TWI502582B - Customer service interactive voice system - Google Patents

Description

Service point voice customer service system

The present invention relates to a customer service system, and more particularly to a voice customer service system for a service point.

The familiar telephone customer service system, such as the registration service of the hospital and the ordering of the catering industry, are all after the user calls the customer service center and interacts with the user through the customer service; or plays the pre-recorded voice and the voice on the phone. Button settings, such as the number keys "0~9" and the two symbol keys "*", "#" correspond to the options, interact with the user to achieve the purpose.

For example, when the user calls the telephone customer service system of the hospital, the telephone customer service system will play the pre-recorded voice first, and inform the numeric keys 0~9 and the two symbol keys *, # respectively corresponding to the options, for example 1: registration, 2: query, 9: the service of the switchboard for you. Then the user selects the desired service, for example, press the number key 1 to select the registration number, and then the system enters the registration process. Then the telephone customer service system dials the voice of the registration process, and informs the number key 1: surgery, 2: internal medicine. Then the user presses the corresponding number key according to their needs. In addition, you can also enter some information in these ways, such as the date of the pre-registration, and the user's ID number.

However, a customer service staff cannot handle more than two things at the same time, so if you have to cope with a large number of customer service needs at the same time, you must arrange multiple customer service personnel to deal with them, so that personnel costs will rise sharply. The interactive method of using the telephone button and playing the pre-recorded voice is not convenient in use except for being not humanized. For example, using a mobile phone or a microphone with a button on the same body, You must listen to the pre-recorded voice, then take the phone away from your ear to your eyes before you can touch the button to make a selection or input. You can't listen to the voice and input at the same time, and sometimes you miss the voice. . Furthermore, it is necessary to wait until the pre-recorded voice is played, the user can know the function corresponding to each button, and the user may waste a lot of time to listen to many functions that are not needed before hearing the required function. . In other words, the user must stand in a passive perspective and wait for the customer service system to give relevant information or choices. This lack of user-friendly and inconvenient customer service system needs improvement.

Accordingly, the object of the present invention is to provide a user-friendly and convenient voice interactive voice customer service system.

The technical means adopted by the present invention for solving the problems of the prior art provides a voice service system for a service point. The voice customer service system receives an audio signal sent by a user to confirm the information, and the voice customer service system includes an audio. The capture mechanism, a first conversion mechanism, a composite integration mechanism, and a second conversion mechanism. The audio capture mechanism extracts a first voice audio material from the audio signal. The first conversion mechanism is coupled to the audio capture mechanism for converting the first voice audio material into a first text material. The structure of the integration organization is to integrate the first text of the first conversion institution into a second text. The second conversion mechanism is connected to the combined integration mechanism for converting the second text data into a second voice audio material, and outputting the second voice audio data, wherein the voice customer service system is set at a service point, and the service point is receiving The audio signal sent by the user is used to confirm the information.

In an embodiment of the invention, the first conversion mechanism and the second conversion mechanism are a cloud processing mechanism.

In an embodiment of the invention, the audio capture mechanism extracts the audio signal with respect to a decibel reference value and extracts the first voice audio material.

In an embodiment of the invention, the structural integration mechanism is based on the first The number and/or content of the preset elements included in the text data are combined and integrated.

In an embodiment of the invention, a deconstructing mechanism is further connected to the first conversion mechanism for deconstructing the first text data into a text deconstructing member, and the deconstructed text deconstructing member is transmitted to the decoupling integration mechanism.

In an embodiment of the present invention, when the first text data is less than the preset element of the preset number of items, the first text data is combined and integrated to become a pre-requisite number of content requirements. The second text data of the element is set, and the audio capturing mechanism captures the next audio signal.

In an embodiment of the present invention, when the first text material has a preset element, the first text data is combined and integrated to form a second text material whose content is a reconfirmed preset element.

In an embodiment of the present invention, when the first text data has the preset number of items, the first text data is combined and integrated into a content system to notify the user that the content has been completed. The second text of the voice customer service.

In an embodiment of the invention, the audio signal is an audio signal from a telephone.

In an embodiment of the invention, the service point is a food ordering device.

Through the technical means adopted by the invention, it is possible to save the manpower by not having a customer service person to specifically receive the customer service call. Moreover, the voice customer service system naturally communicates with the user through voice and recognizes the user's needs through voice recognition, and has the advantages of humanization and simple operation for the user. Therefore, it is particularly suitable for applications where customer service needs are large and the user's purpose is clear, such as the catering industry or the medical industry. This will improve the quality of service while reducing costs.

