some or all of the above-mentioned processing in the analysis unit may be performed, for example, using AI, or may be performed without using AI. For example, the analysis unit can input the voice data to AI and have the AI analyze the voice data.

The generation unit can use a generation AI that has knowledge about a specific business or service. The generation unit uses a generation AI that has knowledge about a specific business or service, such as customer support or medical consultation. The generation unit uses the generation AI to generate an appropriate response to an inquiry about the specific business or service. For example, the generation unit generates an appropriate response to an inquiry about customer support. In this way, by using the generation AI that has knowledge about the specific business or service, a more appropriate response can be generated. Some or all of the above-mentioned processing in the generation unit may be performed, for example, using AI, or may be performed without using AI. For example, the generation unit can cause the AI to execute the generation AI that has knowledge about the specific business or service.

When analyzing voice data, the analysis unit can improve the analysis accuracy by taking into account a specific accent or dialect. For example, when analyzing voice data having an accent of a specific region, the analysis unit applies an accent model of that region. In addition, when analyzing voice data having a specific dialect, the analysis unit can also perform the analysis by taking into account the characteristics of the dialect. Furthermore, when analyzing voice data containing a mixture of multiple accents or dialects, the analysis unit can also perform the analysis by integrating the characteristics of each. In this way, by taking into account a specific accent or dialect, the analysis accuracy is improved. Some or all of the above-mentioned processing in the analysis unit may be performed, for example, using AI, or may be performed without using AI. For example, the analysis unit can input voice data having a specific accent or dialect to AI and have the AI perform the analysis.

The analysis unit can perform filtering to remove background noise when analyzing the voice data. For example, the analysis unit applies filtering to remove background noise before analyzing the voice data. The analysis unit can also perform filtering to remove noise in a specific frequency band. Furthermore, the analysis unit can also perform filtering to remove dynamically changing background noise in real time. This removes background noise, improving analysis accuracy. Some or all of the above-mentioned processing in the analysis unit may be performed using AI, for example, or may be performed without using AI. For example, the analysis unit can input the voice data to AI and cause the AI to remove background noise.

When analyzing the voice data, the analysis unit can adjust the analysis method based on the geographical location information of the user. For example, when the user is in a specific area, the analysis unit performs the analysis taking into account the accent and dialect of that area. In addition, when the user is moving, the analysis unit can also perform the analysis taking into account the accent and dialect of the destination area. Furthermore, when the user is in different areas, the analysis unit can also perform the analysis by integrating the characteristics of each area. In this way, the analysis method can be adjusted by taking into account the geographical location information of the user. Some or all of the above-mentioned processing in the analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the analysis unit can input the geographical location information of the user to AI and cause the AI to adjust the analysis method.

When analyzing the voice data, the analysis unit can analyze the user's social media activity and prioritize analysis of related voice data. For example, the analysis unit prioritizes analysis of voice data related to a specific topic from the user's social media activity. The analysis unit can also prioritize analysis of voice data related to a specific event from the user's social media activity. Furthermore, the analysis unit can also prioritize analysis of voice data related to a specific person from the user's social media activity. In this way, by analyzing the user's social media activity, related voice data can be prioritized. A part or all of the above-mentioned processing in the analysis unit may be performed, for example, using AI or may be performed without using AI. For example, the analysis unit can input the user's social media activity data to AI and cause AI to perform analysis to determine the priority order of related voice data.

When generating a response, the generation unit can adjust the level of detail of the response based on the importance of the inquiry content. For example, the generation unit generates a detailed response to an inquiry of high importance. The generation unit can also generate a concise response to an inquiry of low importance. Furthermore, the generation unit can dynamically adjust the level of detail of the response according to the importance. This allows for a more appropriate response by adjusting the level of detail of the response based on the importance of the inquiry content. Some or all of the above-mentioned processing in the generation unit may be performed using, for example, AI, or may be performed without using AI. For example, the generation unit can input the importance of the inquiry content to the AI and cause the AI to execute processing to adjust the level of detail of the response.

When generating a response, the generation unit can apply different generation algorithms depending on the category of the inquiry. For example, the generation unit applies a specialized generation algorithm to a technical inquiry. The generation unit can also apply a general-purpose generation algorithm to a general inquiry. Furthermore, the generation unit can apply a quick generation algorithm to an urgent inquiry. This allows for a more appropriate response by applying different generation algorithms depending on the inquiry category. Some or all of the above-mentioned processing in the generation unit may be performed using, for example, AI, or may be performed without using AI. For example, the generation unit can input the inquiry category to the AI and cause the AI to execute a process of determining the generation algorithm to be applied.

When generating a reply, the generation unit can determine the priority of the reply based on the time of submission of the inquiry. For example, the generation unit can generate a reply with priority to a recently submitted inquiry. The generation unit can also generate a reply with priority to an inquiry that has been unresolved for a long time. Furthermore, the generation unit can dynamically adjust the priority of the reply depending on the submission time. This allows for a faster response by determining the priority of the reply based on the submission time of the inquiry. Some or all of the above-mentioned processing in the generation unit may be performed using, for example, AI, or may be performed without using AI. For example, the generation unit can input the submission time of the inquiry to the AI and cause the AI to execute a process of determining the priority of the reply.

When generating a reply, the generation unit can adjust the order of replies based on the relevance of the inquiries. For example, the generation unit generates replies preferentially for inquiries with high relevance. The generation unit can also postpone generating replies for inquiries with low relevance. Furthermore, the generation unit can dynamically adjust the order of replies according to the relevance. This allows for a more appropriate reply by adjusting the order of replies based on the relevance of the inquiries. Some or all of the above-mentioned processing in the generation unit may be performed, for example, using AI, or may be performed without using AI. For example, the generation unit can input the relevance of the inquiries to the AI and cause the AI to execute processing to adjust the order of replies.

The voice conversion unit can perform voice filtering to improve the naturalness of the generated voice during voice conversion. For example, the voice conversion unit applies a noise reduction filter to the generated voice. The voice conversion unit can also apply an echo canceling filter to the generated voice. Furthermore, the voice conversion unit can apply a sound quality improvement filter to the generated voice. This improves the naturalness of the generated voice, making it possible to convert the voice into a more natural voice. Some or all of the above-mentioned processing in the voice conversion unit can be performed using, for example, AI, or can be performed without using AI. For example, the voice conversion unit can input the generated voice data to AI and have the AI perform voice filtering.

The voice conversion unit can improve the accuracy of voice conversion by taking into account a specific accent or dialect when generating voice. For example, when generating voice with an accent of a specific region, the voice conversion unit applies an accent model of that region. In addition, when generating voice with a specific dialect, the voice conversion unit can also perform voice conversion by taking into account the characteristics of the dialect. Furthermore, when generating voice in which multiple accents or dialects are mixed, the voice conversion unit can also perform voice conversion by integrating the characteristics of each. In this way, the accuracy of voice conversion is improved by taking into account a specific accent or dialect. Some or all of the above-mentioned processing in the voice conversion unit may be performed using, for example, AI, or may be performed without using AI. For example, the voice conversion unit can input voice data with a specific accent or dialect to AI and have the AI perform voice conversion.

The voice conversion unit can adjust the voice conversion method taking into account the geographical location information of the user when vocalizing. For example, when the user is in a specific area, the voice conversion unit performs voice conversion taking into account the accent or dialect of that area. In addition, when the user is moving, the voice conversion unit can also perform voice conversion taking into account the accent or dialect of the area to which the user is moving. Furthermore, when the user is in different areas, the voice conversion unit can also perform voice conversion by integrating the characteristics of each area. In this way, the voice conversion method can be adjusted by taking into account the geographical location information of the user. Part or all of the above-mentioned processing in the voice conversion unit may be performed using, for example, AI, or may be performed without using AI. For example, the voice conversion unit can input the geographical location information of the user to AI and cause the AI to adjust the voice conversion method.

When vocalizing, the vocalization unit can analyze the user's social media activity and vocalize related voice data preferentially. For example, the vocalization unit preferentially generates voice related to a specific topic from the user's social media activity. The vocalization unit can also preferentially generate voice related to a specific event from the user's social media activity. Furthermore, the vocalization unit can also preferentially generate voice related to a specific person from the user's social media activity. In this way, by analyzing the user's social media activity, related voice data can be preferentially vocalized. A part or all of the above-mentioned processing in the vocalization unit may be performed, for example, using AI or may be performed without using AI. For example, the vocalization unit inputs the user's social media activity data into AI and causes AI to perform vocalization that determines the priority of related voice data.

During fine tuning, the adjustment unit can optimize the generation algorithm by referring to past inquiry data. The adjustment unit, for example, analyzes past inquiry data and optimizes parameters of the generation algorithm. The adjustment unit can also extract specific patterns from past inquiry data and reflect them in the generation algorithm. Furthermore, the adjustment unit can improve the accuracy of the generation algorithm based on past inquiry data. This makes it possible to optimize the generation algorithm by referring to past inquiry data. Some or all of the above-mentioned processing in the adjustment unit may be performed, for example, using AI, or may be performed without using AI. For example, the adjustment unit can input past inquiry data to AI and cause AI to optimize the generation algorithm.

