KR102504445B1 - System and method for supporting artificial intelligence speech recognition and conversation services - Google Patents

Abstract

The present invention relates to a system for supporting artificial intelligence speech recognition and conversation services, comprising: an AI chatbot for extracting first response information corresponding to query content included in query speech of a speaker from a pre-built learning data set and providing the same; and a cloud response server for deriving second response information appropriate for the query content included in the query speech of the speaker transmitted from the AI chatbot, using at least one NLP engine, and providing the derived second response information to the AI chatbot, wherein the AI chatbot compares the first response information and the second response information to select and enunciate response information having delay time and response accuracy meeting preset conditions. Therefore, immediate response without delay in response speed is enabled.

Description

System and method for supporting artificial intelligence speech recognition and conversation services}

The present invention relates to a system and method for supporting artificial intelligence voice recognition and conversation services.

For advanced technology to become a part of daily life, an intuitive and free interface is becoming a must. A more advanced user interface is required for next-generation service robots that aim for advanced intelligent services but must be operated at a certain distance from customers.

However, in the case of newly proposed gestures, kinect interfaces, gaze recognition or interfaces that recognize electrical signals in the brain, robots have not yet passed the demonstration stage, and call center control and voice remain in limited applications. Recognition interface also has technical limitations in the use of robots.

The emergence of smart speakers over the past several years has shown the possibility of expanding smart devices using voice, and has emerged as a key platform leading voice conversation services.

However, negative perceptions such as inaccurate recognition, misinterpretation, limited vocabulary, and slow cloud service speed of smart devices that provide voice interfaces act as obstacles to adopting voice recognition as a user interface for new smart devices. Beyond trial and error, the technology to implement a digital assistant through an indoor service robot is still unstable.

Publication No. 10-2017-0093629 (Title of Invention: Voice Recognition Apparatus and Method, Voice Recognition System)

An object of the present invention is to provide a system and method for supporting artificial intelligence voice recognition and conversation services that can solve the conventional problems.

In order to solve the above problem, a system supporting artificial intelligence voice recognition and conversation service according to an embodiment of the present invention converts first response information matching the query content included in the speaker's query voice from a pre-built learning data set. AI chatbot that extracts and provides; And a cloud response server for deriving second response information suitable for the inquiry contents included in the questioning voice of the speaker transmitted from the AI chatbot using at least one NLP engine, and then providing the derived second response information to the AI chatbot. Including, the AI chatbot compares the first response information and the second response information and selects and utters response information having a delay time and response accuracy that meets a preset condition. A query included in the speaker's voice information a wake-up unit requesting a response to the query content (query word) when detecting content (query word); a local response provider extracting and providing response information corresponding to a response to the query content (query language) from a learned learning data set; And a processor unit composed of a response waiting manager unit that compares response information derived from the local response provider and the cloud response server, selects and ignites response information having a delay time and response accuracy that meets a preset condition, and , When the response standby manager detects rejection or negative language for response information spoken by the wakeup unit, ignites unselected response information, and when the wakeup unit does not detect rejection or negative language within a preset time, the It is characterized in that unselected response information is deleted.
In one embodiment, the cloud response server selects one of at least one NLP engine for receiving response information suitable for the speaker's query content (query language) from the AI chatbot, and then the response derived from the selected NLP engine It is characterized by providing information to an AI chatbot.

delete

delete

delete

delete

delete

delete

delete

delete

In one embodiment, the cloud response server derives response information from each of at least one or more NLP engines according to a round robin fashion, and then selects response information with the highest accuracy among the derived response information to generate an AI chatbot. It is characterized by providing as.

In one embodiment, the cloud response server provides response information in the form of at least one of an automatic speech recognition (ASR) response, a natural language understanding (NLU) response, and a text to speech (TTS) response. It is characterized by doing.

In one embodiment, the cloud response server derives response information from each of at least one NLP engine, analyzes the accuracy rate of the derived response information, and if the analyzed accuracy rate does not meet a predetermined standard, the corresponding NLP engine It is characterized by processing to perform self-learning for.