The specific embodiments of the present invention will be further described by the following examples and the accompanying drawings.

100, 100a‧‧‧ voice customer service system

1‧‧‧Audio capture agency

2‧‧‧First conversion agency

3‧‧‧Deconstruction Agency

4‧‧‧Construction Integration Agency

5‧‧‧Second conversion mechanism

6‧‧‧Audio output mechanism

P, P'‧‧‧ service points

S‧‧‧Remote Server

T‧‧‧Phone

T1‧‧‧ receiving end

T2‧‧‧ vocal

U‧‧‧Users

1 is a system diagram showing a voice customer service system of a service point according to a first embodiment of the present invention; and FIG. 2 is a block diagram showing a voice customer service system of a service point according to a first embodiment of the present invention; The figure shows a system diagram of a voice customer service system of a service point according to a second embodiment of the present invention; and FIG. 4 is a schematic diagram showing the operation of a voice customer service system of a service point according to a second embodiment of the present invention.

Referring to FIG. 1 and FIG. 2, the voice customer service system 100 of the service point according to the first embodiment of the present invention is disposed at a service point P for receiving an audio signal sent by a user U and confirming the information. . The service point is located in a place where the telephone customer service is required. In this embodiment, the service point P is a medical voice service device and is installed in the hospital. In other embodiments, the service point can also be a computer and a multifunction. Telephone business machine, etc. The voice customer service system 100 is connected to a telephone T of the U-side of the user, such as a mobile phone or a home phone, and the audio signal sent by the user U is captured by the service point P by the telephone terminal T1.

The voice customer service system 100 includes an audio capture mechanism 1, a first conversion mechanism 2, a deconstruction mechanism 3, a cooperative integration mechanism 4, a second conversion mechanism 5, and an audio output mechanism 6. The audio capture mechanism 1 extracts a first voice audio material from the audio signal. The first converting mechanism 2 is connected to the audio capturing mechanism 1 for converting the first voice audio material into a first text data. The deconstruction mechanism 3 is connected to the first conversion mechanism for deconstructing the first text data into a text deconstructing member, and the deconstructed text deconstructing member is transmitted to the decoupling integration mechanism 4. The structure integration unit 4 connects the deconstruction mechanism 3, and integrates the character deconstruction means deconstructed by the deconstruction mechanism 3 to become the second character material. The second conversion mechanism 5 is connected to the synthesis integration mechanism 4 for converting the second text data into a second voice audio material, and outputting the second voice audio data to the audio output mechanism 6. The audio output mechanism 6 is connected to the second conversion mechanism 5. Outputting the received second voice audio material to a sounding terminal T2 of the telephone T.

In detail, when the user U dials the phone T to the hospital, the voice customer service system 100 installed at the service point P of the hospital (ie, the medical voice service device) communicates with the phone T through the telecommunication line, and then retrieves the user U's statement. Audio signal. For example, when the user U says "I want to register", the receiving terminal T1 of the telephone T transmits the audio signal and transmits it to the audio capturing mechanism 1 through telecommunication such as radio transmission or telecommunication line. Furthermore, the audio capture mechanism 1 can filter the background sound according to a decibel reference value of the environment background of the user U and extract the first voice audio data from the audio signal. Thereby, the noise of the environment background of the user U is filtered to improve the correct rate of capturing the first voice audio material. For example, when the noise of the environmental background is 50 decibels, and the voice volume change and frequency change of the user's speech are greater than the change of the noise, the audio capture mechanism 1 filters out the 50 dB signal of the noise according to the difference between the two. . In addition, the audio capture mechanism 1 can also convert the audio signal into the first voice audio data in the form of a digital signal for subsequent analysis processing. Of course, the present invention is not limited thereto, and may be converted into a digital signal when the first conversion mechanism 2 is performed, or may be directly extracted from the first audio and audio data and transmitted to the first conversion mechanism 2 for analysis processing without being converted into a digital signal.

The first conversion mechanism 2 is disposed in a remote server S in the embodiment, and after receiving the first voice audio data, the first voice audio data is recognized and converted into the first text data by using STT (Speech to Text). It is then transferred to the deconstruction mechanism 3.