During fine tuning, the adjustment unit can weight the learning data based on the time of inquiry submission. For example, the adjustment unit weights recent inquiry data and reflects the weight in the generation algorithm. The adjustment unit can also weight inquiry data that has been unresolved for a long time and reflect the weight in the generation algorithm. Furthermore, the adjustment unit can dynamically adjust the weighting of the learning data according to the submission time. This allows for more appropriate adjustment by weighting the learning data based on the time of inquiry submission. Some or all of the above-mentioned processing in the adjustment unit may be performed using, for example, AI, or may be performed without using AI. For example, the adjustment unit can input the time of inquiry submission to AI and cause AI to perform weighting of the learning data.

When providing voice, the providing unit can select the optimal voice providing method by referring to the user's past inquiry history. For example, the providing unit selects the optimal voice providing method from the user's past inquiry history. The providing unit can also analyze the user's past inquiry history and select the voice providing method based on a specific pattern. Furthermore, the providing unit can dynamically adjust the voice providing method based on the user's past inquiry history. In this way, the optimal voice providing method can be selected by referring to the user's past inquiry history. A part or all of the above-mentioned processing in the providing unit may be performed, for example, using AI or may be performed without using AI. For example, the providing unit can input the user's past inquiry history to AI and cause AI to execute processing to select the optimal providing method.

When providing voice, the providing unit can select the optimal providing method by taking into account the device information of the user. For example, when the user is using a smartphone, the providing unit selects a voice providing method that matches the screen size. In addition, when the user is using a tablet, the providing unit can also select a voice providing method optimized for a large screen. Furthermore, when the user is using a smartwatch, the providing unit can also select a voice providing method that is simple and highly visible. In this way, the optimal voice providing method can be selected by taking into account the device information of the user. Some or all of the above-mentioned processing in the providing unit may be performed, for example, using AI, or may be performed without using AI. For example, the providing unit can input the device information of the user to the AI and cause the AI to execute a process of selecting the optimal providing method.

The system according to the embodiment is not limited to the above-mentioned example, and various modifications are possible, for example, as follows:

When analyzing the voice data, the analysis unit can improve the accuracy of the analysis by referring to the user's past inquiry history. For example, the analysis unit extracts specific patterns from the past inquiry history and reflects them in the analysis of the voice data. The analysis unit can also learn the user's speech tendencies based on the past inquiry history and improve the accuracy of the analysis. Furthermore, the analysis unit can perform analysis with knowledge of specific tasks or services by referring to the past inquiry history. In this way, by referring to the user's past inquiry history, the accuracy of the analysis is improved and more appropriate responses are possible.

When providing the generated voice, the voice generation unit can adjust the length of the voice taking into account the remaining battery level of the user's device. For example, when the battery level is low, the voice generation unit can provide a short voice that focuses on the main points. When the battery level is sufficient, the voice generation unit can also provide a voice that includes a detailed explanation. Furthermore, when the battery level is moderate, the voice generation unit can provide a voice of an appropriate length. This makes it possible to provide optimal voice by taking into account the remaining battery level of the user's device.

When analyzing the voice data, the analysis unit can monitor the user's speaking speed in real time and dynamically adjust the analysis method. For example, if the user's speaking speed is fast, the analysis unit can increase the analysis speed. Also, if the user's speaking speed is slow, the analysis unit can decrease the analysis speed. Furthermore, if the user's speaking speed fluctuates, the analysis speed can also be dynamically adjusted. This allows for more appropriate analysis by adjusting the analysis method according to the user's speaking speed.

When analyzing the voice data, the analysis unit can estimate the user's age group and apply an analysis method appropriate to the age group. For example, the analysis unit can perform an analysis that takes into account casual language for younger users. Also, for older users, the analysis unit can perform an analysis that takes into account polite language. Furthermore, the analysis can also take into account the frequency of use of specific words and phrases depending on the age group. In this way, by applying an analysis method appropriate to the user's age group, the analysis accuracy can be improved and more appropriate responses can be made.

When providing the generated voice, the voice generation unit can adjust the frequency band of the voice taking into account the user's hearing characteristics. For example, if the user has difficulty hearing high-pitched sounds, the voice generation unit can provide voice with emphasis on low-pitched sounds. Also, if the user has difficulty hearing low-pitched sounds, the voice generation unit can provide voice with emphasis on high-pitched sounds. Furthermore, specific frequency bands can be emphasized or suppressed depending on the user's hearing characteristics. This makes it possible to provide optimal voice by taking into account the user's hearing characteristics.

When generating a response, the generation unit can maintain consistency in the response by referring to the user's past inquiry history. For example, the generation unit generates a response that matches the content of the past inquiry. In addition, the generation unit can generate a response that reflects the user's preferences and tendencies based on the past inquiry history. Furthermore, by referring to the past inquiry history, it is possible to generate a response that is free of inconsistencies. In this way, by referring to the user's past inquiry history, consistency in the response can be maintained, enabling a more appropriate response.

The processing flow of Example 1 is briefly explained below.

Step 1: The analysis unit analyzes the voice data. For example, the analysis unit converts the voice data into text data using a voice recognition technique. The analysis unit can also analyze the contents of the voice data using a natural language processing technique. For example, the analysis unit analyzes the phonemes and phonology of the voice data and models the tempo and intonation of the voice.
Step 2: The generation unit uses the generation AI to generate a response based on the data analyzed by the analysis unit. The generation unit generates a response, for example, using a text generation AI (e.g., LLM). The generation unit can also use the generation AI to generate a response with knowledge about a specific business or service. For example, the generation unit generates an appropriate response to a customer support inquiry.
Step 3: The voice conversion unit converts the response generated by the generation unit into voice. The voice conversion unit converts the text data into voice data, for example, using a voice synthesis technique. The voice conversion unit can also provide the generated voice data to the customer. For example, the voice conversion unit provides the generated voice data to the customer over the telephone or the Internet.

(Example 2)
The customer response automation system according to the embodiment of the present invention is a system that uses large-scale voice data to create a voice tempo and intonation model, and combines it with a generation AI to fully automatically respond to inquiries from customers and the like. This system first uses large-scale voice data to model the voice tempo and intonation. This model learns the rhythm and intonation of natural conversation through analysis of voice data. Next, a response to the customer's inquiry is generated using the generation AI. This generation AI has been fine-tuned in advance and has knowledge of specific business and services. Furthermore, the response content generated by the generation AI is converted into voice using the voice tempo and intonation model. As a result, the text-based response generated by the generation AI is provided to the customer as a natural voice. For example, a quick and accurate response to customer inquiries is possible. In addition, by performing fine tuning, the response content of the generation AI can be made closer to a more accurate one. In this way, the present invention combines large-scale voice data and generation AI to realize automation of customer responses, thereby improving business efficiency and customer satisfaction. As a result, the customer response automation system is able to respond quickly and accurately to inquiries from customers.

The customer support automation system according to the embodiment includes an analysis unit, a generation unit, and a voice conversion unit. The analysis unit analyzes voice data. The analysis unit converts voice data into text data, for example, using voice recognition technology. The analysis unit can also analyze the content of the voice data using natural language processing technology. For example, the analysis unit analyzes the phonemes and phonology of the voice data and models the tempo and intonation of the voice. The generation unit uses a generation AI to generate a response based on the data analyzed by the analysis unit. The generation unit generates a response, for example, using a text generation AI (for example, LLM). The generation unit can also use a generation AI to generate a response that has knowledge about a specific business or service. For example, the generation unit generates an appropriate response to an inquiry about customer support. The voice conversion unit voices the response generated by the generation unit. The voice conversion unit converts text data into voice data, for example, using voice synthesis technology. The voice conversion unit can also provide the generated voice data to the customer. For example, the voice conversion unit provides the generated voice data to the customer via telephone or the Internet. This enables the customer response automation system according to the embodiment to respond quickly and accurately to customer inquiries. Some or all of the above-mentioned processes in the analysis unit, generation unit, and voice conversion unit may be performed, for example, using AI, or may be performed without using AI. For example, the analysis unit can input voice data to the AI and have the AI analyze the voice data. The generation unit can input data analyzed by the analysis unit to the AI and have the AI generate a response. The voice conversion unit can input the response generated by the generation unit to the AI and have the AI convert it into voice.

The analysis unit analyzes the voice data. The analysis unit converts the voice data into text data, for example, using voice recognition technology. Specifically, the voice recognition technology converts the voice signal into digital data, and analyzes the digital data to identify phonemes and phonology. This makes it possible to extract information contained in the voice data in text format. The analysis unit can also analyze the contents of the voice data using natural language processing technology. Natural language processing technology analyzes the grammatical structure and meaning of the text data, and performs understanding based on the context. For example, the analysis unit analyzes the phonemes and phonology of the voice data and models the tempo and intonation of the voice. This makes it possible to grasp the speaker's emotions and intentions more accurately. Furthermore, the analysis unit can also use filtering technology to remove background noise and echoes from the voice data. This makes it possible to improve the quality of the voice data and increase the accuracy of the analysis. The analysis unit combines these technologies to analyze the voice data with high accuracy and output it as text data. The analysis unit can share the results of the analysis of the voice data with other systems and departments, and can, for example, work with a customer support system or database to improve the efficiency of customer support.