A method for supporting an artificial intelligence voice recognition and conversation service according to an embodiment of the present invention for solving the above problems includes detecting user's voice information (speech information) in an AI chatbot; After converting the voice information (speech information) into text in the AI chatbot, the query content in the text is analyzed to extract response information matching the query content from the learned data set, and at the same time, voice information to an external cloud response server providing text information of (speech information); And the AI chatbot compares the response information provided from the cloud response server with the response information extracted from the learning data set built in the AI chatbot, selects response information having delay time and accuracy of response that meets preset conditions, and then utters and, after the uttering step, uttering unselected response information when a user's rejection or negative language is detected for the uttered response information; and deleting the non-selected response information if the user's rejection or negative language is not detected within a preset time, wherein the AI chatbot compares the first response information and the second response information to meet a preset condition. a wake-up unit requesting a response to the query content (query word) included in the speaker's voice information so as to select and utter response information having a delay time and accuracy of response; a local response provider extracting and providing response information corresponding to a response to the query content (query language) from a learned learning data set; and a processor unit configured of a response waiting manager unit that compares response information derived from the local response providing unit and the cloud response server, selects and ignites response information having delay time and response accuracy that meet preset conditions. .

delete

delete

delete

delete

Therefore, using the system and method for supporting artificial intelligence voice recognition and conversation service according to an embodiment of the present invention, there is an advantage that immediate response is possible without delay in response speed.

In addition, if there is an appropriate response among the response information analyzed in each of the local NLP solution supported by the AI chatbot and the set cloud NLP solution, there is an advantage that one of them can be selected and converted into voice data immediately and then output to the user. .

Furthermore, in addition to enabling immediate conversation service with minimal delay, the operating system uses a local NLP engine even in a network or network blackout situation to provide a conversation service between a user and an AI chatbot. has the advantage of being possible.

1 is a network configuration diagram of a system supporting artificial intelligence voice recognition and conversation service according to an embodiment of the present invention.
FIG. 2 is a detailed configuration diagram of the chatbot shown in FIG. 1 .
FIG. 3 is a detailed configuration diagram of a processor unit shown in FIG. 2 .
FIG. 4A is an exemplary view illustrating processes of creation, utterance, and termination of a local conversation session of the AI chatbot shown in FIG. 1 .
FIG. 4B is an exemplary view illustrating an operation process of the cloud server shown in FIG. 1 .
5 is an exemplary diagram illustrating a process of selecting response information uttered by the AI chatbot shown in FIG. 1.
6 is a flowchart illustrating a method for supporting an artificial intelligence voice recognition and conversation service according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail so that those skilled in the art can easily practice the present invention with reference to the accompanying drawings. However, the present invention may be embodied in many different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

Throughout the specification, when a part is said to be "connected" to another part, this includes not only the case where it is "directly connected" but also the case where it is "electrically connected" with another element interposed therebetween. . In addition, when a part "includes" a certain component, this means that it may further include other components, not excluding other components, unless otherwise stated, and one or more other characteristics. However, it should be understood that it does not preclude the possibility of existence or addition of numbers, steps, operations, components, parts, or combinations thereof.

As used throughout the specification, the terms "about", "substantially", etc., are used at or approximating that value when manufacturing and material tolerances inherent in the stated meaning are given, and do not convey an understanding of the present invention. Accurate or absolute figures are used to help prevent exploitation by unscrupulous infringers of the disclosed disclosure. The term "step of (doing)" or "step of" as used throughout the specification of the present invention does not mean "step for".

In this specification, a "unit" includes a unit realized by hardware, a unit realized by software, and a unit realized using both. Further, one unit may be realized using two or more hardware, and two or more units may be realized by one hardware. On the other hand, '~ unit' is not limited to software or hardware, and '~ unit' may be configured to be in an addressable storage medium or configured to reproduce one or more processors. Thus, as an example, '~unit' refers to components such as software components, object-oriented software components, class components, and task components, processes, functions, properties, and procedures. , subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays and variables. Functions provided within components and '~units' may be combined into smaller numbers of components and '~units' or further separated into additional components and '~units'. In addition, components and '~units' may be implemented to play one or more CPUs in a device or a secure multimedia card.