The deconstruction organization 3 is a textual analysis organization that uses the grammatical structure analysis and the meaning analysis to analyze the first text data into various texts, deconstructs the pre-set elements contained in the first text data, and analyzes "I", "Yes" and " A text destructive component such as a registered number. And transmitted to the synthetic integration mechanism 4.

The integration integration mechanism 4 is a dialogue processing mechanism, which constructs an integrated text destruction component according to the dialog processing model. Specifically, the received text destruction components are compared with a database, and the received pre-determination is determined. Set the element to What, and determine the pre-set elements that have not yet been received, and then use the integration to give a corresponding response. For example, the words "I", "Yes" and "Registered" of the text destructor are combined to give a second text corresponding to the response, such as "What do you want to hang?", "I would like to ask if the date you want to register is Which day is equal to and transmitted to the second conversion mechanism 5.

The second conversion mechanism 5 and the first conversion mechanism 2 are disposed in the same remote server S in this embodiment, and the first conversion mechanism 2 and the second conversion mechanism 5 are a cloud processing mechanism, so the first conversion mechanism 2 and the second conversion mechanism 5 can be located in a data processing center outside the hospital and communicate with the organization of the service point P through the network. Of course, the present invention is not limited thereto, and the first conversion mechanism 2 and the second conversion mechanism 5 may also be a voice recognition device or a voice data library integrated with other mechanisms of the service point P. After receiving the second text data, the second conversion mechanism 5 converts the second text data into the second voice audio material by means of TTS (text to speech) and outputs it to the audio output mechanism 6.

The audio output mechanism 6 receives the second voice audio data and converts it into an analog signal for transmitting the telecommunication signal, and then transmits it to the sounding terminal T2 of the telephone T, and gives the user U a corresponding response, and then completes a dialogue. interactive. then. If the user U makes a further correspondence according to the content of the second voice audio material. For example, if the second voice and audio material is "What do you want to hang?", then the user U responds to "I want to hang the dermatology", then the voice customer service system 100 gives the corresponding response and repeats the same operation to make the next time. Dialogue interaction.

In addition, when the preset elements collected by the integrated integration mechanism 4 reach a preset number of items, the voice customer service system completes the entire dialogue interaction. For example, when user U starts talking, he directly states that "I want to hang the dermatology clinic of Dr. Huang Qixian in the afternoon of June 6. My name is Wang Daming, the identity card number is H123456789", and the integration integration agency 4 received it. Text deconstructing components such as "hanging", "afternoon of June 6", "doctor of ○○○", "dermatology clinic", "Wang Daming", "H123456789", etc., when the preset elements collected reach the preset items Number and content At this time, the synthetic integration unit 4 outputs the second text data that is the process of completing the entire voice registration. On the contrary, the second voice audio data corresponding to the unreceived preset element is output to query the user U answer, and then the audio capturing mechanism 1 re-takes a response from the next audio signal of the user U. The second voice audio material. Of course, the present invention is not limited thereto, and the preset elements may be more or different, and may be subdivided according to different large groups, for example, the presets that the text deconstructing component received by the integrated integration mechanism 4 under the registered group needs to reach. The elements are different from the preset elements under the query group. By means of the above grouping, the operation time of the user U and the voice customer service system can be shortened. Through the above technical means, the voice customer service system naturally communicates with the user through voice and recognizes the user's needs through voice recognition, and has the advantages of humanization and simple operation for the user. Moreover, because the user does not need to passively wait for the system problem, he can actively speak the demand first, greatly speed up the customer service speed, and also reduce the manpower of the customer service personnel and reduce the operating cost.

Referring to FIG. 3 and FIG. 4, the service place of the voice customer service system 100a of the service point according to the second embodiment of the present invention is set in a restaurant, such as a restaurant, a fast food restaurant, a delivery restaurant, and a beverage shop. The service point P' is a food ordering device, such as a food point of sale device, and both the first conversion mechanism and the second conversion mechanism are disposed together with the food and beverage outlet device. When the user U calls the restaurant, the voice customer service system 100a of the service point P' interacts with the user U. The user U can directly say what he wants to eat, when he comes to take it, etc., and the voice customer service system 100a actively raises questions about the content that the user U does not mention, and engages with the user U to develop a question and answer interaction. When the number of items and contents of the preset elements are collected, the ordering action is completed. Then, the catering point-of-sale device displays or prints information such as the content of the order and the amount of consumption to the staff of the restaurant to prepare. In this way, the entire ordering process can be accelerated.