The generation unit uses the generation AI to generate a response based on the data analyzed by the analysis unit. The generation unit generates a response using, for example, a text generation AI (e.g., LLM). Specifically, the generation AI receives text data provided by the analysis unit as input and generates an appropriate response based on the content. The generation AI has learned a large amount of text data and has the ability to understand grammar and context, so it can generate natural and fluent responses. The generation unit can also use the generation AI to generate responses with knowledge of specific business operations and services. For example, the generation unit generates an appropriate response to an inquiry about customer support. The generation AI has learned knowledge about specific business operations and services in advance and can also respond to specialized questions. Furthermore, the generation unit can evaluate the quality of the generated response and make corrections as necessary. For example, if the response generated by the generation AI is inappropriate, the generation unit reevaluates the content of the response and generates a more appropriate response. The generation unit can also store the generated response in a database and use it as a reference for future inquiries. This allows the generation unit to generate quick and accurate responses, improving the efficiency and quality of customer support.

The voice conversion unit converts the response generated by the generation unit into voice. The voice conversion unit converts text data into voice data, for example, using voice synthesis technology. Specifically, the voice synthesis technology receives text data as input and generates natural voice based on the content. The voice synthesis technology generates natural and easy-to-listen voice by analyzing combinations of phonemes and phonological elements and adding appropriate intonation and tempo. The voice conversion unit can also provide the generated voice data to customers. For example, the voice conversion unit provides the generated voice data to customers via telephone or the Internet. In the case of telephone, the voice conversion unit transmits the generated voice data to a telephone line in real time and responds directly to the customer. In the case of the Internet, the voice conversion unit distributes the generated voice data in a streaming format so that the customer can listen to the voice through a web browser or a dedicated application. Furthermore, the voice conversion unit can evaluate the quality of the generated voice data and make corrections as necessary. For example, if the intonation or tempo of the voice is unnatural, the voice conversion unit readjusts the voice synthesis technology to generate a more natural voice. The voice conversion unit can also store the generated voice data in a database and use it as a reference for future inquiries. This allows the voice conversion unit to provide fast and accurate voice responses, improving the efficiency and quality of customer service.

The generation unit can include an adjustment unit that performs fine tuning. The adjustment unit performs fine tuning of the generation AI. The adjustment unit improves the accuracy of the response by, for example, adjusting parameters of the generation AI. The adjustment unit can also select training data. For example, the adjustment unit selects data related to a specific business or service and uses it for training the generation AI. In this way, the generation unit can make the response content of the generation AI more accurate by performing fine tuning. Some or all of the above-mentioned processing in the adjustment unit may be performed, for example, using AI, or may be performed without using AI. For example, the adjustment unit can cause the AI to adjust parameters of the generation AI.

The voice conversion unit may include a providing unit that provides the generated voice to the customer. The providing unit provides the generated voice to the customer. The providing unit provides the voice to the customer, for example, via telephone. The providing unit may also provide the voice to the customer via the Internet. For example, the providing unit provides the voice through a website or a mobile app. In this way, by providing the generated voice to the customer, it becomes possible to respond in a natural voice. Some or all of the above-mentioned processing in the providing unit may be performed, for example, using AI, or may be performed without using AI. For example, the providing unit may input the generated voice data to AI and cause the AI to provide the voice.

The analysis unit can analyze multiple pieces of voice data and model the tempo and intonation of the voice. The analysis unit analyzes multiple pieces of voice data, such as telephone voices and recorded voices. The analysis unit analyzes the voice waveform and extracts rhythm patterns. For example, the analysis unit analyzes the phonemes and phonology of the voice data and models the tempo and intonation of the voice. In this way, the rhythm and intonation of natural conversation can be learned by analyzing large-scale voice data. Some or all of the above-mentioned processing in the analysis unit may be performed, for example, using AI, or may be performed without using AI. For example, the analysis unit can input the voice data to AI and have the AI analyze the voice data.

The generation unit can use a generation AI that has knowledge about a specific business or service. The generation unit uses a generation AI that has knowledge about a specific business or service, such as customer support or medical consultation. The generation unit uses the generation AI to generate an appropriate response to an inquiry about the specific business or service. For example, the generation unit generates an appropriate response to an inquiry about customer support. In this way, by using the generation AI that has knowledge about the specific business or service, a more appropriate response can be generated. Some or all of the above-mentioned processing in the generation unit may be performed, for example, using AI, or may be performed without using AI. For example, the generation unit can cause the AI to execute the generation AI that has knowledge about the specific business or service.

The analysis unit can estimate the user's emotions and adjust the analysis method of the voice data based on the estimated user's emotions. For example, when the user is feeling stressed, the analysis unit can slow down the tempo of the voice data and make the intonation gentle. In addition, when the user is relaxed, the analysis unit can also speed up the tempo of the voice data and enrich the intonation. Furthermore, when the user is in a hurry, the analysis unit can also speed up the tempo of the voice data and simplify the intonation. This allows for more appropriate analysis by adjusting the analysis method of the voice data according to the user's emotions. The estimation of emotions is realized using an emotion estimation function using, for example, an emotion engine or a generation AI. The generation AI is, for example, a text generation AI (for example, LLM) or a multimodal generation AI, but is not limited to such examples. A part or all of the above-mentioned processing in the analysis unit may be performed using, for example, an AI, or may be performed without using an AI. For example, the analysis unit can input the user's emotion data to the AI and cause the AI to execute emotion estimation.

When analyzing voice data, the analysis unit can improve the analysis accuracy by taking into account a specific accent or dialect. For example, when analyzing voice data having an accent of a specific region, the analysis unit applies an accent model of that region. In addition, when analyzing voice data having a specific dialect, the analysis unit can also perform the analysis by taking into account the characteristics of the dialect. Furthermore, when analyzing voice data containing a mixture of multiple accents or dialects, the analysis unit can also perform the analysis by integrating the characteristics of each. In this way, by taking into account a specific accent or dialect, the analysis accuracy is improved. Some or all of the above-mentioned processing in the analysis unit may be performed, for example, using AI, or may be performed without using AI. For example, the analysis unit can input voice data having a specific accent or dialect to AI and have the AI perform the analysis.

The analysis unit can perform filtering to remove background noise when analyzing the voice data. For example, the analysis unit applies filtering to remove background noise before analyzing the voice data. The analysis unit can also perform filtering to remove noise in a specific frequency band. Furthermore, the analysis unit can also perform filtering to remove dynamically changing background noise in real time. This removes background noise, improving analysis accuracy. Some or all of the above-mentioned processing in the analysis unit may be performed using AI, for example, or may be performed without using AI. For example, the analysis unit can input the voice data to AI and cause the AI to remove background noise.

The analysis unit can estimate the user's emotions and determine the priority of the voice data to be analyzed based on the estimated user's emotions. For example, when the user is feeling stressed, the analysis unit preferentially analyzes voice data for reducing stress. In addition, when the user is relaxed, the analysis unit can also preferentially analyze voice data for maintaining relaxation. Furthermore, when the user is in a hurry, the analysis unit can also preferentially analyze voice data for responding quickly. This enables more appropriate analysis by determining the priority of the voice data to be analyzed according to the user's emotions. The estimation of emotions is realized using an emotion estimation function using, for example, an emotion engine or a generation AI. The generation AI is, for example, a text generation AI (for example, LLM) or a multimodal generation AI, but is not limited to such examples. A part or all of the above-mentioned processing in the analysis unit may be performed using, for example, an AI, or may be performed without using an AI. For example, the analysis unit can input the user's emotion data to the AI and cause the AI to execute emotion estimation.

When analyzing the voice data, the analysis unit can adjust the analysis method based on the geographical location information of the user. For example, when the user is in a specific area, the analysis unit performs the analysis taking into account the accent and dialect of that area. In addition, when the user is moving, the analysis unit can also perform the analysis taking into account the accent and dialect of the destination area. Furthermore, when the user is in different areas, the analysis unit can also perform the analysis by integrating the characteristics of each area. In this way, the analysis method can be adjusted by taking into account the geographical location information of the user. Some or all of the above-mentioned processing in the analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the analysis unit can input the geographical location information of the user to AI and cause the AI to adjust the analysis method.

When analyzing the voice data, the analysis unit can analyze the user's social media activity and prioritize analysis of related voice data. For example, the analysis unit prioritizes analysis of voice data related to a specific topic from the user's social media activity. The analysis unit can also prioritize analysis of voice data related to a specific event from the user's social media activity. Furthermore, the analysis unit can also prioritize analysis of voice data related to a specific person from the user's social media activity. In this way, by analyzing the user's social media activity, related voice data can be prioritized. A part or all of the above-mentioned processing in the analysis unit may be performed, for example, using AI or may be performed without using AI. For example, the analysis unit can input the user's social media activity data to AI and cause AI to perform analysis to determine the priority order of related voice data.