In this specification, some of the operations or functions described as being performed by a terminal, device, or device may be performed instead by a server connected to the terminal, device, or device. Likewise, some of the operations or functions described as being performed by the server may also be performed by a terminal, apparatus, or device connected to the server.

In this specification, some of the operations or functions described as mapping or matching with the terminal mean mapping or matching the terminal's unique number or personal identification information, which is the terminal's identifying data. can be interpreted as

Hereinafter, a system and method for supporting an artificial intelligence voice recognition conversation service according to an embodiment of the present invention will be described in more detail based on the accompanying drawings.

1 is a network configuration diagram of a system supporting artificial intelligence voice recognition and conversation service according to an embodiment of the present invention, FIG. 2 is a detailed configuration diagram of the chatbot shown in FIG. 1, and FIG. 3 is shown in FIG. Figure 4a is an example diagram explaining the process of creating, uttering, and ending a local conversation session of the AI chatbot shown in FIG. 1, and FIG. 5 is an exemplary diagram illustrating a process of selecting response information uttered by the AI chatbot shown in FIG. 1.

First, as shown in FIG. 1, the system 100 supporting artificial intelligence voice recognition and conversation service according to an embodiment of the present invention is composed of an AI chatbot 200 and a cloud server 300, each component When communicating through a network, the network refers to a connection structure capable of exchanging information between nodes such as a plurality of terminals and servers. Examples of such networks include a local area network (LAN) and a wide area network. It includes a wide area network (WAN), a World Wide Web (WWW), a wired and wireless data communication network, a telephone network, and a wired and wireless television communication network. Examples of wireless data communication networks include 3G, 4G, 5G, 3rd Generation Partnership Project (3GPP), 5th Generation Partnership Project (5GPP), Long Term Evolution (LTE), World Interoperability for Microwave Access (WIMAX), Wi-Fi , Internet (Internet), LAN (Local Area Network), Wireless LAN (Wireless Local Area Network), WAN (Wide Area Network), PAN (Personal Area Network), RF (Radio Frequency), Bluetooth (Bluetooth) network, NFC ( A Near-Field Communication (Near-Field Communication) network, a satellite broadcasting network, an analog broadcasting network, a Digital Multimedia Broadcasting (DMB) network, etc. are included, but not limited thereto.

The AI chatbot 200 compares the first response information extracted from the learning data set corresponding to the response to the query included in the speaker's voice information and the response information to the query from the external cloud server. It may be configured to select and ignite response information having delay time and accuracy of response that meet a preset condition.

In addition, the AI chatbot 200 may be configured to classify and classify talkers based on voice information, learn the query form and query language used by each talker, and store the learned learning data set in a DB.

The AI chatbot 200 may include a processor unit 210, a memory 220 and a communication module 230.

The processor unit 210 may include one or more application processors (APs), one or more communication processors (CPs), or one or more AI processors (artificial intelligence processors). The application processor, communication processor, or AI processor may be each included in different integrated circuit (IC) packages or included in one IC package.

The application processor may control a plurality of hardware or software components connected to the application processor by driving an operating system or an application program, and may process/calculate various data including multimedia data. For example, the application processor may be implemented as a system on chip (SoC). The processor unit 210 may further include a graphic processing unit (GPU).

The communication processor may perform functions of managing a data link and converting a communication protocol in communication with the cloud response server 300 connected through a network. For example, the communication processor may be implemented as an SoC. The communications processor may perform at least part of the multimedia control function.

Also, the communication processor may control data transmission and reception of the communication module 230 . The communication processor may be implemented to be included as at least a part of the application processor.

The application processor or communication processor may load a command or data received from at least one of a non-volatile memory or other components connected thereto into the volatile memory and process the load. Also, the application processor or communication processor may store data received from at least one of the other components or generated by at least one of the other components in a non-volatile memory.

More specifically, the processor unit 210 may include a wakeup unit 211, a local response providing unit 212, and a response standby manager unit 213.

The wake-up detection unit 211 may be configured to generate a voice conversation session for the query and request a response when it detects a query included in voice information of a talker.

The local response providing unit 212 may be configured to extract and provide response information corresponding to a response to the query from a learning data set built in an artificial intelligence robot.