Specifically, as shown in FIG. 4, when the user U calls the restaurant, the voice customer service system 100a of the service point P' outputs "XX beverage shop Hello, please Ask the voice and audio information of what service is required. User U says the audio signal "I want a cup of pearl milk tea". When the structural integration unit 4 receives the text deconstructing components of the pre-set elements such as "a cup" and "pearl milk tea", it is combined and integrated to form a second text, in which the preset elements of the first text data are also collected. The number of preset items is sufficient, so the content of the second text data is a preset element that requires insufficient items, such as "sugar amount", "ice amount", and the like. In this embodiment, the second text is "What is the amount of sugar and ice?". Then, after receiving the second audio and audio data corresponding to the second text data, the user U speaks the next audio signal of "half sugar, less ice", and the audio capturing mechanism 6 captures the next audio signal. Until the preset elements collected by the integration integration organization 4 reach a predetermined number of items, for example, the number of items of the preset elements is five, including the type of beverage, the amount of sugar, the temperature, and the basic information of the customer, the integrated integration mechanism 4 receives When the five preset elements are used, the content of the second text data outputted is to inform the user that the U has completed the current voice service. For example, the second text is "The order service has been completed, thank you for your call." In addition, when the first text data has a preset element, the content of the second text data may be a reconfirmed preset element, whether or not the first text data has a predetermined number of items. For example, the second text is "OK, and I will confirm it with you once, is it a cup of pearl milk tea, half sugar, less ice?". Through the voice customer service system of the present invention, the shortcomings of the conventional catering point-of-sale device are improved, so that the catering industry can improve the quality of the service through the voice customer service system, and at the same time save the manpower of the voice customer service.

The above description is only for the preferred embodiment of the present invention, and those skilled in the art can make other improvements according to the above description, but these changes still belong to the inventive spirit of the present invention and the patents defined below. In the scope.

100‧‧‧Voice Customer Service System

P‧‧‧ Service Point

S‧‧‧Remote Server

T‧‧‧Phone

T1‧‧‧ receiving end

T2‧‧‧ vocal

U‧‧‧Users

Claims (9)

A voice service system for a service point, the voice customer service system receives an audio signal sent by a user to confirm the information, and the voice customer service system includes: an audio capture mechanism for extracting the audio signal a voice audio material; a first converting mechanism connected to the audio capturing mechanism for converting the first voice audio data into a first text data; a deconstructing mechanism connecting the first converting mechanism to give the first text data Deconstructing into a text deconstructing component; a synthesizing integration mechanism, judging the received preset element according to the text deconstructing component, and determining an unreceived presetting element according to the received preset element, and Determining that the received preset element is combined with the preset element that has been determined to have not been received, and becomes a second text data; and a second conversion mechanism connecting the composite integration mechanism Converting the second text data into a second voice audio material, and outputting the second voice audio data, wherein the voice The system is set up to service a service point, the point of service to receive the audio signals sent by the user to confirm the information. The voice customer service system of the service point according to claim 1, wherein the first conversion mechanism and the second conversion mechanism are a cloud processing mechanism. The voice customer service system of the service point according to claim 1, wherein the audio capturing mechanism extracts the audio signal with respect to a decibel reference value and extracts the first voice audio material. The voice customer service system of the service point according to claim 1, wherein the structure is integrated and integrated according to the number and/or content of the preset elements included in the first text data. a voice customer service system for a service point as described in claim 4, wherein the composite integration The mechanism is that when the first text data is less than the preset element of the preset number of items, the first text data is combined and integrated into the second text data of the preset element whose content is an insufficient number of items. And the audio capture mechanism captures the next audio signal. The voice customer service system of the service point according to claim 4, wherein the first integration means is to integrate the first text data into a content system for reconfirmation when the first text material has the preset element The second text of the preset element. The voice customer service system of the service point according to claim 4, wherein the first integration means is to integrate the first text data into a unified element when the first text data has the preset number of items The content is the second text message that informs the user that the voice client service has been completed. The voice customer service system of the service point according to claim 1, wherein the audio signal is an audio signal from a telephone. The voice customer service system of the service point according to claim 1, wherein the service point is a food ordering device.