The generation unit can estimate the user's emotions and adjust the way of expressing the response based on the estimated user's emotions. For example, when the user is stressed, the generation unit generates a response using a gentle expression method. Also, when the user is relaxed, the generation unit can generate a response using a friendly expression method. Furthermore, when the user is in a hurry, the generation unit can generate a response using a concise and quick expression method. This allows for a more appropriate response by adjusting the way of expressing the response according to the user's emotions. The estimation of emotions is realized using an emotion estimation function using, for example, an emotion engine or a generation AI. The generation AI is, for example, a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. A part or all of the above-mentioned processing in the generation unit may be performed using, for example, an AI, or may be performed without using an AI. For example, the generation unit can input the user's emotion data to the AI and cause the AI to execute emotion estimation.

When generating a response, the generation unit can adjust the level of detail of the response based on the importance of the inquiry content. For example, the generation unit generates a detailed response to an inquiry of high importance. The generation unit can also generate a concise response to an inquiry of low importance. Furthermore, the generation unit can dynamically adjust the level of detail of the response according to the importance. This allows for a more appropriate response by adjusting the level of detail of the response based on the importance of the inquiry content. Some or all of the above-mentioned processing in the generation unit may be performed using, for example, AI, or may be performed without using AI. For example, the generation unit can input the importance of the inquiry content to the AI and cause the AI to execute processing to adjust the level of detail of the response.

When generating a response, the generation unit can apply different generation algorithms depending on the category of the inquiry. For example, the generation unit applies a specialized generation algorithm to a technical inquiry. The generation unit can also apply a general-purpose generation algorithm to a general inquiry. Furthermore, the generation unit can apply a quick generation algorithm to an urgent inquiry. This allows for a more appropriate response by applying different generation algorithms depending on the inquiry category. Some or all of the above-mentioned processing in the generation unit may be performed using, for example, AI, or may be performed without using AI. For example, the generation unit can input the inquiry category to the AI and cause the AI to execute a process of determining the generation algorithm to be applied.

The generation unit can estimate the user's emotions and adjust the length of the response based on the estimated user's emotions. For example, when the user is stressed, the generation unit generates a short, to-the-point response. Also, when the user is relaxed, the generation unit can generate a longer response including detailed explanations. Furthermore, when the user is in a hurry, the generation unit can generate a quick, concise response. This allows for a more appropriate response by adjusting the length of the response according to the user's emotions. The estimation of emotions is realized using an emotion estimation function, for example, using an emotion engine or generation AI. The generation AI is, for example, a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. A part or all of the above-mentioned processing in the generation unit may be performed, for example, using AI, or may be performed without using AI. For example, the generation unit can input the user's emotion data to the AI and cause the AI to perform emotion estimation.

When generating a reply, the generation unit can determine the priority of the reply based on the time of submission of the inquiry. For example, the generation unit can generate a reply with priority to a recently submitted inquiry. The generation unit can also generate a reply with priority to an inquiry that has been unresolved for a long time. Furthermore, the generation unit can dynamically adjust the priority of the reply depending on the submission time. This allows for a faster response by determining the priority of the reply based on the submission time of the inquiry. Some or all of the above-mentioned processing in the generation unit may be performed using, for example, AI, or may be performed without using AI. For example, the generation unit can input the submission time of the inquiry to the AI and cause the AI to execute a process of determining the priority of the reply.

When generating a reply, the generation unit can adjust the order of replies based on the relevance of the inquiries. For example, the generation unit generates replies preferentially for inquiries with high relevance. The generation unit can also postpone generating replies for inquiries with low relevance. Furthermore, the generation unit can dynamically adjust the order of replies according to the relevance. This allows for a more appropriate reply by adjusting the order of replies based on the relevance of the inquiries. Some or all of the above-mentioned processing in the generation unit may be performed, for example, using AI, or may be performed without using AI. For example, the generation unit can input the relevance of the inquiries to the AI and cause the AI to execute processing to adjust the order of replies.

The voice conversion unit can estimate the user's emotion and adjust the vocalization expression method based on the estimated user's emotion. For example, when the user is stressed, the voice conversion unit performs vocalization in a gentle voice. Also, when the user is relaxed, the voice conversion unit can perform vocalization in a friendly voice. Furthermore, when the user is in a hurry, the voice conversion unit can perform vocalization in a quick and concise voice. This allows for more appropriate vocalization by adjusting the vocalization expression method according to the user's emotion. The emotion estimation is realized using an emotion estimation function, for example, using an emotion engine or a generation AI. The generation AI is a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. A part or all of the above-mentioned processing in the voice conversion unit may be performed, for example, using AI, or may be performed without using AI. For example, the voice conversion unit can input the user's emotion data to the AI and cause the AI to perform emotion estimation.

The voice conversion unit can perform voice filtering to improve the naturalness of the generated voice during voice conversion. For example, the voice conversion unit applies a noise reduction filter to the generated voice. The voice conversion unit can also apply an echo canceling filter to the generated voice. Furthermore, the voice conversion unit can apply a sound quality improvement filter to the generated voice. This improves the naturalness of the generated voice, making it possible to convert the voice into a more natural voice. Some or all of the above-mentioned processing in the voice conversion unit can be performed using, for example, AI, or can be performed without using AI. For example, the voice conversion unit can input the generated voice data to AI and have the AI perform voice filtering.

The voice conversion unit can improve the accuracy of voice conversion by taking into account a specific accent or dialect when generating voice. For example, when generating voice with an accent of a specific region, the voice conversion unit applies an accent model of that region. In addition, when generating voice with a specific dialect, the voice conversion unit can also perform voice conversion by taking into account the characteristics of the dialect. Furthermore, when generating voice in which multiple accents or dialects are mixed, the voice conversion unit can also perform voice conversion by integrating the characteristics of each. In this way, the accuracy of voice conversion is improved by taking into account a specific accent or dialect. Some or all of the above-mentioned processing in the voice conversion unit may be performed using, for example, AI, or may be performed without using AI. For example, the voice conversion unit can input voice data with a specific accent or dialect to AI and have the AI perform voice conversion.

The voice conversion unit can estimate the user's emotion and determine the priority of voice conversion based on the estimated user's emotion. For example, when the user is stressed, the voice conversion unit generates a voice for reducing stress. In addition, when the user is relaxed, the voice conversion unit can also generate a voice for maintaining relaxation. Furthermore, when the user is in a hurry, the voice conversion unit can also generate a voice for responding quickly. This enables more appropriate voice conversion by determining the priority of voice conversion according to the user's emotion. The emotion estimation is realized using an emotion estimation function using, for example, an emotion engine or a generation AI. The generation AI is, for example, a text generation AI (for example, LLM) or a multimodal generation AI, but is not limited to such examples. A part or all of the above-mentioned processing in the voice conversion unit may be performed using, for example, an AI, or may be performed without using an AI. For example, the voice conversion unit can input the user's emotion data to the AI and cause the AI to execute emotion estimation.

The voice conversion unit can adjust the voice conversion method taking into account the geographical location information of the user when vocalizing. For example, when the user is in a specific area, the voice conversion unit performs voice conversion taking into account the accent or dialect of that area. In addition, when the user is moving, the voice conversion unit can also perform voice conversion taking into account the accent or dialect of the area to which the user is moving. Furthermore, when the user is in different areas, the voice conversion unit can also perform voice conversion by integrating the characteristics of each area. In this way, the voice conversion method can be adjusted by taking into account the geographical location information of the user. Part or all of the above-mentioned processing in the voice conversion unit may be performed using, for example, AI, or may be performed without using AI. For example, the voice conversion unit can input the geographical location information of the user to AI and cause the AI to adjust the voice conversion method.

When vocalizing, the vocalization unit can analyze the user's social media activity and vocalize related voice data preferentially. For example, the vocalization unit preferentially generates voice related to a specific topic from the user's social media activity. The vocalization unit can also preferentially generate voice related to a specific event from the user's social media activity. Furthermore, the vocalization unit can also preferentially generate voice related to a specific person from the user's social media activity. In this way, by analyzing the user's social media activity, related voice data can be preferentially vocalized. A part or all of the above-mentioned processing in the vocalization unit may be performed, for example, using AI or may be performed without using AI. For example, the vocalization unit inputs the user's social media activity data into AI and causes AI to perform vocalization that determines the priority of related voice data.

The adjustment unit can estimate the user's emotion and adjust the fine-tuning parameters based on the estimated user's emotion. For example, when the user is feeling stressed, the adjustment unit adjusts the parameters for reducing stress. In addition, when the user is relaxed, the adjustment unit can also adjust the parameters for maintaining relaxation. Furthermore, when the user is in a hurry, the adjustment unit can adjust the parameters for responding quickly. This allows for more appropriate adjustment by adjusting the fine-tuning parameters according to the user's emotion. The emotion estimation is realized using an emotion estimation function using, for example, an emotion engine or a generation AI. The generation AI is, for example, a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. A part or all of the above-mentioned processing in the adjustment unit may be performed using, for example, an AI, or may be performed without using an AI. For example, the adjustment unit can input the user's emotion data to the AI and cause the AI to execute emotion estimation.