Here, the local response provider 220 may learn the neural network using a program stored in the memory 230 . In particular, a neural network for recognizing data related to query responses may be trained. Here, the neural network may be designed to simulate a human brain structure (eg, a neuron structure of a human neural network) on a computer. A neural network may include an input layer, an output layer, and at least one hidden layer. Each layer may include at least one neuron having a weight, and the neural network may include neurons and synapses connecting the neurons. In the neural network, each neuron may output an input signal input through a synapse as a function value of an activation function for weight and/or bias.

A plurality of network modes may transmit and receive data according to respective connections so as to simulate synaptic activity of neurons that transmit and receive signals through synapses. Here, the neural network may include a deep learning model developed from a neural network model. In the deep learning model, a plurality of network nodes may exchange data according to a convolutional connection relationship while being located in different layers. Examples of neural network models include deep neural networks (DNNs), convolutional neural networks (CNNs), recurrent neural networks, restricted Boltzmann machines, and deep belief networks. ), deep Q-Network, etc., and can be applied to fields such as vision recognition, voice recognition, natural language processing, and voice/signal processing.

That is, the local response providing unit 212 may include a data learning unit that learns a neural network for data classification/recognition. The data learning unit may learn criteria regarding which training data to use to determine data classification/recognition and how to classify and recognize data using the training data. The data learning unit may acquire learning data to be used for learning and learn the deep learning model by applying the obtained learning data to the deep learning model. The data learning unit may be manufactured and mounted in the form of at least one hardware chip. For example, the data learning unit may be manufactured in the form of a dedicated hardware chip for artificial intelligence, and may be manufactured and mounted as a part of a general-purpose processor (CPU) or a graphics-only processor (GPU). Also, the data learning unit may be implemented as a software module. When implemented as a software module (or a program module including instructions), the software module may be stored in a computer-readable, non-transitory computer readable recording medium (non-transitory computer readable media). In this case, at least one software module may be provided to an operating system (OS) or may be provided by an application.

The data learning unit may learn to have a criterion for determining how to classify/recognize predetermined data by using the acquired training data. At this time, the learning method by the data learning unit may be classified into supervised learning, unsupervised learning, and reinforcement learning. Here, supervised learning refers to a method of learning an artificial neural network given a label for training data, and a label is an answer (or a result value) that the artificial neural network must infer when learning data is input to the artificial neural network. can mean Unsupervised learning may refer to a method of training an artificial neural network in a state in which a label for training data is not given. Reinforcement learning may refer to a method of learning to select an action or an action sequence that maximizes a cumulative reward in each state by an agent defined in a specific environment.

In addition, the data learning unit may train the neural network model using a learning algorithm including error backpropagation or gradient descent. When the neural network model is trained, the trained neural network model may be referred to as a learning model. The learning model may be stored in the memory 230 and used to infer results for new input data other than training data.

The response waiting manager unit 213 compares the response information transmitted from the local response providing unit 212 and the cloud response server 300, and the response information having delay time and accuracy of response that meet preset conditions. It may be configured to ignite by selecting.

The response waiting manager unit 213 may include a function of uttering non-selected response information when it detects rejection or negative language for response information uttered by the wake-up unit.

In addition, the response waiting manager unit 213 may include a function of deleting the unselected response information when the wakeup detection unit 211 does not detect rejection or negative language within a preset time.

Meanwhile, the AI chatbot 200 mentioned in the present invention is an intelligent virtual assistant, a virtual personal assistant, an intelligent personal assistant, a conversational agent, a virtual companion, Even if it is used as a term such as virtual assistant, it may be included in the scope of the chatbot of the present invention.

In addition, the AI chatbot 200 may perform functions of natural language processing (NLP) and natural language generation (NLG) through a reinforcement learning algorithm in the field of artificial intelligence. A chatbot can understand human speech and text, analyze the data it has or has learned, and respond with words and text that humans can understand. It can process written language and can even make logical inferences. In addition, chatbots can find information requested by users from unstructured data, and can provide intelligent services that inform and utilize real situations.