During fine tuning, the adjustment unit can optimize the generation algorithm by referring to past inquiry data. The adjustment unit, for example, analyzes past inquiry data and optimizes parameters of the generation algorithm. The adjustment unit can also extract specific patterns from past inquiry data and reflect them in the generation algorithm. Furthermore, the adjustment unit can improve the accuracy of the generation algorithm based on past inquiry data. In this way, the generation algorithm can be optimized by referring to past inquiry data. Some or all of the above-mentioned processing in the adjustment unit may be performed, for example, using AI, or may be performed without using AI. For example, the adjustment unit can input past inquiry data to AI and cause AI to optimize the generation algorithm.

The adjustment unit can estimate the user's emotion and adjust the frequency of fine tuning based on the estimated user's emotion. For example, when the user is stressed, the adjustment unit performs fine tuning frequently to reduce stress. In addition, when the user is relaxed, the adjustment unit can also reduce the frequency of fine tuning to maintain relaxation. Furthermore, when the user is in a hurry, the adjustment unit can perform fine tuning frequently to respond quickly. This allows for more appropriate adjustment by adjusting the frequency of fine tuning according to the user's emotion. The emotion estimation is realized using an emotion estimation function using, for example, an emotion engine or a generation AI. The generation AI is, for example, a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. A part or all of the above-mentioned processing in the adjustment unit may be performed using, for example, an AI, or may be performed without using an AI. For example, the adjustment unit can input the user's emotion data to the AI and cause the AI to perform emotion estimation.

During fine tuning, the adjustment unit can weight the learning data based on the time of inquiry submission. For example, the adjustment unit weights recent inquiry data and reflects the weight in the generation algorithm. The adjustment unit can also weight inquiry data that has been unresolved for a long time and reflect the weight in the generation algorithm. Furthermore, the adjustment unit can dynamically adjust the weighting of the learning data according to the submission time. This allows for more appropriate adjustment by weighting the learning data based on the time of inquiry submission. Some or all of the above-mentioned processing in the adjustment unit may be performed using, for example, AI, or may be performed without using AI. For example, the adjustment unit can input the time of inquiry submission to AI and cause AI to perform weighting of the learning data.

The providing unit can estimate the user's emotions and adjust the method of providing voice based on the estimated user's emotions. For example, when the user is stressed, the providing unit provides voice in a gentle voice. Also, when the user is relaxed, the providing unit can provide voice in a friendly voice. Furthermore, when the user is in a hurry, the providing unit can provide voice in a quick and concise voice. This allows for more appropriate voice provision by adjusting the method of providing voice according to the user's emotions. The estimation of emotions is realized using an emotion estimation function using, for example, an emotion engine or a generation AI. The generation AI is, for example, a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. A part or all of the above-mentioned processing in the providing unit may be performed using, for example, an AI, or may be performed without using an AI. For example, the providing unit can input the user's emotion data to the AI and cause the AI to perform emotion estimation.

When providing voice, the providing unit can select the optimal voice providing method by referring to the user's past inquiry history. For example, the providing unit selects the optimal voice providing method from the user's past inquiry history. The providing unit can also analyze the user's past inquiry history and select the voice providing method based on a specific pattern. Furthermore, the providing unit can dynamically adjust the voice providing method based on the user's past inquiry history. In this way, the optimal voice providing method can be selected by referring to the user's past inquiry history. A part or all of the above-mentioned processing in the providing unit may be performed, for example, using AI or may be performed without using AI. For example, the providing unit can input the user's past inquiry history to AI and cause AI to execute processing to select the optimal providing method.

The providing unit can estimate the user's emotion and determine the priority of voice provision based on the estimated user's emotion. For example, when the user is feeling stressed, the providing unit can provide voice for reducing stress preferentially. In addition, when the user is relaxed, the providing unit can also provide voice for maintaining relaxation preferentially. Furthermore, when the user is in a hurry, the providing unit can also provide voice for responding quickly preferentially. This enables more appropriate voice provision by determining the priority of voice provision according to the user's emotion. The emotion estimation is realized using an emotion estimation function using, for example, an emotion engine or a generation AI. The generation AI is, for example, a text generation AI (for example, LLM) or a multimodal generation AI, but is not limited to such examples. A part or all of the above-mentioned processing in the providing unit may be performed using, for example, an AI, or may be performed without using an AI. For example, the providing unit can input the user's emotion data to the AI and cause the AI to execute emotion estimation.

When providing voice, the providing unit can select the optimal providing method by taking into account the device information of the user. For example, when the user is using a smartphone, the providing unit selects a voice providing method that matches the screen size. In addition, when the user is using a tablet, the providing unit can also select a voice providing method optimized for a large screen. Furthermore, when the user is using a smartwatch, the providing unit can also select a voice providing method that is simple and highly visible. In this way, the optimal voice providing method can be selected by taking into account the device information of the user. Some or all of the above-mentioned processing in the providing unit may be performed, for example, using AI, or may be performed without using AI. For example, the providing unit can input the device information of the user to the AI and cause the AI to execute a process of selecting the optimal providing method.

The system according to the embodiment is not limited to the above-mentioned example, and various modifications are possible, for example, as follows:

When analyzing the voice data, the analysis unit can improve the accuracy of the analysis by referring to the user's past inquiry history. For example, the analysis unit extracts specific patterns from the past inquiry history and reflects them in the analysis of the voice data. The analysis unit can also learn the user's speech tendencies based on the past inquiry history and improve the accuracy of the analysis. Furthermore, the analysis unit can perform analysis with knowledge of specific tasks or services by referring to the past inquiry history. In this way, by referring to the user's past inquiry history, the accuracy of the analysis can be improved and more appropriate responses can be made.

The generation unit can estimate the user's emotions and adjust the tone of the reply based on the estimated user's emotions. For example, if the user is angry, the generation unit generates a reply in a calm and subdued tone. Also, if the user is sad, the generation unit can generate a reply in a gentle tone. Furthermore, if the user is happy, the generation unit can generate a reply in a bright tone. This allows for a more appropriate reply by adjusting the tone of the reply according to the user's emotions.

When providing the generated voice, the voice generation unit can adjust the length of the voice taking into account the remaining battery level of the user's device. For example, when the battery level is low, the voice generation unit can provide a short voice that focuses on the main points. When the battery level is sufficient, the voice generation unit can also provide a voice that includes a detailed explanation. Furthermore, when the battery level is moderate, the voice generation unit can provide a voice of an appropriate length. This makes it possible to provide optimal voice by taking into account the remaining battery level of the user's device.

When analyzing the voice data, the analysis unit can monitor the user's speaking speed in real time and dynamically adjust the analysis method. For example, if the user's speaking speed is fast, the analysis unit can increase the analysis speed. Also, if the user's speaking speed is slow, the analysis unit can decrease the analysis speed. Furthermore, if the user's speaking speed fluctuates, the analysis speed can also be dynamically adjusted. This allows for more appropriate analysis by adjusting the analysis method according to the user's speaking speed.

The generation unit can estimate the user's emotions and customize the content of the reply based on the estimated user's emotions. For example, if the user is feeling anxious, the generation unit can generate a reply with content that gives a sense of security. Also, if the user is excited, the generation unit can generate a reply with content that encourages the user to remain calm. Furthermore, if the user is confused, the generation unit can generate a reply with clear and easy-to-understand content. This allows for a more appropriate reply by customizing the content of the reply according to the user's emotions.

When analyzing the voice data, the analysis unit can estimate the user's age group and apply an analysis method appropriate to the age group. For example, the analysis unit can perform an analysis that takes into account casual language for younger users. Also, for older users, the analysis unit can perform an analysis that takes into account polite language. Furthermore, the analysis can also take into account the frequency of use of specific words and phrases depending on the age group. In this way, by applying an analysis method appropriate to the user's age group, the analysis accuracy can be improved and more appropriate responses can be made.

The generation unit can estimate the user's emotions and adjust the timing of the response based on the estimated user's emotions. For example, if the user is impatient, the generation unit can generate a response quickly. Also, if the user is relaxed, the generation unit can generate a response with a slight delay. Furthermore, if the user is angry, the response can be delayed to give the user time to calm down. In this way, a more appropriate response can be provided by adjusting the timing of the response according to the user's emotions.

When providing the generated voice, the voice generation unit can adjust the frequency band of the voice taking into account the user's hearing characteristics. For example, if the user has difficulty hearing high-pitched sounds, the voice generation unit can provide voice with emphasis on low-pitched sounds. Also, if the user has difficulty hearing low-pitched sounds, the voice generation unit can provide voice with emphasis on high-pitched sounds. Furthermore, specific frequency bands can be emphasized or suppressed depending on the user's hearing characteristics. This makes it possible to provide optimal voice by taking into account the user's hearing characteristics.

When analyzing the voice data, the analysis unit can estimate emotions based on the content of the user's speech and adjust the depth of the analysis depending on the estimated emotion. For example, if the user makes an emotional statement, the analysis unit can perform a detailed analysis. Also, if the user makes a calm statement, the analysis unit can perform a simplified analysis. Furthermore, if the user's emotions fluctuate, the analysis depth can be dynamically adjusted. In this way, more appropriate analysis is possible by estimating emotions based on the content of the user's speech and adjusting the depth of the analysis.

When generating a response, the generation unit can maintain consistency in the response by referring to the user's past inquiry history. For example, the generation unit generates a response that matches the content of the past inquiry. In addition, the generation unit can generate a response that reflects the user's preferences and tendencies based on the past inquiry history. Furthermore, by referring to the past inquiry history, a response without inconsistencies can be generated. In this way, by referring to the user's past inquiry history, consistency in the response can be maintained, enabling a more appropriate response.

The process flow for Example 2 is briefly explained below.

Step 1: The analysis unit analyzes the voice data. For example, the analysis unit converts the voice data into text data using a voice recognition technique. The analysis unit can also analyze the contents of the voice data using a natural language processing technique. For example, the analysis unit analyzes the phonemes and phonology of the voice data and models the tempo and intonation of the voice.
Step 2: The generation unit uses the generation AI to generate a response based on the data analyzed by the analysis unit. The generation unit generates a response, for example, using a text generation AI (e.g., LLM). The generation unit can also use the generation AI to generate a response with knowledge about a specific business or service. For example, the generation unit generates an appropriate response to a customer support inquiry.
Step 3: The voice conversion unit converts the response generated by the generation unit into voice. The voice conversion unit converts the text data into voice data, for example, using a voice synthesis technique. The voice conversion unit can also provide the generated voice data to the customer. For example, the voice conversion unit provides the generated voice data to the customer over the telephone or the Internet.

The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input for the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is generative AI such as ChatGPT (registered trademark) (Internet search <URL: https://openai.com/blog/chatgpt>). The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating a voice, text data indicating a text, and image data indicating an image (e.g., still image data or video data) is input. The data generation model 58 infers the input inference data according to the instruction indicated by the prompt, and outputs the inference result in one or more data formats such as voice data, text data, and image data. The data generation model 58 includes, for example, text generation AI, image generation AI, and multimodal generation AI. Here, inference refers to, for example, analysis, classification, prediction, and/or summarization. The identification processing unit 290 performs the above-mentioned identification processing while using the data generation model 58. The data generation model 58 may be a fine-tuned model to output an inference result from a prompt that does not include an instruction, in which case the data generation model 58 can output an inference result from a prompt that does not include an instruction. In the data processing device 12, etc., the data generation model 58 includes a plurality of types, and the data generation model 58 includes an AI other than the generation AI. The AI other than the generation AI may be, for example, linear regression, logistic regression, decision tree, random forest, support vector machine (SVM), k-means clustering, convolutional neural network (CNN), recurrent neural network (RNN), generative adversarial network (GAN), or naive Bayes, and may perform various processes, but is not limited to such examples. The AI may also be an AI agent. In addition, when the processing of each part described above is performed by AI, the processing is performed in part or in whole by AI, but is not limited to such examples. In addition, processing performed by AI, including the generating AI, may be replaced with rule-based processing, and rule-based processing may be replaced with processing performed by AI, including the generating AI.

The processing by the data processing system 10 described above is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the smart device 14, but may also be executed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the smart device 14. The specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the smart device 14 or an external device, and the smart device 14 acquires or collects information required for processing from the data processing device 12 or an external device.

Each of the multiple elements including the above-mentioned analysis unit, generation unit, and voice conversion unit is realized, for example, by at least one of the smart device 14 and the data processing device 12. For example, the analysis unit is realized by the processor 46 of the smart device 14, and analyzes voice data and converts it into text data. The generation unit is realized, for example, by the specific processing unit 290 of the data processing device 12, and generates a response based on the analyzed data. The voice conversion unit is realized, for example, by the control unit 46A of the smart device 14, and converts the generated response into voice data and provides it to the customer. The correspondence between each unit and the device or control unit is not limited to the above-mentioned example, and various changes are possible.

[Second embodiment]
FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

As shown in FIG. 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

The data processing device 12 includes a computer 22, a database 24, and a communication I/F 26. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I/F 26 are also connected to the bus 34. The communication I/F 26 is connected to a network 54. Examples of the network 54 include a WAN and/or a LAN.

The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I/F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

The microphone 238 receives instructions and the like from the user by receiving voice uttered by the user. The microphone 238 captures the voice uttered by the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs the voice according to instructions from the processor 46.

Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures the user's surroundings (e.g., an imaging range defined by an angle of view equivalent to the field of vision of a typical able-bodied person).

The communication I/F 44 is connected to the network 54. The communication I/Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I/Fs 44 and 26 is performed in a secure state.

Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotion, including estimation and prediction of the user's emotion, but is not limited to such examples. Furthermore, the estimation and prediction of emotion also includes, for example, analysis of emotions.

In the smart glasses 214, the specific processing is performed by the processor 46. The storage 50 stores the specific processing program 60. The processor 46 reads the specific processing program 60 from the storage 50 and executes the read specific processing program 60 on the RAM 48. The specific processing is realized by the processor 46 operating as the control unit 46A in accordance with the specific processing program 60 executed on the RAM 48. The smart glasses 214 also have a data generation model and an emotion identification model similar to the data generation model 58 and the emotion identification model 59, and can use these models to perform processing similar to that of the specific processing unit 290.

Note that a device other than the data processing device 12 may have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 obtains a processing result (such as a prediction result) using the data generation model 58 by communicating with the server device having the data generation model 58. In addition, the data processing device 12 may be a server device, or may be a terminal device owned by a user (for example, a mobile phone, a robot, a home appliance, etc.).

The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input for the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

The data generation model 58 is a so-called generative AI. An example of the data generation model 58 is a generative AI such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating a voice, text data indicating a text, and image data indicating an image (e.g., still image data or video data) is input. The data generation model 58 infers the input inference data according to the instruction indicated by the prompt, and outputs the inference result in one or more data formats such as voice data, text data, and image data. The data generation model 58 includes, for example, a text generation AI, an image generation AI, and a multimodal generation AI. Here, inference refers to, for example, analysis, classification, prediction, and/or summarization. The identification processing unit 290 performs the above-mentioned identification processing while using the data generation model 58. The data generation model 58 may be a fine-tuned model so as to output an inference result from a prompt that does not include an instruction, in which case the data generation model 58 can output an inference result from a prompt that does not include an instruction. In the data processing device 12, etc., the data generation model 58 includes a plurality of types, and the data generation model 58 includes an AI other than the generation AI. The AI other than the generation AI may be, for example, a linear regression, a logistic regression, a decision tree, a random forest, a support vector machine (SVM), a k-means clustering, a convolutional neural network (CNN), a recurrent neural network (RNN), a generative adversarial network (GAN), or a naive Bayes, and may perform various processes, but is not limited to such examples. The AI may also be an AI agent. In addition, when the processing of each part described above is performed by an AI, the processing is performed in part or in whole by the AI, but is not limited to such examples. In addition, the processing performed by AI, including the generating AI, may be replaced with rule-based processing, and the rule-based processing may be replaced with processing performed by AI, including the generating AI.

The data processing system 210 according to the second embodiment performs the same processing as the data processing system 10 according to the first embodiment. The processing by the data processing system 210 is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the smart glasses 214, but may be executed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the smart glasses 214. In addition, the specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the smart glasses 214 or an external device, etc., and the smart glasses 214 acquires or collects information required for processing from the data processing device 12 or an external device, etc.

Each of the multiple elements including the above-mentioned analysis unit, generation unit, and voice conversion unit is realized, for example, by at least one of the smart glasses 214 and the data processing device 12. For example, the analysis unit is realized by the processor 46 of the smart glasses 214, and analyzes voice data and converts it into text data. The generation unit is realized, for example, by the specific processing unit 290 of the data processing device 12, and generates a response based on the analyzed data. The voice conversion unit is realized, for example, by the control unit 46A of the smart glasses 214, and converts the generated response into voice data and provides it to the customer. The correspondence between each unit and the device or control unit is not limited to the above-mentioned example, and various changes are possible.

[Third embodiment]
FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

As shown in FIG. 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

The data processing device 12 includes a computer 22, a database 24, and a communication I/F 26. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I/F 26 are also connected to the bus 34. The communication I/F 26 is connected to a network 54. Examples of the network 54 include a WAN and/or a LAN.

The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I/F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

The microphone 238 receives instructions and the like from the user by receiving voice uttered by the user. The microphone 238 captures the voice uttered by the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs the voice according to instructions from the processor 46.

Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures the user's surroundings (e.g., an imaging range defined by an angle of view equivalent to the field of vision of a typical able-bodied person).

The communication I/F 44 is connected to the network 54. The communication I/Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I/Fs 44 and 26 is performed in a secure state.

Figure 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Figure 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotion, including estimation and prediction of the user's emotion, but is not limited to such examples. Furthermore, the estimation and prediction of emotion also includes, for example, analysis of emotions.

In the headset type terminal 314, the specific processing is performed by the processor 46. The storage 50 stores the specific program 60. The processor 46 reads the specific program 60 from the storage 50 and executes the read specific program 60 on the RAM 48. The specific processing is realized by the processor 46 operating as the control unit 46A in accordance with the specific program 60 executed on the RAM 48. The headset type terminal 314 has a data generation model and an emotion identification model similar to the data generation model 58 and the emotion identification model 59, and can also perform processing similar to that of the specific processing unit 290 using these models.