To this end, the chatbot according to an embodiment of the present invention is a pattern recognition (Pattern Recognition) in which a machine identifies figures, texts, voices, etc., information search, question and answer, system so that a machine recognizes a language commonly used by humans Natural Language Processing, which includes automatic translation and interpretation functions; Semantic Web, which can understand the meaning of information data and even make logical inferences; and Text Mining, which finds new and useful information from unstructured text data. Text Mining), it can perform the function of context-aware computing that informationizes real situations in virtual space and provides user-centered intelligent services by utilizing them.

At this time, in one embodiment of the present invention, a character may be given to the chatbot, that is, an identity may be implanted so as to have various personalities. It has been revealed through various studies that the probability of recognizing a chatbot as a person increases when an identity is implanted in it. Humans endowing robots with personalities are helpful because robot users can more easily understand their behavior and enable more friendly interactions, which can lead to higher preference.

Various methods exist as a personality type analysis tool, but in one embodiment of the present invention, the DISC model can be used to classify types through behaviors that are easily revealed rather than internal factors such as personal temperament that are difficult to observe. This is because the DISC model is suitable for identifying the personality suitable for artificial intelligence because it is easy to grasp the overall preference of users due to its simple classification. The two dimensions that are the criteria for DISC analysis are object (day/person) and speed (slow/fast).

Personality types can be divided into four types: Dominance (D), Influences (I: Influences), Conscintiousness (C), and Steadiness (S). For example, D-types who are driven are work-oriented and value quick decisions and momentum.

Social type I type is good at influencing others, active and fun personality.

Type C is a prudent type, quiet, analytical, and precise.

Type S is a stable, relaxed and rational personality. These personality types can be assumed to fit the role of the chatbot.

Let's assume that D-type chatbots can be suitable for task-oriented chatbots where fast and accurate work processing is important, sociable I-types can be suitable as companions when bored, and cautious C-types or stable S-types can be suitable as counseling chatbots. can In addition, even if the character is defined as described above, the user's taste and preference according to the type may be reflected and applied.

In addition, an embodiment of the present invention may further include anthropomorphism, which is an interaction design that makes the interactive agent feel like a human. At this time, the quality of interaction between humans and agents can be improved by eliciting positive social responses using studies related to Computers Are Social Actors (CASA).

The memory 220 may include a built-in memory or an external memory. The built-in memory includes volatile memory (eg, dynamic RAM (DRAM), static RAM (SRAM), synchronous dynamic RAM (SDRAM), etc.) or non-volatile memory (eg, one time programmable ROM (OTPROM)), It may include at least one of programmable ROM (PROM), erasable and programmable ROM (EPROM), electrically erasable and programmable ROM (EEPROM), mask ROM, flash ROM, NAND flash memory, NOR flash memory, etc.). According to one example, the embedded memory may take the form of a solid state drive (SSD). The external memory is a flash drive, for example, compact flash (CF), secure digital (SD), micro secure digital (Micro-SD), mini secure digital (Mini-SD), extreme digital (xD), or It may include a memory stick and the like.

Next, the communication module 230 may be configured to request a response to the query from an external cloud response server 300 and receive response information from the cloud response server.

The communication module 230 may include a wireless communication module or an RF module. The wireless communication module may include, for example, Wi-Fi, BT, GPS or NFC. For example, the wireless communication module may provide a wireless communication function using a radio frequency. Additionally or alternatively, the wireless communication module includes a network interface or modem for connecting the user device 100 to a network (eg, Internet, LAN, WAN, telecommunication network, cellular network, satellite network, POTS or 5G network, etc.) can include

The RF module may be responsible for transmitting and receiving data, for example, transmitting and receiving RF signals or called electronic signals. For example, the RF module may include a transceiver, a power amp module (PAM), a frequency filter, or a low noise amplifier (LNA). In addition, the RF module may include components for transmitting and receiving electromagnetic waves in free space in wireless communication, for example, conductors or wires.

Next, the cloud response server 300 selects any one of at least one NLP engine that provides response information suitable for the query included in the voice information of the speaker from the AI chatbot 200, and then selects the selected NLP engine It may be configured to provide the response information derived from to the AI chatbot 200.