Note that a device other than the data processing device 12 may have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 obtains a processing result (such as a prediction result) using the data generation model 58 by communicating with the server device having the data generation model 58. In addition, the data processing device 12 may be a server device, or may be a terminal device owned by a user (for example, a mobile phone, a robot, a home appliance, etc.).

The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input for the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

The data generation model 58 is a so-called generative AI. An example of the data generation model 58 is a generative AI such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating a voice, text data indicating a text, and image data indicating an image (e.g., still image data or video data) is input. The data generation model 58 infers the input inference data according to the instruction indicated by the prompt, and outputs the inference result in one or more data formats such as voice data, text data, and image data. The data generation model 58 includes, for example, a text generation AI, an image generation AI, and a multimodal generation AI. Here, inference refers to, for example, analysis, classification, prediction, and/or summarization. The identification processing unit 290 performs the above-mentioned identification processing while using the data generation model 58. The data generation model 58 may be a fine-tuned model to output an inference result from a prompt that does not include an instruction, in which case the data generation model 58 can output an inference result from a prompt that does not include an instruction. In the data processing device 12, etc., the data generation model 58 includes a plurality of types, and the data generation model 58 includes an AI other than the generation AI. The AI other than the generation AI may be, for example, linear regression, logistic regression, decision tree, random forest, support vector machine (SVM), k-means clustering, convolutional neural network (CNN), recurrent neural network (RNN), generative adversarial network (GAN), or naive Bayes, and may perform various processes, but is not limited to such examples. The AI may also be an AI agent. In addition, when the processing of each part described above is performed by AI, the processing is performed in part or in whole by AI, but is not limited to such examples. In addition, the processing performed by AI, including the generating AI, may be replaced with rule-based processing, and the rule-based processing may be replaced with processing performed by AI, including the generating AI.

The data processing system 310 according to the third embodiment performs the same processing as the data processing system 10 according to the first embodiment. The processing by the data processing system 310 is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the headset type terminal 314, but may also be executed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the headset type terminal 314. In addition, the specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the headset type terminal 314 or an external device, and the headset type terminal 314 acquires or collects information required for processing from the data processing device 12 or an external device.

Each of the multiple elements including the above-mentioned analysis unit, generation unit, and voice conversion unit is realized, for example, by at least one of the headset type terminal 314 and the data processing device 12. For example, the analysis unit is realized by the processor 46 of the headset type terminal 314, and analyzes voice data and converts it into text data. The generation unit is realized, for example, by the specific processing unit 290 of the data processing device 12, and generates a response based on the analyzed data. The voice conversion unit is realized, for example, by the control unit 46A of the headset type terminal 314, and converts the generated response into voice data and provides it to the customer. The correspondence between each unit and the device or control unit is not limited to the above-mentioned example, and various changes are possible.

[Fourth embodiment]
FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

As shown in FIG. 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

The data processing device 12 includes a computer 22, a database 24, and a communication I/F 26. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I/F 26 are also connected to the bus 34. The communication I/F 26 is connected to a network 54. Examples of the network 54 include a WAN and/or a LAN.

The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I/F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

The microphone 238 receives instructions and the like from the user by receiving voice uttered by the user. The microphone 238 captures the voice uttered by the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs the voice according to instructions from the processor 46.

Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS image sensor or a CCD image sensor, and captures the user's surroundings (e.g., an imaging range defined by an angle of view equivalent to the field of vision of a typical able-bodied person).

The communication I/F 44 is connected to the network 54. The communication I/Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I/Fs 44 and 26 is performed in a secure state.

The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and legs. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and legs. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotion, including estimation and prediction of the user's emotion, but is not limited to such examples. Furthermore, the estimation and prediction of emotion also includes, for example, analysis of emotions.

In the robot 414, the specific processing is performed by the processor 46. The storage 50 stores the specific program 60. The processor 46 reads the specific program 60 from the storage 50 and executes the read specific program 60 on the RAM 48. The specific processing is realized by the processor 46 operating as the control unit 46A in accordance with the specific program 60 executed on the RAM 48. The robot 414 has a data generation model and an emotion identification model similar to the data generation model 58 and the emotion identification model 59, and can also perform processing similar to that of the specific processing unit 290 using these models.

Note that a device other than the data processing device 12 may have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 obtains a processing result (such as a prediction result) using the data generation model 58 by communicating with the server device having the data generation model 58. In addition, the data processing device 12 may be a server device, or may be a terminal device owned by a user (for example, a mobile phone, a robot, a home appliance, etc.).

The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating the user input for the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

The data generation model 58 is a so-called generative AI. An example of the data generation model 58 is a generative AI such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating a voice, text data indicating a text, and image data indicating an image (e.g., still image data or video data) is input. The data generation model 58 infers the input inference data according to the instruction indicated by the prompt, and outputs the inference result in one or more data formats such as voice data, text data, and image data. The data generation model 58 includes, for example, a text generation AI, an image generation AI, and a multimodal generation AI. Here, inference refers to, for example, analysis, classification, prediction, and/or summarization. The identification processing unit 290 performs the above-mentioned identification processing while using the data generation model 58. The data generation model 58 may be a fine-tuned model to output an inference result from a prompt that does not include an instruction, in which case the data generation model 58 can output an inference result from a prompt that does not include an instruction. In the data processing device 12, etc., the data generation model 58 includes a plurality of types, and the data generation model 58 includes an AI other than the generation AI. The AI other than the generation AI may be, for example, linear regression, logistic regression, decision tree, random forest, support vector machine (SVM), k-means clustering, convolutional neural network (CNN), recurrent neural network (RNN), generative adversarial network (GAN), or naive Bayes, and may perform various processes, but is not limited to such examples. The AI may also be an AI agent. In addition, when the processing of each part described above is performed by AI, the processing is performed in part or in whole by AI, but is not limited to such examples. In addition, the processing performed by AI, including the generating AI, may be replaced with rule-based processing, and the rule-based processing may be replaced with processing performed by AI, including the generating AI.

The data processing system 410 according to the fourth embodiment performs the same processing as the data processing system 10 according to the first embodiment. The processing by the data processing system 410 is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the robot 414, but may also be executed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the robot 414. In addition, the specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the robot 414 or an external device, etc., and the robot 414 acquires or collects information required for processing from the data processing device 12 or an external device, etc.

Each of the multiple elements including the above-mentioned analysis unit, generation unit, and voice conversion unit is realized, for example, by at least one of the robot 414 and the data processing device 12. For example, the analysis unit is realized by the processor 46 of the robot 414, and analyzes voice data and converts it into text data. The generation unit is realized, for example, by the specific processing unit 290 of the data processing device 12, and generates a response based on the analyzed data. The voice conversion unit is realized, for example, by the control unit 46A of the robot 414, and converts the generated response into voice data and provides it to the customer. The correspondence between each unit and the device or control unit is not limited to the above-mentioned example, and various modifications are possible.

The emotion identification model 59, which serves as an emotion engine, may determine the emotion of the user according to a specific mapping. Specifically, the emotion identification model 59 may determine the emotion of the user according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the emotion of the robot, and the identification processing unit 290 may perform identification processing using the emotion of the robot.

9 is a diagram showing an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive emotions are arranged. Emotions that represent states and actions arising from a state of mind are arranged on the outer sides of the concentric circles. Emotions are a concept that includes emotions and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions that occur in the brain are arranged. On the right side of the concentric circles, emotions that are generally induced by situational judgment are arranged. On the upper and lower sides of the concentric circles, emotions that are generally generated from reactions that occur in the brain and are induced by situational judgment are arranged. In addition, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

These emotions are distributed in the three o'clock direction of emotion map 400, and usually fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness takes precedence over internal sensations, resulting in a sense of calm.

The inside of emotion map 400 represents what is going on inside the mind, and the outside of emotion map 400 represents behavior, so the further out you go on emotion map 400, the more visible (expressed in behavior) the emotions become.

Here, human emotions are based on various balances such as posture and blood sugar level, and when these balances are far from the ideal, it indicates an unpleasant state, and when they are close to the ideal, it indicates a pleasant state. Emotions can also be created for robots, cars, motorcycles, etc., based on various balances such as posture and remaining battery power, so that when these balances are far from the ideal, it indicates an unpleasant state, and when they are close to the ideal, it indicates a pleasant state. The emotion map may be generated, for example, based on the emotion map of Dr. Mitsuyoshi (Research on speech emotion recognition and emotion brain physiological signal analysis system, Tokushima University, doctoral dissertation: https://ci.nii.ac.jp/naid/500000375379). The left half of the emotion map is lined with emotions that belong to an area called "reaction" where sensation is dominant. The right half of the emotion map is lined with emotions that belong to an area called "situation" where situation recognition is dominant.

The emotion map defines two emotions that encourage learning. The first is the negative emotion around the middle of "repentance" or "reflection" on the situation side. In other words, this is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the positive emotion around "desire" on the response side. In other words, this is when the robot has positive feelings such as "I want more" or "I want to know more."

The emotion identification model 59 inputs user input to a pre-trained neural network, obtains emotion values indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple learning data that are combinations of user input and emotion values indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in Figure 10. Figure 10 shows an example in which multiple emotions, "peace of mind," "calm," and "reassuring," have similar emotion values.

In the above embodiment, an example was given in which a specific process is performed by one computer 22, but the technology disclosed herein is not limited to this, and distributed processing of the specific process may be performed by multiple computers, including computer 22.

In the above embodiment, an example has been described in which the specific processing program 56 is stored in the storage 32, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a Universal Serial Bus (USB) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

The specific processing program 56 may also be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

The various processors listed below can be used as hardware resources for executing specific processes. Examples of processors include a CPU, which is a general-purpose processor that functions as a hardware resource for executing specific processes by executing software, i.e., a program. Examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which are processors with a circuit configuration designed specifically to execute specific processes. All of these processors have built-in or connected memory, and all of these processors execute specific processes by using the memory.

The hardware resource that executes the specific process may be composed of one of these various processors, or may be composed of a combination of two or more processors of the same or different types (e.g., a combination of multiple FPGAs, or a combination of a CPU and an FPGA). The hardware resource that executes the specific process may also be a single processor.

As an example of a configuration using a single processor, first, there is a configuration in which one processor is configured by combining one or more CPUs with software, and this processor functions as a hardware resource that executes a specific process. Secondly, there is a configuration in which a processor is used that realizes the functions of the entire system, including multiple hardware resources that execute a specific process, on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the various processors mentioned above as hardware resources.

More specifically, the hardware structure of these various processors can be an electric circuit that combines circuit elements such as semiconductor elements. The specific processing described above is merely an example. It goes without saying that unnecessary steps can be deleted, new steps can be added, and the processing order can be changed without departing from the spirit of the invention.

In the above example, the first to fourth embodiments have been described separately, but some or all of these embodiments may be combined. Also, the smart device 14, smart glasses 214, headset terminal 314, and robot 414 are only examples, and they may be combined with each other, or may be other devices. Also, in the above example, the first and second embodiments have been described separately, but these may be combined.

The above description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure, and are merely an example of the technology of the present disclosure. For example, the above explanation of the configuration, function, action, and effect is an explanation of an example of the configuration, function, action, and effect of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above description and illustrations, within the scope of the gist of the technology of the present disclosure. Also, in order to avoid confusion and to make it easier to understand the parts related to the technology of the present disclosure, the above description and illustrations omit explanations of technical common sense that do not require particular explanation to enable the implementation of the technology of the present disclosure.

All publications, patent applications, and technical standards mentioned in this specification are incorporated by reference into this specification to the same extent as if each individual publication, patent application, and technical standard was specifically and individually indicated to be incorporated by reference.

(Appendix 1)
an analysis unit that analyzes the voice data;
a generation unit that generates a response based on the data analyzed by the analysis unit;
a voice generation unit that voices the response generated by the generation unit;
A system comprising:
(Appendix 2)
The generation unit is
The system according to claim 1, further comprising an adjustment unit for fine tuning.
(Appendix 3)
The voice conversion unit is
The system according to claim 1, further comprising a providing unit for providing the generated voice to a customer.
(Appendix 4)
The analysis unit is
2. The system according to claim 1, further comprising: analyzing a plurality of pieces of voice data and modeling the tempo and intonation of the voice.
(Appendix 5)
The generation unit is
The system described in claim 1, characterized in that it uses a generative AI that has knowledge about a specific business or service.
(Appendix 6)
The analysis unit is
The system according to claim 1, further comprising: estimating a user's emotion; and adjusting a method of analyzing the voice data based on the estimated user's emotion.
(Appendix 7)
The analysis unit is
2. The system of claim 1, further comprising: a processor configured to generate a speech data stream for speech recognition based on a particular accent or dialect;
(Appendix 8)
The analysis unit is
2. The system according to claim 1, further comprising a filter process for removing background noise when analyzing voice data.
(Appendix 9)
The analysis unit is
The system according to claim 1, further comprising: estimating a user's emotion; and determining a priority order of voice data to be analyzed based on the estimated user's emotion.
(Appendix 10)
The analysis unit is
The system of claim 1, further comprising: adjusting an analysis method based on a user's geographic location information when analyzing voice data.
(Appendix 11)
The analysis unit is
The system of claim 1, further comprising: when analyzing voice data, analyzing relevant voice data preferentially based on the user's social media activity.
(Appendix 12)
The generation unit is
The system according to claim 1, further comprising: estimating a user's emotion; and adjusting a reply expression method based on the estimated user's emotion.
(Appendix 13)
The generation unit is
The system according to claim 1, further comprising: a step of adjusting a level of detail of a reply based on the importance of the inquiry content when generating the reply.
(Appendix 14)
The generation unit is
2. The system of claim 1, wherein when generating a response, different generation algorithms are applied depending on the category of the query.
(Appendix 15)
The generation unit is
The system of claim 1, further comprising: estimating a user's emotion; and adjusting a length of the reply based on the estimated user's emotion.
(Appendix 16)
The generation unit is
2. The system of claim 1, wherein when generating a response, the response is prioritized based on when the query was submitted.
(Appendix 17)
The generation unit is
2. The system of claim 1, wherein when generating responses, the order of responses is adjusted based on the relevance of the query.
(Appendix 18)
The voice conversion unit is
2. The system of claim 1, further comprising: estimating a user's emotion; and adjusting a voice expression method based on the estimated user's emotion.
(Appendix 19)
The voice conversion unit is
2. The system of claim 1, further comprising: performing voice filtering during voice generation to improve the naturalness of the generated voice.
(Appendix 20)
The voice conversion unit is
2. The system of claim 1, further comprising: a voice generating system that takes into account specific accents or dialects to improve voice generation accuracy.
(Appendix 21)
The voice conversion unit is
The system of claim 1, further comprising: estimating a user's emotion; and determining a priority of speech generation based on the estimated user's emotion.
(Appendix 22)
The voice conversion unit is
The system according to claim 1, further comprising: a voice generation method that takes into account a user's geographic location information during voice generation.
(Appendix 23)
The voice conversion unit is
The system of claim 1, further comprising: analyzing a user's social media activity and prioritizing the conversion of relevant audio data to audio during conversion.
(Appendix 24)
The adjustment unit is
3. The system of claim 2, further comprising: estimating a user's emotion; and adjusting fine-tuning parameters based on the estimated user's emotion.
(Appendix 25)
The adjustment unit is
The system according to claim 2, wherein during fine tuning, the generation algorithm is optimized by referring to past query data.
(Appendix 26)
The adjustment unit is
3. The system of claim 2, further comprising: estimating a user's emotion; and adjusting a frequency of fine-tuning based on the estimated user's emotion.
(Appendix 27)
The adjustment unit is
3. The system of claim 2, wherein during fine tuning, the training data is weighted based on the time of query submission.
(Appendix 28)
The providing unit is
The system of claim 3, further comprising: estimating a user's emotion; and adjusting a manner of providing audio based on the estimated user's emotion.
(Appendix 29)
The providing unit is
The system according to claim 3, wherein when providing voice, the system refers to the user's past inquiry history to select the optimal method of providing the voice.
(Appendix 30)
The providing unit is
The system according to claim 3, further comprising: estimating a user's emotion; and determining a priority of audio provision based on the estimated user's emotion.
(Appendix 31)
The providing unit is
The system according to claim 3, wherein when providing voice, the system selects an optimal method of providing voice by taking into consideration device information of the user.

10, 210, 310, 410 Data processing system 12 Data processing device 14 Smart device 214 Smart glasses 314 Headset type terminal 414 Robot

Claims (10)

an analysis unit that analyzes the voice data;
a generation unit that generates a response based on the data analyzed by the analysis unit;
a voice generation unit that voices the response generated by the generation unit;
A system comprising: The generation unit is
The system according to claim 1 , further comprising an adjustment unit for fine tuning. The voice conversion unit is
The system according to claim 1 , further comprising a providing unit for providing the generated voice to a customer. The analysis unit is
2. The system according to claim 1, further comprising: analyzing a plurality of pieces of voice data to model the tempo and intonation of the voice. The generation unit is
The system of claim 1, further comprising: a generative AI having knowledge of a particular business or service. The analysis unit is
The system according to claim 1 , further comprising: estimating a user's emotion; and adjusting a method of analyzing the voice data based on the estimated user's emotion. The analysis unit is
2. The system of claim 1, wherein when analyzing speech data, a particular accent or dialect is taken into account to improve analysis accuracy. The analysis unit is
2. The system according to claim 1, further comprising a filtering process for removing background noise when analyzing the voice data. The analysis unit is
The system according to claim 1 , further comprising: estimating a user's emotion; and determining a priority order of voice data to be analyzed based on the estimated user's emotion. The analysis unit is
The system of claim 1, wherein when analyzing the voice data, the analysis method is adjusted based on the user's geographic location information.

Images