In addition, the cloud response server 300 derives response information from each of at least one NLP engine according to a round robin fashion, and selects response information with the highest accuracy among the derived response information to generate an AI chatbot. It may be a configuration provided as (200).

The cloud response server 300 may provide response information in the form of at least one of an automatic speech recognition (ASR) response, a natural language understanding (NLU) response, and a text to speech (TTS) response. .

In the automatic voice recognition response, a response to a query is provided only in a fixed form of information, a natural language interpretation response is a response to information about a result of natural language interpretation, and a TTS response is a response based on TTS technology. It may be a response to information converted into a voice signal.

In addition, the cloud response server 300 derives response information from each of the at least one NLP engine, analyzes the accuracy rate of the derived response information, and if the analyzed accuracy rate does not meet a preset standard, the corresponding NLP model It may include a function of processing to perform self-learning for.

6 is a flowchart illustrating a method for supporting an artificial intelligence voice recognition and conversation service according to an embodiment of the present invention.

As shown in FIG. 6, the method for supporting artificial intelligence voice recognition and conversation service according to an embodiment of the present invention (S700) detects user's voice information (speech information) in the AI chatbot 200 (S710) Then, after converting voice information (speech information) into text, analyzing the content of the query in the text, extracting response information matching the content of the query from the learned data set, and at the same time to the external cloud response server (300) Text information of voice information (speech information) is provided (S720).

Thereafter, the AI chatbot 200 receives response information corresponding to the query content of the voice information (speech information) provided by the cloud response server 300, and then first response information extracted from the learning data set built in the AI chatbot and second response information of an external cloud server, select response information having a delay time and response accuracy that meets a preset condition, and then convert the response information into voice information and speak (S730).

In addition, after the uttering step, if a user's rejection or negative language is detected for the uttered response information, a process of uttering unselected response information may be further included.

The method may further include deleting the non-selected response information if the user's rejection or negative language is not detected within a preset time after the utterance step.

Therefore, using the system and method for supporting artificial intelligence voice recognition and conversation service according to an embodiment of the present invention, there is an advantage that immediate response is possible without delay in response speed.

In addition, if there is an appropriate response among the response information analyzed in each of the local NLP solution supported by the AI chatbot and the set cloud NLP solution, there is an advantage that one of them can be selected and converted into voice data immediately and then output to the user. .

Furthermore, in addition to enabling immediate conversation service with minimal delay, the operating system uses a local NLP engine even in a network or network blackout situation to provide a conversation service between a user and an AI chatbot. has the advantage of being possible.

“~ unit” used in one embodiment of the present invention may be implemented as a hardware component, a software component, and/or a combination of hardware components and software components. For example, devices and components described in the embodiments may include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA), It may be implemented using one or more general purpose or special purpose computers, such as a programmable logic unit (PLU), microprocessor, or any other device capable of executing and responding to instructions. A processing device may run an operating system (OS) and one or more software applications running on the operating system. A processing device may also access, store, manipulate, process, and generate data in response to execution of software. For convenience of understanding, there are cases in which one processing device is used, but those skilled in the art will understand that the processing device includes a plurality of processing elements and/or a plurality of types of processing elements. It can be seen that it can include. For example, a processing device may include a plurality of processors or a processor and a controller. Other processing configurations are also possible, such as parallel processors.

Software may include a computer program, code, instructions, or a combination of one or more of the foregoing, which configures a processing device to operate as desired or processes independently or collectively. You can command the device. Software and/or data may be any tangible machine, component, physical device, virtual equipment, computer storage medium or device, intended to be interpreted by or to provide instructions or data to a processing device. , or may be permanently or temporarily embodied in a transmitted signal wave. Software may be distributed on networked computer systems and stored or executed in a distributed manner. Software and data may be stored on one or more computer readable media.

The method according to the embodiment of the present invention may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer readable medium. The computer readable medium may include program instructions, data files, data structures, etc. alone or in combination. Program commands recorded on the medium may be specially designed and configured for the embodiment or may be known and usable to those skilled in computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic media such as floptical disks. - includes hardware devices specially configured to store and execute program instructions, such as magneto-optical media, and ROM, RAM, flash memory, and the like. Examples of program instructions include high-level language codes that can be executed by a computer using an interpreter, as well as machine language codes such as those produced by a compiler. The described hardware devices may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

The foregoing may be modified and modified by those skilled in the art without departing from the essential characteristics of the present invention. Therefore, the embodiments disclosed in the present invention are not intended to limit the technical idea of the present invention, but to explain, and the scope of the technical idea of the present invention is not limited by these embodiments. The protection scope of the present invention should be construed according to the claims below, and all technical ideas within the equivalent range should be construed as being included in the scope of the present invention.

100: system
200: AI chatbot
210: processor unit
211: wake-up detection unit
212: local response providing unit
213: response waiting manager unit
220: memory
230: communication department
300: cloud response server

Claims (11)

an AI chatbot extracting and providing first response information corresponding to the inquiry contents included in the speaker's query voice from a pre-built learning data set; and
A cloud response server for deriving second response information suitable for the inquiry content included in the query voice of the speaker delivered by the AI chatbot using at least one NLP engine, and then providing the derived second response information to the AI chatbot include,
The AI chatbot compares the first response information and the second response information to select and utter response information having a delay time and response accuracy that meets a preset condition,
a wake-up unit requesting a response to the query content (query word) when detecting the query content (query word) included in the speaker's voice information;
a local response provider extracting and providing response information corresponding to a response to the query content (query language) from a learned learning data set; and
A processor unit configured of a response waiting manager unit that compares response information derived from the local response providing unit and the cloud response server, selects and ignites response information having a delay time and response accuracy that meets a preset condition,
The response waiting manager unit
When the wake-up unit detects rejection or negative language for response information uttered, unselected response information is ignited,
The system for supporting an artificial intelligence voice recognition conversation service, characterized in that, if the wake-up unit does not detect rejection or negative language within a predetermined time, the unselected response information is deleted.
delete delete delete According to claim 1,
The cloud response server
After selecting any one of at least one NLP engine for receiving response information suitable for the speaker's query content (query language) from the AI chatbot, response information derived from the selected NLP engine is provided to the AI chatbot. A system that supports artificial intelligence voice recognition conversation service.
According to claim 1,
The cloud response server
After deriving response information from each of at least one NLP engine according to a round robin fashion, selecting the response information with the highest accuracy among the derived response information and providing it to an AI chatbot Artificial intelligent voice, characterized in that A system that supports cognitive conversation services.
According to claim 1,
The cloud response server
An artificial intelligence voice recognition conversation characterized by providing at least one type of response information among an automatic speech recognition (ASR) response, a natural language understanding (NLU) response, and a text to speech (TTS) response. systems that support the service.
According to claim 7,
The cloud response server
After deriving response information from each of at least one NLP engine, analyzing the accuracy rate of the derived response information, and processing to perform self-learning for the corresponding NLP engine when the analyzed accuracy rate does not meet a predetermined standard. A system that supports an artificial intelligence voice recognition conversation service characterized by
Detecting the user's voice information (speech information) in the AI chatbot;
After converting the voice information (speech information) into text in the AI chatbot, the query content in the text is analyzed to extract response information matching the query content from the learned data set, and at the same time, voice information to an external cloud response server providing text information of (speech information); and
In the AI chatbot, the response information provided from the cloud response server and the response information extracted from the learning data set built in the AI chatbot are compared, and response information having delay time and response accuracy that meets preset conditions is selected, and then uttered contains steps,
After the ignition step,
uttering non-selected response information when a user's rejection or negative language is detected for the uttered response information; and
Deleting the unselected response information if the user's rejection or negative language is not detected within a preset time,
The AI chatbot compares the first response information and the second response information to select and utter response information having a delay time and response accuracy that meets a preset condition,
a wake-up unit requesting a response to the query content (query word) when detecting the query content (query word) included in the speaker's voice information;
a local response provider extracting and providing response information corresponding to a response to the query content (query language) from a learned learning data set; and
An artificial processor including a processor unit including a response standby manager unit that compares response information derived from the local response providing unit and the cloud response server, selects and ignites response information having delay time and response accuracy that meet preset conditions. How to support Intelligent Voice Recognition Conversation Services. delete delete

Images